最新刊期
卷 58 ， 期 1 ， 2026
- “据最新报道，中国水利水电深部地下工程面临设计方法和围岩灾变问题，亟需突破关键技术挑战，为深部能源开发提供技术指导。”摘要：SignificanceAdvancing into the Earth’s deep interior has become a key direction in China’s current strategic scientific and technological development. As a crucial component of the national clean energy system, hydropower projects are extending their underground construction toward depths exceeding one kilometer and even deeper. These deep geological environments are typically characterized by extreme conditions such as high in-situ stress, elevated geothermal gradients, and high pore pressure. The associated disaster-inducing factors differ significantly from those in shallower zones, revealing the limitations of conventional engineering design methodologies and technical standards in such settings. Existing project statistics indicate that geological hazards, including rockbursts, spalling, and large deformations, occur frequently at depths greater than 600 meters. Moreover, the large-scale, complex spatial structure and high disturbance intensity of deep underground caverns amplify the risk of cascading failures in surrounding rock masses. Traditional depth-based classifications are no longer sufficient to fully characterize the high-energy geological environment encountered in deep hydraulic and hydropower projects. There is an urgent need to systematically redefine the boundaries of “deep” environments from the integrated perspective of stress, temperature, and pore pressure, and to establish corresponding disaster prevention and engineering design systems, thereby enhancing the scientific rigor and safety of deep underground hydropower construction.ProgressBased on a systematic review of how the concept of “deep” is defined across industries, including mining, transportation, hydraulic and hydropower, this study examines its relevance in the context of hydraulic and hydropower engineering, in which the unique structural and environmental characteristics of large-scale underground hydraulic structures have been taken into account. From this perspective, the study defines the notion of “deep” in hydraulic and hydropower underground projects through three interrelated dimensions: the evolution of stress-dominated mechanical regimes, the classification of high-temperature thermal hazards, and the progressive increase in pore pressure. Within this framework, underground spaces at depths of less than 600 meters, depths between 600 and 1 000 meters and depths exceeding 1 000 meters are respectively identified as conventionally controllable zones, hazard-intensified zones and ultra-deep complex zones, which require specialized technical assessment and justification. Building on this refined classification, the study further investigated the critical challenges existing throughout the full life cycle of deep hydraulic and hydropower projects. There are five core technical directions, where significant breakthroughs are urgently needed: 1) Precise exploration and real-time sensing during drilling: There is a need to develop directional drilling and synchronous coring technologies adapted to environments with high geothermal gradients and high pore pressures. These technologies should be integrated with sensors capable of withstanding extreme conditions to enable real-time transmission of deep geological data and inversion of dynamic parameters, thereby providing high-resolution geological constraints for the layout and design of underground cavern groups; 2) Design theory for large-scale deep-buried cavern groups: It is necessary to establish a coordinated design framework that incorporates tunnel axis optimization, spacing control, and structural synergy under true triaxial stress conditions. This theoretical system aims to overcome the limitations of traditional two-dimensional assumptions and enhance the adaptability of design strategies to the challenges posed by deep geological environments; 3) Engineering layout and intelligent construction: Advancing collaborative excavation technologies—such as stepwise and zoned excavation—alongside multi-parameter blasting optimization methods is essential. These should be supported by targeted support systems, including high-performance rock bolts, sprayed concrete, and composite linings, to ensure safe construction under conditions of high stress and elevated temperature; 4) Disaster prevention and proactive control: For typical deep geological hazards such as rockbursts, large deformations, and high-permeability seepage, it is critical to establish an integrated prevention and control framework based on the “source-pathway-receptor” model. This framework should incorporate technologies such as curtain grouting, pre-relief blasting, and zoned drainage and guidance systems to improve disaster response preparedness and promote coordinated stability between surrounding rock and supporting structures; 5) Smart operation, maintenance, and emergency evacuation: The construction of a multi-parameter coupled sensing system and a digital twin platform is necessary to establish a closed-loop management mechanism encompassing early warning, emergency response, and post-event recovery. This system should integrate monitoring of thermal hazards, seepage, and deformation with the deployment of escape routes and emergency power systems, thus enabling proactive safety management over the long service life of deep underground hydraulic and hydropower facilities.Conclusions and ProspectsAs hydraulic and hydropower underground projects extend into kilometer-scale depths, the types of risks, their underlying mechanisms, and the corresponding prevention strategies in design, construction, and operation undergo fundamental transformation. Within this context, a depth of 600 meters can be considered the upper applicability limit of current technical systems, whereas 1 000 meters represents a critical threshold that demands special attention. To address these challenges, a zoned and hierarchical engineering response framework should be established to provide more targeted solutions across five key technological domains. First, high-precision exploration and enhanced sensing of thermo-hydro-mechanical coupling parameters are essential to characterize deep geological environments accurately. Second, multi-objective collaborative design methods must be formulated under true triaxial stress conditions to improve structural adaptability. Third, intelligent construction equipment and processes should be developed, guided by multi-source information, to ensure safety and efficiency during excavation. Fourth, integrated risk-control strategies combining blockage, drainage, and pressure relief need to be deployed to achieve proactive disaster management. Fifth, smart operation and maintenance systems should be built on big-data platforms and predictive modeling to support long-term monitoring and resilience. In conclusion, these pathways are expected to substantially enhance engineering safety, operational reliability, and economic viability of deep underground hydraulic and hydropower projects under extreme geological conditions. In addition, they will provide replicable and transferable technical support for deep-earth resource development and the implementation of major infrastructure projects.关键词：hydraulic and hydropower engineering;deep engineering geology;geological environment;surrounding rock catastrophe;technical challenges182|234|0更新时间：2026-01-29
INVITED CONTRIBUTION
- “在新能源储能技术领域，专家构建了抽水蓄能电站综合效益评价模型，为项目决策提供理论支持和实践依据。”摘要：ObjectiveThe comprehensive evaluation of pumped storage power plants is of critical importance for ensuring that these systems, which play a pivotal role in grid regulation, renewable energy integration, and the achievement of national carbon reduction targets, are operating optimally. In practice, however, evaluation processes often suffer from incomplete indicator systems, imbalanced subjective and objective weight assignments, and a lack of robust methodologies capable of adequately reflecting uncertainties and fuzziness in complex systems. Therefore, the present study proposes a novel comprehensive benefit evaluation method for pumped storage power plants, which integrates game theory-based combined weighting with cloud model theory. The objective is to develop a systematic, objective, and adaptable evaluation framework that provides scientific support for decision-making and operational optimization in pumped storage power plants.MethodsFirstly, based on an extensive literature review, this study employed social network analysis (SNA) to screen the key indicators required for the comprehensive benefit evaluation of pumped storage power plants. Through centrality analysis, it removed redundant indicators. It merged overlapping concepts, which ultimately formed an evaluation indicator system comprising eight primary categories (financial evaluation, national economic evaluation, technical benefits, dynamic benefits, static benefits, grid benefits, comprehensive sustainability, and social benefits) and 30 corresponding secondary indicators. Secondly, it used a nine-level linguistic operator to quantify the relative importance of adjacent indicators, and it applied a recursive formula to compute subjective weights. The CRITIC method quantified indicator variability and conflict by calculating the standard deviation and correlation coefficient, generating objective weights. Then, game theory integrated and optimized the two sets of weights by constructing an objective function that minimized the deviation between subjective and objective weights. The optimal combination coefficient was then determined to achieve a dynamic balance in weight allocation. Finally, a comprehensive benefit evaluation model was developed based on cloud model theory. Expert rating data for each indicator were first normalized, and an inverse cloud generator computed the cloud numerical characteristics (expectation Ex, entropy En, and hyper-entropy He) to capture the inherent uncertainty of the data. Then, using the predetermined combination weights of secondary indicators, the cloud numerical characteristics of primary indicators, and the overall evaluation were synthesized step by step. A forward cloud generator in Matlab generated a comprehensive benefit cloud map, and the effectiveness level was determined through spatial matching and the maximum closeness criterion. In addition, a case study on the Ziyunshan pumped storage power plant was conducted to validate the practical applicability of the proposed model.Results and DiscussionsThe evaluation results of the comprehensive benefits of the Ziyunshan pumped storage power plant showed that its overall evaluation cloud diagram (Ex=7.026 0, En=0.749 4, He=0.271 6) closely aligned with the standard cloud diagram for “good overall benefits” (Ex=7.000 0, En=0.670 0, He=0.300 0), verifying the effectiveness of the constructed model. The analysis of key indicators revealed that the peak shaving and valley filling coal-saving benefit (Ex = 8.400 0) fell within the good range, with an estimated annual reduction of 411 000 tons of coal consumption for thermal power generation, which verified the pumped storage power plants' pivotal role in load balancing and carbon reduction. The black start capability (Ex = 7.600 0) fell within the good benefit range, with an actual response time 75% shorter than that of conventional thermal power plants, which highlighted its technical advantage in enhancing grid resilience. However, the annual start-up frequency (Ex = 5.600 0) reached only 68% of its designed value, reflecting the inadequate adaptability of dispatch strategies to load fluctuations. In terms of financial and economic evaluation, the loan repayment period (Ex = 5.600 0) and asset-liability ratio (Ex = 5.000 0) corresponded to a repayment term of 15 years and a high debt ratio of 65%, revealing long-term debt repayment pressure and capital structure risks. Regarding comprehensive sustainability benefits, the expectation values of environmental indicators, such as impacts on water quality, noise, and air, were all below 6.0, classifying them in the average benefit range. Although the construction wastewater reuse rate reached 85%, the suspended sediment concentration in the reservoir area remained as high as 0.318 kg/m³, and noise levels in sensitive areas exceeded the standard by 10 dB, indicating the need for further optimization of environmental management measures. These results indicated that although pumped storage power plants demonstrated significant operational benefits, continuous improvements in environmental management measures were still required to comply with stricter environmental regulations.ConclusionsThe study concludes that the proposed evaluation method, which integrates game theory-based combined weighting with cloud model analysis, provides a scientifically rigorous and robust framework for assessing the comprehensive benefits of pumped storage power plants. Its successful application to the Ziyunshan pumped storage power plant case study confirmed that the evaluation results closely align with actual operational performance, validating the effectiveness and reliability of the approach. Accordingly, the integration of advanced weighting techniques with fuzzy quantitative modeling enhances the objectivity of the evaluation and provides meaningful insights for identifying key areas for improvement. These findings indicate that the method has significant potential for broader application in evaluating energy storage systems and other complex engineering projects, ultimately contributing to more informed decision-making and improved operational efficiency.关键词：pumped storage power plant;game theory;cloud model;comprehensive benefit evaluation;social network analysis method;G1 method;CRITIC method199|168|0更新时间：2026-01-29
- “在地质学领域，专家建立了三维离散裂隙网络模型，模拟断层剪切过程，揭示了裂隙网络剪切损伤演化及其对渗流行为的控制机制，为地质研究开辟了新方向。”摘要：ObjectiveAccurately predicting permeability evolution during fault shear in fractured rock masses under constant normal stiffness (CNS) boundary conditions is of fundamental importance for the design and optimization of geothermal reservoir stimulation, subsurface energy extraction, and fluid management in deep rock engineering. In enhanced geothermal systems (EGS), hydraulic stimulation commonly induces shear slip along pre-existing faults and fracture networks, resulting in complex and competing hydraulic responses. On one hand, fault shear dilation can significantly enhance permeability by generating preferential flow channels; on the other hand, the associated stress redistribution and normal stress amplification may induce closure and damage of surrounding fracture networks, leading to permeability degradation. Despite extensive experimental and numerical studies on shear‒flow coupling in single fractures, the permeability evolution of three-dimensional discrete fracture networks (DFNs) interacting with a shearing fault under CNS conditions remains poorly understood. Existing studies often focus on either fault permeability enhancement or fracture closure effects in isolation, while neglecting the coupled competition between shear-induced dilation and network-scale compression damage. Moreover, the influence of aperture anisotropy, boundary stiffness, and stress constraints on flow anisotropy and channelization patterns has not been systematically quantified. The objective of this study is therefore to quantitatively investigate the dual mechanisms governing permeability evolution during fault shear under CNS conditions: 1) permeability enhancement driven by fault shear dilation and channelization, and 2) permeability reduction induced by compression damage of the surrounding fracture network. Particular emphasis is placed on evaluating the effects of fracture aperture anisotropy, normal stiffness, and boundary stress on damage evolution, permeability anisotropy, and flow partitioning between faults and fracture networks.MethodsA three-dimensional DFN model incorporating a through-going fault was developed to simulate shear‒flow coupling processes under CNS boundary conditions. The DFN consists of multiple randomly oriented fractures with statistically prescribed aperture distributions, embedded within a cubic rock domain. Fracture apertures were assumed to follow a truncated Gaussian distribution with spatial correlations, characterized by a mean aperture μ0 ranging from 1 to 4 mm and a standard deviation σ0 ranging from 0.3 to 1.2 mm. This formulation captures the inherent heterogeneity and anisotropy of natural fracture apertures while avoiding unrealistically negative values. Fault shear was simulated by imposing incremental shear displacements (uₛ: 0 to 200 mm) under CNS constraints, such that normal stress evolved dynamically in response to shear-induced dilation. The mechanical response of fractures to normal stress was governed by a nonlinear hyperbolic closure relationship, enabling progressive aperture reduction and damage accumulation in the DFN during shear. A fracture damage rate (RD), defined based on the reduction of average fracture aperture relative to the maximum allowable closure, was introduced to quantify the degree of DFN damage at each shear stage. Fluid flow within the DFN and fault was modeled using the Reynolds equation, assuming laminar flow through rough-walled fractures. Numerical simulations were conducted in COMSOL Multiphysics, with flow applied along the x-, y-, and z-directions to evaluate permeability anisotropy. Equivalent permeability coefficients were calculated using the cubic law, and permeability evolution ratios were derived relative to the initial, undeformed state. To assess the influence of mechanical boundary conditions, a parametric study was conducted by varying the normal stiffness (kₙ: 0.25 to 1.00 GPa/m) and boundary stress (σy: 1 to 4 MPa). Flow partitioning between the fault and the DFN was quantified by calculating the proportion of total volumetric flux transmitted through each component, allowing for detailed analysis of channelization and hydraulic dominance.Results and DiscussionsSimulation results reveal that permeability evolution during fault shear is governed by a pronounced competition between fault-induced permeability enhancement and DFN compression-induced permeability reduction. This competition is strongly modulated by fracture aperture anisotropy, mechanical boundary conditions, and flow direction. Increasing aperture standard deviation σ0 significantly reduced the DFN damage rate RD across all shear stages. For example, at μ0 = 1 mm and us = 200 mm, RD decreased from 34.68% for σ0 = 0.3 mm to 12.35% for σ0 = 1.2 mm, representing a reduction of approximately 64%. This effect arises from the increased heterogeneity of fracture apertures, which promotes stress redistribution and limits the spatial extent of fracture closure. Highly heterogeneous apertures also facilitate localized channelized flow, resulting in pronounced groove flow patterns within both the DFN and the fault. Normal stiffness kₙ and boundary stress σy exerted a strong control on DFN damage evolution. Higher kₙ and σy amplified normal stress accumulation during shear, leading to accelerated fracture closure and increased RD. The most significant damage increments, reaching up to 45%, occurred during the initial shear stage (uₛ: 0 to 40 mm), corresponding to rapid stress build-up under CNS conditions. Beyond this stage, damage accumulation gradually stabilized as shear progressed into a residual regime. Flow simulations demonstrated marked permeability anisotropy induced by fault shear. Permeability along the shear-parallel z-axis increased by two to three orders of magnitude due to fault dilation and the development of continuous high-aperture channels. In contrast, permeability along the x- and y-directions decreased by approximately 60% to 80%, reflecting dominant DFN compression and loss of interconnected flow pathways. Flow partitioning analysis showed that fluid progressively concentrated within the fault as shear displacement increased. At uₛ = 200 mm, more than 94% of the total flow was transmitted through the fault for most boundary conditions, rendering the surrounding DFN hydraulically negligible. This dominance of fault channel flow was further enhanced by higher kₙ and σy, which suppressed DFN permeability while promoting fault-controlled transport. Notably, at large shear displacements (us > 160 mm), shear-induced stresses significantly exceeded the imposed boundary stress σy, resulting in a diminished influence of σy on permeability evolution. This indicates a transition from boundary-controlled to shear-dominated hydraulic behavior, with important implications for long-term reservoir performance.ConclusionsThis study provides a comprehensive quantitative framework for understanding permeability evolution in fractured rock masses during fault shear under CNS conditions. The results highlight that permeability enhancement and reduction mechanisms coexist and compete throughout the shear process, with their relative dominance controlled by fracture aperture anisotropy, normal stiffness, and boundary stress. From an engineering perspective, the findings suggest that stimulation strategies in geothermal reservoirs should explicitly account for DFN damage induced by shear-related stress amplification, rather than assuming monotonic permeability enhancement. High aperture heterogeneity and controlled shear displacements may be leveraged to promote stable channelized flow while minimizing network-scale permeability loss. The strong localization of flow within faults at large shear displacements also underscores the need to manage fault-dominated flow paths to avoid premature thermal breakthrough or uneven reservoir depletion.关键词：fractured rock mass;permeability;shear;discrete fracture network110|112|0更新时间：2026-01-29
- “在岩石破坏强度预测领域，专家基于摩尔应力圆理论建立了非线性Mohr-Coulomb强度准则，考虑中主应力影响，验证了其在复杂应力状态下的适用性和工程应用价值。”摘要：ObjectiveAccurate prediction of the strength of deeply buried hard rock is crucial for ensuring the safety and stability of deep rock engineering. Currently, two key factors influencing the strength prediction of deep-buried hard rock are high stress and the intermediate principal stress. High stress produces nonlinear deformation characteristics in deeply buried hard rock, and the intermediate principal stress substantially affects the deformation and failure behavior of hard rock. This study proposes a modified Mohr‒Coulomb strength criterion for hard rock, which simultaneously considers the nonlinear strength characteristics under large-range variations in confining pressure and the influence of the intermediate principal stress on rock strength.MethodsFirst, an analysis of Mohr’s stress circle under true triaxial conditions was conducted. Based on the theory of Mohr’s stress circle, it was determined that rock strength comprised cohesive and frictional strength components. The intermediate principal stress significantly influenced the cohesive strength component, while its effect on the frictional strength component remained negligible. A generalized functional form of the Mohr‒Coulomb (M‒C) strength criterion under true triaxial stress conditions was proposed, which attributed the influence of the intermediate principal stress on rock strength to an intermediate principal stress effect function. The rationality of this functional form was preliminarily validated using true triaxial test data from four types of hard rock. Then, an analysis was conducted on commonly used rock strength criteria that reduced to the M‒C criterion under certain conditions, including the D‒P criterion, the Mogi criterion, and the modified Lade criterion. The results revealed that the significant deviations in their strength predictions primarily resulted from inaccuracies in the mathematical characterization of the intermediate principal stress effect function. Therefore, the specific forms of the intermediate principal stress effect functions corresponding to the D‒P, Mogi, and modified Lade criteria were examined in detail. On this basis, a parameter η, which represented the difference between the tensile and compressive strengths of the rock, and a parameter γ, which characterized the influence of the intermediate principal stress (σ2) on rock strength under constant minimum principal stress (σ3), were introduced. A new intermediate principal stress effect function was constructed. The classical M‒C criterion was modified, establishing a nonlinear M‒C strength criterion that accounted for the influence of the intermediate principal stress by integrating the critical state characteristics of rock. Then, methods for determining the parameters were provided. Finally, the applicability of the new criterion was validated using true triaxial test data from six types of hard rock, and a comparative analysis was conducted with four traditional strength criteria.Results and DiscussionsPredictions from this criterion were compared to true triaxial test data from six types of hard rock, along with a comparative analysis against four traditional strength criteria to validate the applicability of the proposed modified M‒C strength criterion that considered the influence of the intermediate principal stress. The results demonstrated that the proposed criterion effectively captured the observed trend in which rock strength initially increased and then decreased with the rise of the intermediate principal stress, as well as the significant strength differences exhibited under the conditions of σ2=σ3 and σ2=σ1. Under different confining pressure σ3 conditions, the variation patterns captured by this strength criterion remained consistent. However, as the ∂σ1/∂σ3 ratio gradually decreased, the increase in rock strength became more moderated with rising σ3, indicating that the enhancing effect of the intermediate principal stress on rock strength exhibited a decaying characteristic with increasing stress levels. When σ2=σ3, the criterion reduces to a nonlinear M‒C strength criterion that accounts for the critical state of rock. With increasing σ3, the lateral deformation constraint effect continuously strengthened, which led to a persistent increase in the curvature of the strength envelope and reflected the transition of rock mechanical behavior from brittle to ductile. When σ3
- “勾头透水丁坝研究揭示了其结构设计对河道缓流促淤的影响，为工程应用提供理论参考。”摘要：ObjectiveThe selection of the hook-head permeable spur dike's structural design and body parameters is of great importance for reducing river flow velocity and promoting siltation. Compared to conventional dike designs, the permeable design exhibits enhanced coordination and stability. It effectively overcomes the constraints of solid dike design in complex engineering applications. In addition, it enhances the flow structure surrounding the dike, mitigates scouring and siltation, and increases the navigability and flood control capacity of the river. However, research on the effects of body parameters on water flow structure remains limited. This study presents an in-depth investigation of the influence of hook-head angle and length modifications on water flow structure around the dike. The objective is twofold: to provide a theoretical foundation for optimizing body parameter design of the hook-head permeable spur dike in practical projects and to raise the adoption and advancement of this dike type.MethodsThe research utilized a physical model test with a hook-head permeable spur dike. An ADV was employed to accurately measure the three-dimensional instantaneous flow velocity around the dike under different hook-head angles and lengths. The test apparatus consisted of a linear open channel, in which the permeable holes of the spur dike model were arranged in a double-layer rectangular configuration, characterized by a controlled permeability of 20%. The measurement sections, which numbered five in total, were positioned circumferentially around the spur dike, with 22 measurement points established at each section. Each measurement point underwent continuous measurement for 30 seconds. The collected data underwent processing, during which the theoretical calculation methods were applied to calculate the slow flow rate, flow field uniformity, turbulence intensity, and turbulence energy around the spur dike. The dimensionless processing was conducted to enhance the intuition of the data. The study further analyzed the specific effects of different body parameters on the flow velocity, slow flow rate, flow field uniformity, turbulence intensity, and the change rule of turbulence energy.Results and DiscussionsThe findings of the research indicated that the hook-angle and length of the permeable spur dike had a substantial impact on the flow structure downstream of the dike when integrated with the hook head design. It was observed that as the hook-angle increased, the flow velocity downstream of the dike decreased, while the turbulence intensity and turbulence energy increased. Specifically, when the hook-angle was set at 60°, rather than 30° and 45°, the flow velocity downstream of the dike body decreased by 2.87% and 1.61%, respectively. Simultaneously, the retardation rate increased by 6.79% and 3.69%, the turbulence intensity increased by 18.95% and 7.58%, and the turbulence energy increased by 27.39% and 5.93%, respectively. Increases in hook-length resulted in decreases in downstream flow velocity and turbulence intensity of the dike body. At hook-length D/2, the flow velocity downstream of the dike body decreased by 8.75% and 3.11%, and the turbulence intensity decreased by 3.86% and 1.97%, compared to D/4 and D/3. In the downstream reflux zone and transition zone, the flow velocity fluctuation was characterized by an "M" type distribution. The turbulence intensity fluctuation was most intense in the downstream reflux zone, and the variation in the transition zone reached a maximum, which was about 1.7 to 2.3 times that in the reflux zone. The turbulence energy was influenced by the hook-angle, with a maximum observed in the area between the permeable hole and the outer part of the dike head. As the hook-angle increased, the peak value gradually deflected toward the right bank. In addition, compared to the conventional permeable straight dike, the average flow velocity in the downstream near-dike area of the hook-head permeable dyke dike decreased by approximately 17%, which substantiated the effectiveness of the hook-head permeable dyke dike in reducing flow velocity.ConclusionsThe study demonstrated a negative correlation between the downstream flow velocity of the dike and the hook-angle and length. As the hook-angle and length increase, the slowing effect of the spur dike on the water flow becomes more pronounced. In contrast, the flow velocity in the downstream main stream area exhibited an upward trend, indicating that increasing hook-angle and length can enhance river navigability and improve the flow environment. In addition, the turbulence intensity downstream of the dike was found to be positively correlated with the hook-angle and negatively correlated with the hook-length. Therefore, an increase in hook-angle aggravates the disturbance of the water flow by the spur dike, resulting in higher turbulence intensity. An appropriate extension of the hook-length can weaken the turbulence intensity of the water flow and raise its movement toward a more stable state.关键词：hook-head permeable spur dike;hook-angle;hook-length;water flow structure;turbulence intensity79|125|0更新时间：2026-01-29
- “在非饱和土体湿度场迁移领域，专家提出了动态平衡饱和度等预测公式，揭示了湿度场变化规律，为边坡稳定性评估提供理论基础。”摘要：ObjectiveThe increasing incidence of rainfall-induced embankment landslides corresponds with the recent rise in extreme rainfall events. During precipitation, moisture field migration occurs within unsaturated embankment soils, which modifies matric suction distribution and soil strength characteristics, threatening embankment stability. Accurately characterizing the spatiotemporal distribution pattern of the water content field and predicting its moisture field migration mechanism is a key prerequisite for advancing slope stability evaluation methods and optimizing geological disaster prevention strategies. Although current research extensively documents macroscopic principles governing moisture field movement, the prediction equation of the moisture field in soil remains scarce and inadequately validated. Prevailing analytical models predominantly assume complete soil saturation behind wetting fronts, but actual field conditions contradict this premise, which influences model accuracy. Therefore, the moisture field migration prediction model is proposed to reveal the migration pattern of the soil moisture field and provide a theoretical basis for improving slope instability warning methods.MethodsFirstly, using a self-developed one-dimensional soil column test system, typical embankment filling sand was selected as the research material, and an unsaturated soil column structure was constructed through layered compaction. The test adopted the rainfall intensity control method, with 35 mm/h as the benchmark rainfall intensity, and simulated rainfall conditions using an axial uniform spraying system. Moisture sensors were installed along the depth of the soil column to monitor the dynamic response characteristics of the volumetric moisture content of different soil layers in real time to accurately and continuously capture the moisture migration process. Secondly, using COMSOL finite element software, the modified Cambridge model, unsaturated empirical formulas, and the VG model were introduced to establish one-dimensional soil column models, and the numerical model was validated using the test results. The variation law of the moisture field was preliminarily revealed through both the model test and numerical simulation. Thirdly, based on the indoor test and simulation results, the main influencing factors of the soil dynamic equilibrium saturation degree were analyzed, and a prediction formula for the soil dynamic equilibrium saturation degree was proposed and validated through the numerical simulation model. Then, based on the assumption of a homogeneous soil model, the quantitative relationship between the rainfall volume and rainwater retention in soil was analyzed, the soil dynamic equilibrium saturation prediction formula was introduced, and the prediction formula for the dynamic equilibrium saturation of unsaturated soil was established and verified through indoor tests and numerical simulations. In addition, the quantitative relationship between rainfall volume and the amount of rainwater required to achieve full soil saturation, as well as the prediction formula for groundwater level rise, was established, and the feasibility of these prediction formulas was verified through experiments and numerical models. Finally, based on the above formulas, the effects of wet front migration and groundwater level rise under different compaction degrees (0.75, 0.80, 0.85, and 0.90) and rainfall intensities (5, 10, 15, 20, 25, 30, and 35 mm/s) were discussed, and the migration pattern of the unsaturated soil moisture field under different rainfall intensities was revealed, providing an important theoretical basis for embankment landslide warning.Results and Discussions1) The results of the model tests and numerical simulations showed that the changes in soil volumetric water content under rainfall exhibit five distinct stages, namely the initial stage, the first growth stage, the dynamic equilibrium stage, the second growth stage, and the complete saturation stage. Among them, the first growth stage occurs due to the influence of rainfall on the soil, during which the wetting front migrates from the soil surface to the bottom and causes a gradual increase in soil volumetric moisture content from top to bottom. After the wetting front passes through the soil, the volumetric water content enters the dynamic equilibrium stage. As the groundwater level rises from below, the volumetric water content enters the second growth stage, and eventually, all soil reaches the fully saturated stage. 2) The feasibility of the dynamic equilibrium saturation prediction formula, the wetting front migration prediction formula, and the groundwater level fluctuation prediction formula for unsaturated soil under rainfall conditions was double verified through indoor tests and numerical simulations. The results showed that the maximum error between the dynamic equilibrium saturation prediction formula and the simulation was 4%, the maximum error between the wetting front migration prediction formula and both the simulation and test was 8.5%, and the maximum error between the groundwater level fluctuation prediction formula and both the simulation and test was 8.5%. The trend of the humidity field change curve calculated by these formulas was consistent with the experimental and numerical simulation results. Therefore, these formulas better predicted the migration behavior of the soil moisture field and were suitable for studying unsaturated soil moisture field migration. 3) The unit pore volume, soil permeability, and soil compaction degree of unsaturated soil showed a clear inverse relationship. The reduction in unit pore volume accelerates the migration of the unsaturated soil moisture field, whereas the decrease in soil permeability slows the migration speed of the moisture field. When the soil compaction degree is less than or equal to 0.80, the migration of the humidity field is mainly controlled by the unit pore volume, and the migration speed increases with increasing compaction degree. When the soil compaction degree exceeds 0.80, the migration of the moisture field is mainly controlled by soil permeability, and the migration speed decreases with a higher compaction degree. 4) The migration speed of the unsaturated soil moisture field showed a positive correlation with rainfall intensity, indicating that the migration speed increases as rainfall intensity increases. However, as rainfall intensity continues to increase, its rising effect on the migration speed of the humidity field gradually weakens. 5) At any given moment, the influence of soil compaction and rainfall intensity on the migration of the humidity field remains consistent and does not change with longer rainfall duration.ConclusionsThe study demonstrates that the dynamic equilibrium saturation prediction formula, wetting front migration prediction formula, and groundwater level fluctuation prediction formula effectively predict soil moisture field migration. In addition, the effects of soil compaction degree and rainfall intensity on moisture field migration are clarified. The findings are expected to provide a theoretical basis for improving slope instability warning methods and enhancing the capacity for slope stability assessment and disaster prevention.关键词：moist front;groundwater level;dynamic equilibrium saturation prediction;humidity field migration prediction;compaction degree;rainfall intensity98|123|0更新时间：2026-01-29
- “在土质边坡可靠度分析领域，专家采用逐步协方差矩阵分解方法和改进粒子群算法，为3维土质边坡失效概率分析提供高效方法。”摘要：ObjectiveThree-dimensional (3D) slope reliability analysis encounters two primary challenges: 1) the stochastic modeling of soil properties involves complex spatial variability, requiring extensive data processing and imposing substantial computational demands; and 2) brute-force trial-and-error methods for locating critical failure surfaces are computationally inefficient, hindering real-world implementation despite the need for precise stability evaluations in civil infrastructure. Therefore, this study employs the covariance matrix decomposition method to generate 3D lognormal random fields for soil parameters, enabling efficient modeling of spatial variability. The particle swarm optimization (PSO) algorithm is refined with enhanced termination criteria and integrated with the 3D Bishop method to search for the minimum factor of safety (F). This approach significantly improves the accuracy and computational efficiency of slope reliability analysis.MethodsThe core of this framework lies in employing a stepwise covariance matrix decomposition method to generate 3D lognormal random fields for slope soil parameters. The stepwise covariance matrix decomposition efficiently partitions the problem into tractable components, drastically reducing computational demands while enabling high-resolution random field generation with limited resources by systematically decomposing the covariance matrix. This decomposition facilitated rapid simulation of spatially variable soil parameters without excessive overhead. An enhanced PSO algorithm was proposed for 3D slope stability analysis to complement the random field modeling. The algorithm integrated the 3D Bishop method, a well-established limit equilibrium technique, into PSO's global search mechanism, combining PSO's exploratory capacity with the Bishop method's precise stability calculations. This interaction enhanced the identification of critical slip surfaces. Refined iteration termination criteria were incorporated to expedite convergence toward the minimum factor of safety (F) and the corresponding critical surface, improving computational efficiency for reliability analysis. The failure probability of 3D slopes was assessed through Monte Carlo simulation, which accounted for inherent soil property uncertainties. The method's accuracy and effectiveness were validated through numerical examples that systematically analyzed diverse slip surface geometries (for example, cylindrical and cylindrical-ellipsoidal combinations).Results and DiscussionsThe proposed method was applied to analyze Example 1. Excluding spatial variability of soil parameters and assuming a cylindrical sliding surface, a deterministic analysis yielded an F of 1.352 4, which closely matched previous result of 1.352 5. For a sliding surface combining cylindrical and ellipsoidal shapes, F decreased as the failure surface width (B) increased, gradually approaching the two-dimensional analysis outcome. When spatial variability of soil parameters was incorporated, the calculated failure probability (Pf) range of 0.090 5‒0.091 6 (average 9.11%) for Example 1 closely aligned with previous finding of 9.25%, with a relative error of approximately 1.5%. The comparative analysis of safety factors and failure probabilities confirmed the accuracy of the proposed method and the computational procedure. Systematic studies were conducted for various slip surface forms that encompassed combinations of cylindrical and cylindrical + ellipsoidal shapes, as well as different correlation lengths and coefficients of variation. When the sliding surface was cylindrical, and the characteristic length scale in the out-of-plane direction (ly) was set to 20 m, Pf gradually decreased as B increased. This observation indicated that the spatial variability of soil parameters was one of the pivotal factors contributing to the enhanced stability assessment accuracy of 3D slope analysis compared to its two-dimensional counterparts. When ly approaches infinity (ly→+∞), as B increases, Pf gradually rises while the average safety factor (Fav ) decreases. These findings indicated that B significantly influences the stability of 3D slopes. When the coefficient of variation for cohesion (Vc) increased from 0.1 to 0.6, Pf rose from 0.0 to 0.25, reflecting an approximate increment of 0.25. When the coefficient of variation for the internal friction angle (Vφ) increased from 0.1 to 0.4, Pf increased from 0.04 to 0.14, corresponding to an approximate increment of 0.10. When the correlation coefficient between cohesion and the internal friction angle (ρc,φ) increased from -0.5 to 0.5, Pf rose from 0.055 to 0.070, indicating an approximate increment of 0.015. These findings demonstrated that Vc, Vφ, and ρc,φ all exert significant influences on the failure probability. When the characteristic length scale in the x-direction (lx) increased from 10 m to 40 m, Pf increased by approximately 0.03. In contrast, when ly increased from 10 to 40 m, Pf increased by approximately 0.075. This result indicated that ly had a more pronounced impact on slope failure probability compared to lx. In addition, when the characteristic length scale in the z-direction (lz) increased from 1 to 4 m, Pf increased by approximately 0.05, indicating an influence on failure probability.ConclusionsThe results demonstrate that the enhanced PSO algorithm, when integrated with the 3D Bishop method, markedly improves the computational efficiency of 3D slope reliability analysis without compromising accuracy. When only spatial variability is considered, the failure probability decreases as the width of the failure surface increases. In contrast, when only the 3D effect is considered, the failure probability increases with the width of the failure surface. When both factors are considered simultaneously, their effects on the failure probability tend to offset each other, resulting in a significantly lower failure probability for 3D slopes compared to 2D slopes. When accounting for the variability of soil parameters and 3D effects, the failure probability of slopes initially increases and then decreases with the widening of the failure surface. This pattern indicates that 3D effects dominate when the failure surface is narrow, because geometric constraints and stress redistribution play a critical role. In contrast, spatial variability of soil properties becomes the primary factor when the failure surface exceeds a critical width, because material heterogeneity governs the failure process. The analysis reveals a critical failure surface width that marks the transition from geometry-controlled to material-controlled failure mechanisms in slope stability assessment.关键词：soil slope;failure probability;CMD;three-dimensional effect;spatial variability149|102|0更新时间：2026-01-29
- “在抽水蓄能电站建设领域，专家提出了考虑外水压力折减系数的深孔高压压水试验参数计算方法，为优化设计方案和节约工程造价提供重要参考。”摘要：ObjectiveThe accuracy of high-pressure water injection test parameters in deep boreholes is crucial for the construction of pumped storage power stations. Currently, conventional computing methods for water injection test parameters are still used for high-pressure scenarios, leading to significant errors in deep borehole applications. This paper optimizes the calculation of additional hydraulic pressure by introducing an external water pressure reduction coefficient, thereby enhancing the reliability of deep borehole high-pressure water injection test parameters and providing more reliable guidance for engineering design.MethodsFirstly, the calculation method stipulated in the current specifications is analyzed. By examining three groundwater level scenarios: above the test section, within the test section, and below the test section, it is identified that the additional hydraulic pressure calculation is based on the assumption that “the external water pressure exerted by the borehole water on the test section equals its full hydrostatic pressure.” This overestimated external water pressure is the primary cause of parameter errors in deep-borehole high-pressure water injection tests. For optimization, an external water pressure reduction coefficient is introduced to recalibrate the additional hydraulic pressure calculation under all three groundwater conditions. This adjustment brings the external water pressure closer to actual field conditions and yields more rational computational parameters. Finally, the method is validated through a case study at a pumped storage power station. The reduction coefficient is derived by interpolating permeability rates from water injection tests in deep boreholes. The revised approach calculates key parameters, including permeability rate and splitting pressure under high-pressure water injection test conditions, and demonstrates significant improvements compared to standard calculation results.Results and DiscussionsThe current specification for the pressure calculation baseline in high-pressure water injection tests is based on the assumption that the external water pressure reduction coefficient is not considered, which may lead to non-negligible errors in deep borehole parameter calculations. According to the revised additional hydraulic pressure calculation formula incorporating the external water pressure reduction coefficient, when the borehole water level is above or within the test section, additional hydraulic pressure is influenced by the external water pressure reduction coefficient. Generally, additional hydraulic pressure calculated with the external water pressure reduction coefficient is greater than that specified in current standards, and shows a negative correlation with the external water pressure reduction coefficient. When the groundwater level is below the test section, additional hydraulic pressure remains unaffected by the external water pressure reduction coefficient. The calculation formula is applied to the deep hole of a pumped storage power station in Shaanxi province and compared with the standard method. The results show that: In intact to moderately intact granite formations, the failure pattern of the pressure‒flow (P‒Q) curve in high-pressure water injection tests deviates from the conventional five-type classification. Instead, it exhibits a composite characteristic of “laminar flow in the initial phase and cracking behavior in the later phase”. The pressure‒flow‒time (P‒Q‒t) curve demonstrates distinct stages demarcated by the splitting pressure: during the first stage, incremental pressure increases yield no significant flow rate growth, while the second stage features a brittle-failure pressure drop accompanied by a sharp flow surge. Permeability rates calculated via the standard method for conventional water injection tests range 0.03～3.85 Lu (average: 0.20 Lu). In contrast, the reduction coefficient method yields 0.01～1.26 Lu (average: 0.11 Lu), representing reductions of 3.7%～72.8% (average: 44.1%) compared to the standard method. The reduction magnitude increases with borehole depth, notably reaching 71.1% for average permeability in powerhouse sections. The water permeability of the high-pressure water injection test is 0.19～0.26 Lu by the standard method, and 0.12～0.14 Lu by the reducing coefficient method, which is 36.0%～41.8% lower than the former. The ratio of the high-pressure water injection test to the conventional water injection test is 12.0 to 16.9 times, the water permeability of the high-pressure water injection test is much higher than that of the conventional one. The splitting pressure calculated by the standard method ranges from 5.52 to 7.81 MPa, and the splitting pressure calculated by the reduction coefficient method ranges from 10.14 to 12.20 MPa. The value of the splitting pressure obtained by the reduction coefficient method is much higher than that by the standard method. This study experimentally validates the impact of the external water pressure reduction coefficient on high-pressure water injection tests in intact to moderately intact granite formations, however, this method is also feasible in other diverse geological strata.ConclusionsCompared to the standard method, the modified reduction coefficient approach enables more accurate characterization of rock mass permeability and fracturing pressure, thereby effectively mobilizing the rock’s anti-seepage potential. This methodology provides critical guidance for optimizing designs and reducing project costs. The construction of pumped storage power stations is experiencing rapid expansion, accompanied by a prevailing trend of increasingly deeper boreholes, the accuracy of high-pressure water injection test parameters obtained from deep boreholes holds critical significance for pumped storage projects. It is strongly advised that external water pressure reduction coefficients be fully incorporated into high-pressure water injection test calculations for pumped storage power stations. By comparatively analyzing the impacts of both current code methods and the reduction coefficient method on engineering design, a more scientific balance between design safety and construction costs can be achieved, ultimately enhancing the quality of survey and design works for pumped storage power stations.关键词：pumped storage power station;High-pressure water injection test;External water pressure reduction coefficient;Splitting pressure;Water permeability rate107|44|0更新时间：2026-01-29
- “最新研究突破大坝变形监测难题，基于样本熵的自适应预测模型精度高、适用性强，为大坝安全运行提供重要理论支撑。”摘要：ObjectiveDeformation monitoring is one of the most direct and critical methods for assessing dam safety, and numerous studies have shown that deformation data typically exhibit significant changes prior to structural damage. The operating environment of high dams and large reservoirs is complex, and accurately predicting dam deformation is essential for timely identification of potential risks. However, the accuracy of deformation prediction is influenced by several factors, including data type, quantity, quality, patterns, and the prediction model itself. Research indicates that different models perform differently depending on the characteristics of the deformation sequence, making it difficult to ensure reliable predictions using a single model. To address the limitations of current models, such as poor applicability, large fluctuations in prediction accuracy, and weak generalization ability, this paper proposes an adaptive prediction model for dam deformation based on the "decomposition–optimization–prediction–reconstruction" framework.MethodsThe dam deformation time series is first decomposed into several sub-mode sequences of different frequencies using the intrinsic computing expressive empirical mode decomposition with adaptive noise (ICEEMDAN) algorithm. Because high accuracy prediction is closely related to the extraction of key features from deformation data, sample entropy is introduced to quantify the complexity of each sub-mode sequence and to classify them according to entropy size. With prediction accuracy as the evaluation criterion, different hyperparameter optimization algorithms (e.g., dragonfly algorithm, whale optimization algorithm, and grey wolf optimizer) are combined with appropriate prediction models (e.g., long short-term memory network, gated recurrent unit network, and temporal convolutional neural network) for the classified sub-mode sequences. An adaptive criterion linking sample entropy, the optimization algorithm, and the prediction model is proposed. Then, the best combination of hyperparameter optimization algorithm and prediction model is determined. Finally, the deformation prediction results are obtained by reconstructing the predicted results of different sub-mode sequences, thereby forming an adaptive prediction model for dam deformation based on sample entropy.Results and DiscussionsEngineering applications demonstrate that, compared with traditional statistical regression models, the proposed model reduces the root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) by more than 60% and 90% in the training and testing sets, respectively, indicating high prediction accuracy and strong applicability. While the prediction of individual models and the proposed model is comparable on the training set, significant differences are observed in the testing set. For example, using the multi-correlation coefficient R as an indicator, the R values of all models are relatively high and similar in the training set, whereas significant differences occur in the testing set. The minimum increase in R achieved by the adaptive model is 23.5% compared with the TCN model and 35.1% compared with commonly used statistical regression models. These results indicate that the proposed model effectively extracts key features from deformation monitoring data and exhibits strong adaptability and generalization capability. With the increasing automation of dam monitoring systems, the growing number of monitoring points and increasingly complex data types place higher demands on prediction timeliness and accuracy.ConclusionsThe proposed prediction model not only improves prediction accuracy but supports real-time calculation, enabling integration into modern online dam safety monitoring systems. By effectively utilizing deformation prediction data to evaluate the safety status of dams, this model demonstrates substantial theoretical value and practical engineering significance.关键词：dam;deformation;sample entropy;optimization;adaptive prediction model251|242|0更新时间：2026-01-29
- “在高速铁路建设领域，专家通过试验研究揭示了红层泥岩填料强度特性的影响规律及改良机理，为红层泥岩弃渣资源化利用提供理论参考。”摘要：ObjectiveRed mudstone, a prevalent geological material in China, is frequently encountered during slope excavation and tunnel construction projects. These activities inevitably produce substantial amounts of red mudstone waste, creating significant challenges for disposal and environmental management. In addition, red mudstone waste exhibits high susceptibility to weathering, fragmentation, and water-induced disintegration and softening, which limits its engineering applications. Currently, it primarily serves as subgrade fill material. If red mudstone waste is improved for use as subgrade fill material, it reduces waste generation, expands application scenarios, and contributes to waste reduction, pollution control, and resource efficiency.MethodsThis study investigated the shear strength characteristics of red mudstone fill material and lime-stabilized fill material, and analyzed the influence of the degree of saturation on the strength properties of red mudstone fill. The effects of lime content and curing age on the strength characteristics and dilatancy behavior of stabilized fill were clarified. In addition, scanning electron microscopy (SEM), mercury intrusion porosimetry (MIP), and X-ray diffraction (XRD) tests were conducted to reveal the stabilization mechanisms of lime-stabilized red mudstone fill material at the microscopic level.Results and DiscussionsThe experimental results indicated that the cohesion of red mudstone fill material initially increased with the degree of saturation but then decreased, and reached its peak near the optimum water content. This strength variation demonstrated a typical "peak effect" in which cohesion followed a quadratic function relationship with degree of saturation, while the internal friction angle decreased based on a power function. The cohesion of red mudstone fill material reached its maximum near the optimum water content, whereas the internal friction angle significantly increased under drier conditions, which indicated that red mudstone fill material achieved higher shear strength in a relatively dry state. Compared to the fill material at its optimum water content, the cohesion of saturated red mudstone fill material significantly decreased to only 12.5% of its peak value, while the internal friction angle remained at 86% of its optimum state, which indicated that the substantial reduction in shear strength after saturation was primarily attributed to cohesion loss. Drainage and waterproofing measures should be emphasized in red mudstone subgrade construction to mitigate water-induced softening and excessive settlement. After lime stabilization, the cohesion of saturated red mudstone fill increased by more than 6.5 times, and the internal friction angle increased by over 1.4 times, which significantly enhanced the shear strength. In addition, the shear behavior transitioned from strain hardening and shear contraction in untreated samples to strain softening and shear dilation in lime-stabilized samples. The pore structure of lime-stabilized fill remained essentially unchanged after saturation compared to the unsaturated state, and retained a bimodal distribution. With increasing curing age, the proportion of micropores decreased, while nanopores became more prevalent. In addition, lime treatment maintained the particle integrity of red mudstone fill material and formed a three-dimensional interlocking framework that strengthened interparticle bonding, markedly improving shear strength compared to untreated red mudstone fill material.ConclusionsThe findings of this study provide theoretical guidance for the design, construction, and engineering application of improved red mudstone subgrades and further raise the resource utilization of red mudstone waste in high-speed railway construction.关键词：high speed railway;subgrade fill material:red mudstone;lime reinforcement;microstructure117|144|0更新时间：2026-01-29
- “在梯级泵站优化调度领域，专家建立了以碳排放量最小为目标的优化调度模型，探索了决策变量对目标函数的量化影响，并验证了基于改进龙格库塔算法（TRUN）的优化调度方法的优越性，为梯级泵站的优化运行提供解决方案。”摘要：ObjectiveLow operating efficiency, high energy consumption, and substantial carbon emissions are common problems in the operation of cascade pumping stations. To improve operational efficiency and support the "dual carbon" objective, an optimization scheduling model for cascade pumping stations is established with the goal of minimizing carbon emissions, and the Runge-Kutta algorithm (RUN) is introduced to solve the model. In addition, to address the lack of a quantitative sensitivity analysis in the optimal scheduling of cascade pumping stations, the Sobol global sensitivity method is employed to quantitatively evaluate the influence of key parameters on carbon emissions. To overcome the tendency of RUN falling into local optima due to insufficient initial population diversity and boundary stagnation, an improved Runge-Kutta (TRUN) algorithm based on a Tent chaotic map is proposed.MethodsFirst, an optimal scheduling model for cascade pumping stations was developed with carbon emission minimization as the objective function. Second, the Sobol method was used to analyze the sensitivity of the head of each pumping station and the flow rate of each unit to carbon emissions, thereby quantifying the impact of decision variables on the objective function. Third, an optimized scheduling method based on TRUN was proposed. While retaining the exploration characteristics of the RUN algorithm, Tent chaos mapping was introduced to enhance the diversity of the initial population, accelerate convergence, and improve solution accuracy. Additionally, a Tent boundary mapping strategy was adopted to regenerate boundary values, further improving optimization efficiency. Six benchmark functions, including unimodal, multimodal, and fixed-dimension functions, were used to verify the performance of TRUN and the effectiveness of the improvement strategies. Finally, a three-stage pumping station was selected as a case study, in which the Sobol method was used to determine the sensitivity ranking of system parameters, and TRUN was applied to obtain the optimal scheduling scheme.Results and DiscussionsThe mean values and standard deviations of six benchmark functions, including unimodal (Schwefel 2.21 (f1), Rosenbrock (f2)), multimodal (Schwefel (f3), Rastrigin (f4)), and fixed-dimension (Hartman (f5), Shekel (f6)), were calculated using the TRUN, RUN, TPSO, PSO, TGA, and GA algorithms. TRUN, RUN, TPSO, PSO, TGA, and GA achieved 4, 2, 0, 0, 1, and 0 optimal solutions, respectively, verifying the superiority of TRUN and the effectiveness of the proposed improvement strategies. Based on this, the Sobol global sensitivity analysis and TRUN-based optimization scheduling method were applied to a three-stage pumping station. The sensitivity ranking of system parameters, in descending order, was as follows: flow rate of each unit in the first-stage pumping station, flow rate of each unit in the second-stage pumping station, head of the first-stage pumping station, flow rate of each unit in the third-stage pumping station, head of the second-stage pumping station, and head of the third-stage pumping station. These results provide quantitative guidance for daily operational decision-making. In single-stage pumping station optimization, TRUN achieved 67, 56, and 46 optimal solutions out of 100 comparison runs, showing a clear advantage over the other algorithms. In cascade pumping station optimization, compared with the current operating scheme, the TRUN-based scheduling scheme reduced carbon emissions by 249 485 kg/a, outperforming RUN, TPSO, PSO, TGA, and GA, and confirming the effectiveness of the proposed algorithm.ConclusionsThe results demonstrate that the proposed TRUN algorithm exhibits excellent optimization performance. The TRUN-based optimal scheduling method for cascade pumping stations effectively improves system operational efficiency, and its optimization results are superior to those obtained using RUN, TPSO, PSO, TGA and GA. In addition, the Sobol global sensitivity analysis provides quantitative insights into the influence key parameters on carbon emissions, offering valuable references for operational decision-making of cascade pumping station systems.关键词：improved runge kutta optimizer;cascade pumping station;optimal scheduling;carbon emissions;sensitivity analysis279|521|0更新时间：2026-01-29
- “在地下工程领域，专家基于PFC软件构建了考虑地应力与随机节理网络耦合的爆破模型，揭示了节理密度、埋深、装药直径对爆破裂纹演化及岩体破碎行为的影响规律，为高地应力条件下节理岩体的爆破行为及其工程应用提供解决方案。”摘要：ObjectiveWith the ongoing depletion of shallow mineral resources, deep mining operations often exceeding depths of 1 000 m have become increasingly common. In such high-stress environments, the coupled effect of in situ stress and pre-existing joint networks significantly influences rock blasting performance. However, most traditional studies focus on homogeneous or regularly jointed rocks, which fail to represent the geometric randomness and complexity of natural joint distributions. This study aims to investigate the mechanisms of blast-induced damage and fragmentation in jointed rock masses under varying in situ stress conditions. Specifically, it examines how joint density, burial depth, and charge diameter interactively influence microcrack propagation, fractal complexity, and energy dissipation behavior during blasting, with the goal of providing guidance for efficient blasting in deep rock masses.MethodsA two-dimensional numerical blasting model was developed using the Particle Flow Code in 2D (PFC2D), with a discrete fracture network (DFN) incorporated to construct a synthetic rock mass (SRM). The joint planes were modeled using the Smooth Joint Model (SJM), and the intact rock matrix was represented by the Parallel Bond Model (PBM). The blasting load was applied through the particle expansion method, and the loading curve was defined by a half-sine waveform. The micro-mechanical parameters of the model were calibrated against laboratory test results of limestone samples, including uniaxial compression, triaxial compression, and Brazilian tensile tests to ensure the physical accuracy of the simulation. The simulations covered five joint densities (P10=2～10 m-1), five burial depths (300～1 500 m), and five charge diameters (0.010～0.018 m). The output indicators included microcrack count, peak stress, energy distribution (kinetic, strain, and frictional energy), fractal dimension of crack networks, and block size distribution. In addition, a custom FISH script was developed to identify fragmented blocks and quantify post-blast fragment size distribution.Results and discussionsThe results showed that joint density had a non-monotonic influence on fragmentation behavior. At low joint densities, microcrack formation was enhanced due to localized stress concentrations near joint tips. However, as density increased, joint interfaces acted as preferential energy dissipation paths, which reduced the formation of new cracks and decreased the fractal dimension of the crack network. Peak stress propagation showed an exponential attenuation pattern, with up to a 25.97% reduction at high joint densities due to enhanced wave scattering and reflection. Increasing in-situ stress led to a marked reduction in microcrack count and fractal dimension, which indicated more localized and constrained damage zones. Under high stress conditions, microcracks tended to align along the direction of maximum principal stress, particularly in jointed media. A stronger inhibitory effect of joint density on microcrack development was observed at lower stress levels, which highlighted the interactive coupling between these two parameters. Energy analysis revealed that strain energy storage decreased with increasing joint density, while frictional dissipation became more prominent. Kinetic energy remained relatively insensitive to joint configuration. Block size distribution analysis indicated a shift toward finer fragmentation with increased joint density, which demonstrated higher energy efficiency in rock breakage under joint-controlled failure modes. Regarding charge diameter, increasing the diameter enlarged the fragmented zone and increased the number of microcracks. However, beyond a critical diameter (approximately 0.016 m), both the microcrack count and fractal complexity plateaued, which indicated saturation in the fragmentation mechanism. At this stage, a further increase in charge size contributed to over-crushing and inefficient energy utilization.ConclusionsThis study establishes a PFC2D-based numerical model that couples in-situ stress with randomly distributed joints to simulate the dynamic blasting behavior of jointed rock masses. The simulation results indicate that increasing joint density reduces crack complexity and encourages localized fracture patterns, whereas in-situ stress further restricts crack propagation and raises directional growth along the principal stress direction. The interaction between joint density and stress is nonlinear, with more pronounced suppression of fracture development at lower stress levels. Although larger charge diameters expand the fragmentation zone, the growth in crack complexity reaches saturation beyond a threshold, indicating limited improvement in breakage efficiency. Overall, the model provides theoretical support and design guidance for efficient blasting in deep, jointed rock masses.关键词：Jointed rock masses;high in situ stress;Single-hole blasting model;fractal dimension;particle flow80|74|0更新时间：2026-01-29
- “在岩体工程稳定性领域，专家建立了砂岩蠕变本构模型，验证了损伤统计模型准确性，为工程应用提供参考。”摘要：ObjectiveThe creep of rock mass is very important for long-term project stability, and the constitutive model best reflects the mechanical characteristics and deformation patterns of the rock creep process. In recent years, extensive studies on creep characteristics of rocks have been conducted, and many creep constitutive models have been derived from classical elastoplastic mechanics. These models reveal the creep behavior of specific rocks, but they often do not sufficiently take the influence of creep parameters into account, or they overlook the certain factorsduring researchs, so establishing a model that comprehensively incorporates parameters and accurately reveals the creep mechanical behavior of rocks is of significant value.MethodsThe Nishihara model in the classical elastoplastic framework was selected, and damage was introduced to compensate for its inability to reveal the nonlinear creep characteristics of rocks. The Weibull distribution function was applied as the damage statistical function, and the SMP criterion was adopted as the strength criterion of sandstone to derive the damage variables under three-dimensional conditions. A conventional triaxial compression test was conducted using the MTS-815 rock rigidity testing machine, and the applicability of the damage statistical model was verified by comparing the compression test results with the damage evolution curve. The yield function was applied to fit the yield surface of sandstone to determine the yield stress under each confining pressure, and a fracture plastic body was incorporated to represent the plastic deformation of the primary fracture. The components of each part of the model were then integrated to derive the constitutive equation in accordance with the mechanical boundary. The STAC600-600 rock rheology testing machine was utilized to perform the rock classification rheological test, and the fitting results of the creep model were analyzed. The model parameters were optimized to improve the fitting accuracy. Finally, the fitting results of this model were compared to those of other models to analyze the advantages and disadvantages of the approach.Results and DiscussionsThe damage variables derived from the SMP criterion and Weibull distribution function showed strong consistency with the rock deformation behavior under conventional triaxial compression. Damage did not develop during the early loading stage, and when the deviator stress exceeded the yield stress, damage developed rapidly until destruction ofwhole rock was happened. The damage evolution trend indicated that the fracture unit of the rock accumulated rapidly once the deviator stress reached the yield limit, and the rate of damage accumulation varied under different confining pressures, with lower confining pressure leading to a faster accumulation rate. The improved Nishihara model was applied to fit the creep deformation curve and produced satisfactory results; however, the fitting results of the accelerated creep curve under 8, 12, and 20 MPa confining pressures all shifted forward. The Poisson ratio μ was refitted as a function related to confining pressure, and the accuracy improved significantly. An algebraic relationship existed between the key model parameters u and m and the confining pressure, and the constitutive parameters under each confining pressure were obtained after establishing this relationship. Compared to the fitting results of the other two models, the fitting accuracy of this model was higher, and the variable “time”, which often caused errors, did not participate in the parameter calculation. The damage parameters were directly related to stress and strain, which improved accuracy to a certain extent.ConclusionsThe results indicate that the damage statistical creep constitutive model accurately fits the creep deformation characteristics of rocks under three-dimensional conditions, expands the range of influencing factors for constitutive parameters, and provides practical guidance for related engineering applications.关键词：rock creep;SMP criterion;damage variables;weibull distribution;Nishihara model133|145|0更新时间：2026-01-29
- “最新研究开发出一种低碳、经济、高效的固废基胶凝材料，能有效固化盾构渣土，提高其作为路基填料的性能，为盾构渣土资源化利用提供新途径。”摘要：ObjectiveBackfilling of subgrade soil is an important approach for the resource utilization of shield tunnel muck, but the muck produced by shield construction, which exhibits high water content and high compressibility, is challenging to meet subgrade filling requirements. At present, commonly used soil stabilizers such as cement and lime present issues that include high energy consumption, high carbon emissions, insufficient strength, and poor stability. Therefore, it is urgent to develop low-carbon, economical, and efficient curing modified materials to achieve the resource utilization of shield tunnel muck. In this study, the whole solid waste curing agent (CRGD) prepared from industrial wastes such as carbide slag, red mud, ground granulated blast furnace slag, and desulfurized gypsum is utilized to solidify shield tunnel muck, and the performance and improvement mechanism of the improved shield tunnel muck as roadbed filler are examined.MethodsThe test soil was the shield tunnel muck produced during the excavation of the silty clay stratum in a section of Jinan Rail Transit Line 4. A solid waste-based cementitious material (CRGD) was developed to solidify the shield tunnel muck by using the common solid waste red mud, desulfurized gypsum, granulated blast furnace slag, and carbide slag in Shandong Province as raw materials. It consisted of 38% carbide slag (CS), 25% red mud (RM), 25% granulated blast furnace slag powder (GGBS), and 12% desulfurized gypsum (DG) in mass ratio. Lime, a material widely used in solidified soil, was used as a comparative experimental group. First, the compaction curve was drawn through the compaction test to determine the maximum dry density and the optimum moisture content. Then, the specimens were prepared based on the optimum moisture content for the determination of the CBR value, unconfined compressive strength (UCS), and water stability. The UCS of the specimens was measured based on the specification after curing the specimens for 7 d, 14 d, and 28 d, and the water stability coefficient was calculated. Finally, low-field nuclear magnetic resonance (LF-NMR), X-ray diffraction (XRD), and scanning electron microscopy (SEM) were utilized to analyze the hydration products and microstructure of the solidified muck and to reveal its curing mechanism.Results and DiscussionsThe results of the compaction test showed that with the increase in dosage, the compaction curve shifted toward higher water content and lower dry density. After adding 4% CRGD, the wopt of the shield tunnel muck increased from 15.9% to 17.9%, and the ρdmax decreased from 1.81 to 1.60 g/cm3. Under the same dosage conditions, the wopt of lime-solidified muck was higher than that of CRGD-solidified muck, but its ρdmax was lower than that of CRGD-solidified muck. The unconfined compressive strength of the uncured shield tunnel muck was only 0.718 MPa, which did not meet the requirement of 1.5 MPa in the specification. After curing, the UCS increased significantly and increased with higher dosage and longer curing age. The strength increase of CRGD-solidified muck was more significant than that of lime-solidified muck. At 28 days of curing, the UCS of the solidified muck with 4% to 8% CRGD content increased by 230.5%, 334.8%, 389.8%, 477.6%, 525.1%, respectively. At 7% dosage, the UCS of CRGD-solidified muck cured for 28 days reached 4.15 MPa. At the same time, the water stability of the unsolidified shield muck was extremely poor, and the specimen did not maintain its integrity after soaking in water, disintegrating quickly in a short time. The water stability was significantly enhanced after CRGD curing, and the specimen maintained good integrity after immersion for 1 day. The water stability coefficient of the shield tunnel muck after curing treatment was consistent with the change pattern of the UCS and increased with higher dosage and longer curing age. At 7% dosage, the water stability coefficient of CRGD-solidified muck after 28 days of curing reached 83%, which was 20% higher than that of lime-solidified muck. In addition, the bearing capacity of CRGD-solidified muck was also higher than that of lime-solidified muck. At 4% dosage, the CBR value increased from 3.16% to 58.14%, which was 17.40 times higher than that of unsolidified muck and far higher than the design specification requirements. Under the same dosage, the CBR value of CRGD-solidified muck was higher than that of lime-solidified muck , and the difference between the two increased with dosage. When the content reached 8%, the CBR value of CRGD-solidified muck was 217.98%, while that of lime-solidified muck was 111.55%. The porosity of the solidified muck after immersion decreased with increasing content, and the porosity of the CRGD-solidified muck was significantly lower than that of the lime- solidifiedmuck . With higher CRGD content, the pores in the range of 0.001~0.1 μm were relatively reduced, while the pores in the range of 0.1~1.0 μm were relatively increased. After curing, the diffraction peak intensity of minerals such as quartz and mica in the sample weakened, and some mineral peaks almost disappeared, which indicated that the curing agent had a pozzolanic reaction with the silicon-aluminum active minerals in the muck . Compared to lime-solidified muck, AFt also appeared in CRGD-solidified muck, which indicated that the calcium-aluminum active substances in the system reacted with DG, and this reaction was an important reason why the strength of CRGD-solidified muck was higher than that of lime-solidified muck. SEM images showed that shield tunnel muck contained many holes and cracks. After CRGD curing, more amorphous hydration gels and needle-like products were observed. Combined with EDS analysis, these products were identified as C‒(A)‒S‒H gel and AFt. The needle-like ettringite interacted with the muck matrix to form a skeleton structure, which effectively filled the muck pores and improved the compactness and strength of the solidified muck. Based on the current market price, the cost of lime required to treat 1 ton of shield tunnel muck with 4% dosage was 18.000 yuan, while the cost of CRGD required was only 5.160 yuan, saving 12.840 yuan per ton of shield tunnel muck.ConclusionsThe results show that the UCS, water stability coefficient, and CBR values of the modified shield tunnel muck used as roadbed filler are significantly higher than those of lime-solidified muck. This improvement is achieved by using carbide slag, red mud, granulated blast furnace slag, and desulfurization gypsum to prepare the solid waste-based cementitious material (CRGD) for shield tunnel muck solidification. The components of CRGD interact with each other and react with the active silicon-aluminum component in the muck through a pozzolanic reaction, mainly generating C‒(A)‒S‒H gel and needle-like ettringite . These products effectively fill the muck pores, reduce porosity, and improve the compactness and strength of the solidified muck. In addition, treating 1 ton of shield tunnel muck requires only 5.160 yuan, indicating favorable economic and environmental benefits.关键词：shield tunnel muck;subgrade;solid waste-based cementitious materials;CBR value;water stability231|248|0更新时间：2026-01-29
HYDRAULIC ROCK MECHANICS AND ENGINEERING
- “在地质灾害领域，专家通过灰岩相似材料法向碰撞试验，揭示了落石碰撞碎裂过程中能量转化规律，为治理方案设计提供支撑。”摘要：ObjectivePrior to incorporating fragmentation into predictive trajectory models, further research is required to gain a comprehensive understanding of its impact. This study explores the complexities of energy dissipation and fragmentation through experiments using rock spheres of varying diameters. The key outcomes are examined, including impact force, energy dissipation characteristics, and fragmentation conversion mechanisms. To quantify the degree of fragmentation, a novel index based on elastic strain energy is introduced, revealing the intricate relationship between energy dissipation and fragmentation.MethodsThis study employs the Zuojiaying rockfall in Guizhou as a geological prototype. Rockfall similarity materials were selected to prepare high-brittleness, low-strength rock samples with mechanical properties analogous to those of limestone. Uniaxial compression tests and Chevron Notched Brazilian Disk (CCNBD) tests were conducted under a range of strain rate conditions to enable a systematic analysis of the effects of strain rate on the material's mechanical properties, energy dissipation, and fracture toughness. To further investigate the dynamic characteristics of normal rockfall impacts, a custom-designed normal impact testing device was developed, featuring an automatic release system for precise control. Particle Image Velocimetry (PIV) and ImagePy image processing technology were employed to analyze pivotal parameters, including impact velocity and fragment size distribution. Furthermore, the effects of rockfall size and impact velocity on impact force, coefficient of restitution, and fragmentation degree were examined. Finally, the study focused on energy conversion and fragmentation characteristics under different impact conditions, proposed a fragmentation quantification index based on elastic strain energy, and compared the energy conversion and fragmentation differences between normal and inclined slab impacts.Results and DiscussionsBased on the geological prototype of the Zuojiaying rockfall, rock-like materials were prepared. As the strain rate increased, the characteristic stress exhibited an overall increasing trend, while fracture toughness initially increased and then decreased. Elastic strain energy was continuously accumulated within the rock before the peak and was rapidly released after the peak. Due to the initiation and propagation of cracks, dissipated energy continued to grow. However, before the peak, its value was lower than that of elastic strain energy, while after the peak, the growth rate of dissipated energy increased, surpassing that of elastic strain energy. The results of the normal impact tests conducted on the rock-like spheres revealed four dynamic fragmentation characteristics and four stages of dynamic fragmentation. Following fragmentation, the larger fragments underwent further compression, resulting in additional fragmentation, with a significantly greater number of smaller fragments than larger fragments. Due to the action of high hoop tensile strain rates, the smaller fragments detached from the larger fragments at high speeds. As the falling height (or impact velocity) and specimen size increased, the impact force also increased. Additionally, an increase in the elastic modulus of the impacted slab led to an increase in the impact force. The energy conversion analysis of the dynamic fragmentation process demonstrates that the pre-impact kinetic energy, converted from the initial potential energy, dissipates in the form of slab deformation energy, elastic wave dissipation energy, elastoplastic deformation energy, fracture energy, and post-impact kinetic energy of the fragments. Among these, elastoplastic deformation energy, post-impact kinetic energy, and fracture energy account for more than 80% of the total energy dissipation. As the pre-impact kinetic energy increases, elastic wave dissipation energy, slab deformation energy, post-impact kinetic energy, elastoplastic deformation energy, and fracture energy all increase to varying degrees. Meanwhile, the energy conversion rate is negatively correlated with the initial total energy. Based on this analysis, a dynamic fragmentation degree index B, based on the degree of elastic strain energy release, is proposed. As impact velocity and impact force increase, the fragmentation degree curves of B and B3 (area ratio) show consistency, indicating that this index effectively describes the degree of rockfall fragmentation and accurately reflects the energy conversion during the fragmentation process. Additionally, a comparison of normal and inclined slab impact test results reveals that fragmentation degree and energy conversion rate are primarily influenced by the incident angle and impact velocity, with fragmentation degree negatively correlated with the coefficient of restitution.ConclusionsThe fragmentation process during impact exhibits distinct stages and varying degrees of severity, with energy dissipation primarily consisting of elastoplastic deformation energy and fracture energy, which together account for as much as 91.94% of the total. The kinetic energy retained by the remaining fragmented blocks comprises only 8.05% of the total, and the energy conversion rate is inversely proportional to the initial total energy. To quantify the degree of fragmentation, a metric based on elastic strain energy has been defined and validated for its feasibility in analyzing the fragmentation process of specimens during impact. A further comparison with inclined slab impact tests reveals that the impact angle significantly influences the degree of specimen fragmentation and controls the efficiency of energy conversion.关键词：Rockfall;Normal impact;Energy conversion;Degree of fragmentation109|274|0更新时间：2026-01-29
-
Experimental Study on the Mechanism of Progressive Collapse Resistance of Base-isolated Structure AI导读
“在建筑抗震领域，专家通过非线性静力推覆试验，研究了基础隔震结构抗连续倒塌性能及其参数影响规律，为提高建筑抗震性能提供新思路。”摘要：ObjectiveThe fundamental difference between seismic isolation structures and ordinary seismic structures lies in the incorporation of an isolation layer, which effectively mitigates earthquake input energy. The variation in structural measures within the isolation layer leads to different levels of resistance to progressive collapse. As a result, existing research on the progressive collapse resistance of conventional seismic structures does not fully apply to seismic isolation structures. In addition, there is a clear need to develop a more comprehensive understanding of how the stiffness of the isolation layer and the floor tie effect influence the collapse performance and load transfer mechanisms of seismic isolation structures, which requires further empirical validation.MethodsThis study aims to examine the performance of base isolation structures against continuous collapse and to assess the impact of various parameters by utilizing an office building located within a hospital in Ningxia as the reference model. It encompasses the design of substructure tests, working condition design, and test parameter design. The test model underwent comprehensive adjustments while considering the unit test conditions and several limiting factors. Nonlinear static pushover tests were performed on three 1/3-scale reinforced concrete foundation isolation substructures using the dismantled member method. For clarity, the substructural member with a standard rubber isolation bearing was designated as S1, the member with a lead-core rubber isolation bearing as S2, and the substructural member with a flange plate as S3. These test specimens were subjected to static loading under displacement control. Initially, the loading was controlled at 5 mm per stage, and upon reaching the maximum beam mechanism load, it transitioned to controlled loading at 10 mm per stage. The force states, damage modes, and resistance mechanisms of the three specimens were systematically documented. Comparative analyses of bearing deformations, strain variations, deflection curves, and load-carrying capacities were conducted. Ultimately, by employing the principle of energy balance, the dynamic response and collapse resistance of the three specimens were evaluated.Results and DiscussionsThe findings revealed that damage predominantly occurred in regions of higher bending moments, particularly near the center column, while no cracking was observed at the side column nodes. The specimens in the base isolation structure exhibited weak lateral confinement. Upon attaining the peak beam mechanism value, the increase in bearing capacity decelerated due to the energy dissipation of the bearing, which hindered significant capacity growth during the second peak. Therefore, categorizing the resistance of base isolation structures into beam and composite mechanisms was proved to be challenging. The analysis primarily identified two stages, namely the beam mechanism and the composite mechanism. Specimens S1 and S3 exhibited reinforcement rupture at the bottom of beam A, whereas the failure of specimen S2 occurred in the plastic hinge area at the beam end on the right side of the center column, which was attributed to the high horizontal stiffness of the seismic isolation bearing. In seismic isolation structures, deformation primarily occurs in the bearing, which can delay the failure of the plastic hinge at the beam ends of the frame following the formation of a plastic hinge at the beam end of the center column. The peak value of the beam mechanism for specimen S2 was 1.14 times greater than that of specimen S1, and its ultimate bearing capacity surpassed that of S3 by 1.11 times. This indicated that the ultimate load of the base seismic isolation structure increased with the stiffness of the bearings. Similarly, the peak value of the beam mechanism for specimen S3 was 1.09 times higher than that of S1, with an ultimate bearing capacity 1.13 times greater than that of S1. The beams and slabs synergistically resisted the upper load, and the tie effect of the floor slab significantly enhanced the structural resistance to continuous collapse. The tensile action of the floor slab improved the anti-continuous collapse capacity of the base isolation structure. The angular and horizontal displacements in the base isolation structure enabled the substructures, with plastic hinges at the beam ends of the center columns, to exhibit rotational deformation capacity. This delayed the failure of the plastic hinges at the beam ends of the frame and enhanced the anti-continuous collapse performance of the structure. The maximum dynamic loads for S1, S2, and S3 were recorded as 118.8, 140.1, and 129.4 kN, respectively, which were significantly lower than the corresponding static loads. The duration of the dynamic load action for S1 was 1.13 times longer. Despite the short duration of the dynamic load impact, the structure incurred damage when the load reached its peak. Based on the test outcomes and a simplified evaluation method for structural dynamic response grounded in energy balance principles, the residual load capacity of S2 was found to be 28.9% less than that of S1. The residual load capacity of S3 was 81.6% less than that of S1 and 74.1% less than that of S2. These findings indicated that, in the event of medium bearing failure, the rubber isolation bearing in the base isolation structure enhances the residual capacity of the structure to resist continuous collapse. However, increasing the horizontal stiffness of the seismic isolation layer has a minimal effect on the ability of the base isolation structure to resist continuous collapse.ConclusionsUnlike traditional seismic-resistant structures, the progressive collapse resistance mechanism of base-isolated structures was divided into two stages: the beam mechanism and the composite mechanism. The horizontal stiffness provided a long peak load plateau, leading to a significant increase in both rotational and horizontal displacements. The failure mode of beams was characterized by the prior yielding of steel bars near the middle joints; tensile steel bars at both ends also yielded during the beam mechanism stage, while compressive steel bars remained unyielded. Additionally, the slab tension effect effectively enhanced the bearing capacity of isolated structures under the beam mechanism. Energy was dissipated by isolation bearings, resulting in substantial rotation of beams, yet the concrete at the beam ends of edge columns remained intact. The weak constraint of the isolation layer delayed the failure of plastic hinges at the ends of frame beams. Based on the energy equivalence principle and the DoD progressive collapse criterion, an evaluation of the residual progressive collapse resistance of beam-column substructures in base-isolated structures was conducted. It was found that the increase in the horizontal stiffness of the isolation layer had little influence on the residual progressive collapse resistance of base-isolated structures.关键词：Base-isolated structure;progressive collapse;Collapse resistance;Pushdown analysis method;failure mode144|269|0更新时间：2026-01-29 - “攀枝花阳光隧道工程研究揭示了隧道开挖对围岩和结构的影响，为隧道安全施工提供科学依据。”摘要：ObjectiveThe current research on excavation disturbance in mountain tunnels is mostly confined to primarily focused on surrounding rock deformation and stability, tunnel structural load-bearing capacity and design, as well as aspects such as asymmetric loading, excavation impacts and control. However, there is no precedent or well-documented case study for the excavation of tunnels in sensitive construction areas that involves dismantling existing high slope retaining structures, and the key issues and challenges involved are rarely reported. Therefore, it is necessary to propose a reasonable and feasible transfer system scheme specifically for the excavation of breaking anti-slide pile in mountain tunnels. This would involve clarifying the mechanical response characteristics of surrounding rocks, tunnels and sensitive structures under excavation disturbance, and elucidating the plastic deformation patterns and potential failure modes of the surrounding rocks.MethodsDepending on the Panzhihua Sunshine tunnel project and the principle of similarity ratio, a 1:40 scale model test device was constructed to investigate the mechanical response characteristics of the surrounding rock and its structure under the disturbance of tunnel excavation by carrying out the model test of tunnel breaking anti-slide pile. Based on the model test, a 3D fine model considering the interaction of tunnel, slope, conversion system, sensitive building, and surrounding rock was developed using numerical methods and compared with the test results to verify the accuracy and reliability of the two research methods. Therefore, the force characteristics of the transfer structure, and the influence characteristics of the existing sensitive buildings could be further analyzed, and the plastic deformation regulation, as well as its potential damage mechanism of the surrounding rock, could be clarified. The model test and numerical calculation are in good agreement with the displacement law and value at the key points. The numerical calculation results are generally slightly larger than the test monitoring data, and the maximum difference is 10% indicating that the methods used are reasonable.Results and DiscussionsThe results show that under the influence of anti-slide piles breaking during tunnel excavation, the surrounding rock mainly underwent settlement with lateral displacement, which significantly increased before the arrival of the excavation face and gradually stabilized after passing through. Moreover, the maximum displacement of the slope caused by the tunnel excavation is 1.19 mm, and the influence on the surrounding rock from tunnel construction is well controlled. Stress disturbance redistribution is induced by tunnel excavation, and the tunnel-surrounding rock contact pressure and strain increase caused by local surrounding rock self-weight adjustment, with the maximum pressure in the shallow overburden area and the maximum strain increased by 28.3% and 58.28×10-6, respectively. In addition to the transient deformation, tunnel excavation is obviously affected by the time effect, which shows the lagging of soft rock deformation. Tunnel excavation disturbance diffusely and attenuatively transferred to the conversion structure and caused its flexure deformation, with the maximum strain being 30% of the maximum value of the tunnel structure. The flexure deformation is inversely correlated with the distance from the excavation face, and a sudden increase in strain occurs during the anti-slide piles breaking process. With transient deformation characteristics, the impact of the existing slope is mainly in the range of the neighboring piles of the breaking anti-slide piles. The breaking of anti-slide piles is merely impacted by the tunnel excavation, with the existing slope stability factor being higher than 1.35 and in an overall stable state. Combining with the structural responses of the surrounding rock and tunnel, it was concluded that the conversion structure could effectively inhibit the mechanical turbulence and ensure the tunneling safety. The neighboring existing buildings mainly experienced the sections of a small deformation, significantly-increased deformation, slo-wly-increased deformation, and deformation stabilization during the tunnel excavation process. The weak area of the space staggered tunnel is in the hance position of the upper tunnel above the connecting line, and preventing the plastic damage on the hance position of the oblique upper tunnel should be focused on during the construction.ConclusionsThe research results provide a scientific basis for the scheme design and field implementation of the background project, and have achieved perfect construction results. It has become a regional landmark building that provides a good reference for similar projects.关键词：mountain tunnel;breaking anti-slide piles;transfer system;excavation disturbance;force deformation;model test;numerical sim-ulation124|327|0更新时间：2026-01-29
- “在深埋高应力隧道领域，专家探讨了TBM岩爆地层防控方法，为提高TBM施工效率和安全性提供解决方案。”摘要：ObjectiveIn order to study the rock burst prevention and control method of open TBM in deep-buried high-stress tunnels and solve the problem of high risk and low efficiency of open TBM construction in rock burst strata, the coping techniques of open TBM in different strata of rock burst were analyzed. Based on the numerical analysis of the TBM shield against rock burst, the simplified basis of the model and the principle of determining the calculation boundary were proposed, which provided a reference for the simulation calculation of the TBM shield against rock burst.MethodsThe commonly used 10 m level open TBM was selected as the research object, and the dynamic simulation software was used to analyze the structural stress and structural deformation characteristics of the top shield under three levels of medium rock burst, strong rock burst, and extremely strong rock burst. In terms of model simplification, the rock block parameters under three rock burst levels were determined according to the existing research results, including rock block shape, rock block size, rock block ejection velocity, and rock block number. For conservative considerations, the simulation calculation assumed that a rock burst occurred above the entire top shield area, and the velocity direction of the rock blocks generated was along the radial direction of the shield. For the sake of simplifying the calculation, only one layer of rock blocks was calculated. In terms of the calculation boundary, the influence of the side shield and the advanced detection device on the top shield was ignored. Assuming that the jacking cylinder was locked, the connecting hole between the jacking cylinder and the top shield was constrained, and the changes of rock block characteristics were not considered, a rigid constraint was imposed on the rock. In addition, relying on a railway tunnel project in Southwest China, this paper summarizes and analyzes the prevention and control methods for different levels of rock burst in the construction process of dual-structure TBM, and puts forward the coping technology of TBM in different levels of rock burst strata.Results and DiscussionsAccording to the numerical simulation results, the structural stress and deformation characteristics of the shield under different levels of rock burst were analyzed. The conclusions were as follows: 1) the maximum impact force borne by the shield under the three levels of rock burst was 4 998 kN for medium rock burst, 16 564 kN for strong rock burst, and 70 106 kN for extremely strong rock burst; 2) The maximum impact force borne by a single jacking cylinder was 5 618 kN for medium rock burst, 11 293 kN for strong rock burst, and 18 873 kN for extremely strong rock burst; 3) The maximum deformation of the top shield was 15.2 mm for medium rock burst, 18.6 mm for strong rock burst, and 77.3 mm for extremely strong rock burst. The maximum deformation was located on both sides of the tail of the top shield, which was similar to the deformation of the top shield under actual site rock burst; 4) The maximum effective plastic strain of the shield structure was 0.03 for medium rock burst, 0.13 for strong rock burst, and 0.36 for extremely strong rock burst, indicating that large-area plastic and local cracking occurred in the shield structure under the action of extremely strong rock burst, which might affect the storage space of steel bars. Through the summary and analysis of the prevention and control method for different levels of rock burst in the construction process of dual-structure TBM, the coping strategies for different levels of rock burst were put forward: 1) passive prevention and control concept was the main idea for light rock burst and medium rock burst; 2) In strong rock burst strata, the control strategy of passive prevention and control concept plus active prevention and control concept was adopted. The surrounding rock was sealed by shotcreting in time to reduce the expansion of rock burst, and the L1 area at the shield tail was protected by impact-resistant anchor rods and the McNally support system. The "steel pipe sheet plus auxiliary propulsion" method could also be used to pass through strong rock burst strata; 3) Active prevention and control concept should be given priority in extremely strong rock burst formation. If the existing TBM was unable to carry out efficient advanced stress relief and the pilot tunnel method was not the optimal solution, it was recommended to use the drilling and blasting method to pilot the tunnel through the extremely strong rock burst area. For the prevention and control of TBM under the condition of extremely strong rock burst, it was proposed to develop a TBM with the ability of tunnel face advanced processing, so as to improve the adaptability of TBM to complex rock burst strata.ConclusionsThrough the anti-rock burst simulation analysis of the top shield, the impact force of different levels of rock burst on the shield and jacking cylinder was obtained. The reliability of the numerical simulation method was verified by comparing the simulation deformation results with the actual shield deformation on site. According to the calculation results, the rock burst resistance of the TBM host area can be improved by strengthening the strength of the shield shell and main bearing structure, improving the bending section coefficient of the shield tail structure, and reducing the cantilever length of the shield tail. In the light and medium rock burst strata, the control strategy of passive protection can be adopted. In the strong rock burst strata, the control strategy of passive prevention and control concept, plus active prevention and control concept, should be adopted. In the extremely strong rock burst strata, the concept of active prevention and control should be considered, and it is urgent to develop a TBM with the advanced processing ability of the tunnel face. This paper describes the prevention and control methods of different levels of rock burst, and puts forward the calculation method of the top shield against rock burst, which can provide some reference for the formulation of disposal measures when the TBM passes through rock burst strata.关键词：rock burst;drilling and blasting method;TBM method;prevention and control concepts;prevention and control techniques;top shield;numerical simulation125|175|0更新时间：2026-01-29
- “在隧道支护设计领域，专家从三维时空效应出发，通过数值模拟揭示黏弹性围岩隧道变形特性，为流变软岩隧道支护设计提供理论基础。”摘要：ObjectiveThe design of support structures for tunnels in viscoelastic rocks requires consideration of the complex, superimposed spatial influence of the tunnel face and the rheological properties of the surrounding rock. This study investigates the spatiotemporal evolution of the longitudinal displacement of tunnels in viscoelastic surrounding rocks through extensive numerical simulations and establishes a regression model for the displacement release coefficient using the response surface method. This research provides a theoretical foundation for the design of tunnel support systems in rheological soft rock.MethodsIt examines the space-time evolution behavior of the fictitious support pressure in viscoelastic surrounding rock and derives an analytical expression for this pressure. A numerical simulation model of the longitudinal deformation curve of viscoelastic surrounding rock is developed, revealing the longitudinal displacement profile (LDP) characteristics of the viscoelastic constitutive model. An empirical formula for the displacement release coefficient is derived based on the spatiotemporal dynamic evolution mechanism.Results and DiscussionsThe study analyzes the effects of varying surrounding rock parameters on LDP, noting that the LDP under viscoelastic conditions demonstrates similarities to, but apparent differences from, those under elastic and elastoplastic conditions. Specifically, the undisturbed region of surrounding rock beyond a certain distance from the unexcavated face shows a displacement release rate of zero. Near the excavation face, the displacement release rate of the surrounding rock gradually increases and then rises sharply. This rate is affected by both temporal and spatial factors, which dynamically evolve depending on the interaction between the excavation speed and the time-dependent characteristics of the surrounding rock. Beyond a certain distance after tunnel face excavation, the displacement release rate is controlled solely by the time effect, making the spatial effect negligible. Viscoelastic surrounding rocks exhibit a superimposition of time effects on the spatial effect range, defined as space-time effects. The region where the spatial effect disappears is influenced only by the time effect, corresponding to the traditional rheological stage. The displacement release rate of viscoelastic surrounding rock within the space-time effect zone correlates with the excavation speed. An empirical formula applicable to viscoelastic surrounding rock for determining the displacement release coefficient is derived utilizing the dynamic interaction between time and space effects. This coefficient, when reduced to a purely spatial effect, aligns with the existing elastic displacement release coefficient and extends its applicability to unexcavated surrounding rock. This study also investigates the space-time evolution behavior of the fictitious support force in viscoelastic surrounding rock and develops an empirical formula for this force. The fictitious support force for the viscoelastic surrounding rock is determined based on the elastic surrounding rock solution using the Laplace transformation. The longitudinal deformation curve of the surrounding rock, as established in this study, facilitates the calculation of the fictitious support force.ConclusionsAnalysis of various parameters indicates that the variation in the fictitious support force is related to the tunnel excavation speed. When the excavation speed is sufficiently slow, the viscoelastic surrounding rock behaves similarly to elastic rock, and the influence of the time effect on the fictitious support force becomes minimal, being primarily governed by the spatial effect. In contrast, at higher excavation speeds where the time effect becomes significant, the fictitious support force and the displacement release coefficient for viscoelastic surrounding rock cannot be derived directly through inversion; they must be coupled with the surrounding rock characteristic curve, thus being influenced by both time and spatial effects. The empirical formula for the fictitious support force, considering space-time effects, demonstrates that for different surrounding rock parameters, the fictitious support force in viscoelastic surrounding rock is more sensitive to excavation speed.关键词：tunnel engineering;viscoelastic rock;Surrounding rock deformation;Longitudinal displacement profile;Fictitious support pressure;Space-time effect197|591|0更新时间：2026-01-29
- “在桥梁工程领域，专家提出了悬索桥主缆张力监测新方法，为桥梁安全评估提供可靠依据。”摘要：ObjectiveLong-term, effective, and accurate measurement of the main cable force in suspension bridges is essential for the timely detection of structural anomalies and the evaluation of structural health. At present, long-term monitoring of the main cable force faces two significant challenges: low testing accuracy and high testing cost. This study presents a long-term monitoring approach for the main cable force in suspension bridges based on splay saddle displacement.MethodsFirst, the equilibrium states of two different boundary forms, the sliding type and swing type, were discussed separately, and the calculation method for the resultant force of the anchor cable of the suspension bridge was derived. The long-span suspension bridge was regarded as a series and parallel spring system composed of various structural parts, and its structural members were simplified as springs to form an equivalent stiffness model of the series and parallel system. The cable stiffness in this model was considered a series combination of elastic tensile stiffness and linear stiffness. Based on different load forms of the cable, the longitudinal stiffness of the parabolic and catenary cables was calculated separately. The constraint stiffness of the bridge tower on the main cable was derived using the energy method and considering the influence of the second-order effect. Then, based on the balance principle at the splay saddle and the equivalent simplified stiffness model of the suspension bridge, a method for calculating the tension of the main cable using the splay saddle displacement was proposed. Taking a suspension bridge with a 660 m main span as an example, the accuracy and effectiveness of the proposed method were verified through finite element modeling analysis. A low-cost main cable force monitoring scheme was designed using a simple custom bracket and a splay saddle displacement sensor. The bridge's monitoring data for 181 days were preprocessed, and the correlation model between the splay saddle displacement and temperature was established through a polynomial fitting method. The temperature-normalized splay saddle displacement value was then obtained. The average daily displacement of the splay saddle was taken as the representative value, and both the histogram and Q‒Q plot were drawn to analyze the statistical characteristics of the data. Finally, the ARIMA model was established by selecting the first 121 cable force monitoring data points to predict the trend of data variation during the last 60 days.Results and discussionsBased on the established equivalent stiffness model of the suspension bridge, the translational stiffness of the IP point of the cable saddle along its supporting surface was 1 281.7 and 1 814.9 kN/mm, respectively, which were 1.71% and 1.96% different from the 1 259.7 and 1 779.4 kN/mm calculated by ANSYS. This finding indicated that the calculation results of the proposed method were accurate and reliable. The least squares method was utilized to fit the correlation model of the saddle displacement and temperature, and the results showed an obvious linear correlation between the two. The analysis of single-day monitoring data revealed that the variation in the main cable force on that day was less than 540 kN, and the finite element calculation results were consistent with those obtained by the proposed method, with a maximum difference of 8.8 kN, accounting for only 1.66% of the daily variation in the main cable force. This finding confirmed that the method achieved high accuracy. The histogram of splay saddle displacement data presented a bell-shaped curve with a high center and low sides, approximately symmetrical, and the trend and magnitude of the probability density function curve fitted based on the normal distribution were consistent. In the Q‒Q plot, most of the data points were located within the 95% confidence interval, densely distributed in the middle, and symmetrically aligned near the reference line on both sides. It was concluded that the splay saddle displacement data followed a normal distribution. The analysis of 181 days of monitoring data showed that the main cable force changed randomly, with no evident pattern or long-term trend. The maximum, minimum, and average values of the main cable force were 117 559, 114 712, and 115 919 kN, respectively. The average and maximum stresses of the main cable reached only 67.5% and 68.5% of the standard limits, indicating that the structure is in good condition and has sufficient bearing capacity. The variation range of the main cable force was 2 847 kN, accounting for only 2.5% of the mean cable force, indicating that most of the main cable force originated from dead load, while the variation caused by live load was minimal. Compared to the measured data, the established ARIMA model provided a smoother predicted value curve for the last 60 days of the main cable force. The maximum prediction deviation was 358 kN, representing only 0.3% of the mean main cable force, indicating that the established ARIMA model achieved high prediction accuracy.ConclusionsThe proposed method for calculating the main cable force of a suspension bridge based on splay saddle displacement is highly accurate and reliable. There is a clear linear correlation between splay saddle displacement and temperature. The data of splay saddle displacement after temperature normalization follow a normal distribution, and the fluctuation of the main cable force remains minimal during the monitoring period. Most of the main cable force is attributed to the dead load. The main cable structure of the bridge is in good condition and possesses sufficient bearing capacity. The established ARIMA model can accurately predict the magnitude and trend of changes in the main cable force, providing a reliable basis for assessing structural anomalies and performance degradation.关键词：bridge engineering;Long-span suspension bridge;Main cable force monitoring;Data processing and analysis;Time Series Model170|452|0更新时间：2026-01-29
ENGINEERING AND ENVIRONMENT IN CHALLENGING MOUNTAINOUS REGIONS
- “在深渊微生物采样领域，研发了全海深微生物多级过滤取样装置，有效提升了样本质量，为深渊微生物基因库建立提供技术支撑。”摘要：ObjectiveThe oceans are the origin of life and the largest gene pool on Earth. However, human exploration and research on the oceans remain very limited. Based on relevant data, only 5% of the ocean area has been explored, and merely 0.01% has been sampled. The abyssal ocean, with a depth greater than 6 000 m, accounts for 45% of the total ocean depth. The high-pressure, low-temperature, and dark ecological conditions of the abyssal zone give rise to a unique ecosystem that contains the largest microbial reservoir in the entire aquatic system, encompassing nearly 75% of the prokaryotic biomass and 50% of its production. Numerous studies demonstrate that in situ filtration represents the most effective approach to obtaining microbial samples from seawater. However, most existing sampling devices function only at depths shallower than 6 000 m, and very few operate reliably for extended periods in abyssal waters. This study introduces a seawater microbial sampling device applicable at all depths, featuring in situ filtration and preservation capabilities. The degradation of microbial RNA is effectively prevented by adding RNAlater preservation solution to the filtered container, ensuring sample quality and supporting subsequent genetic material research. The device is successfully employed for in situ enrichment of microorganisms at a sea depth exceeding 10 000 m in the Mariana Trench.MethodsThe principal scheme design and detailed design of each component of the in situ filtration, sampling, and preservation system for abyssal microorganisms were conducted. First, a multi-stage membrane system was adopted to enhance filtration efficiency, and the inlet and outlet check valves, along with volume compensator structures, were utilized to achieve contamination-free and pressure-surge-free sampling. Second, the advantages and disadvantages of the membrane-based and cartridge-based schemes were compared, and flat filters and microporous membranes were selected as the filtration components. Then, a deep-water motor and a deep-water pump were designed for operation in abyssal environments, and an in situ fixed preservation structure was developed to process the collected microbial samples and improve sample quality.Results and discussionsThe study independently designed and developed a microbial sampling system with in situ filtration and in situ preservation functions suitable for 10 000 m abyssal environments by referring to and summarizing domestic and international research on deep-sea microbial samplers. The work included the conceptual scheme and detailed design of each component. For practical application, high-pressure environment tests were performed for both the deep-water motor and the complete system, followed by field sea trials during the TS15 and TS21 voyages. The entire system was successfully tested in the field, verifying its capability to operate in 10 000 m-class abyssal environments. Compared to existing deep-sea microbial samplers worldwide, the system designed and developed in this study demonstrated the following advantages: The entire system was designed for 10 000 m abyssal operations and was capable of functioning at all sea depths; the multi-stage membrane filtration system was successfully applied in actual sea trials, and the combination of filter membranes with different pore sizes enabled the preliminary screening of microorganisms in seawater, reduced clogging in small-pore membranes, and enhanced filtration efficiency. The design of the inlet and outlet one-way valve structures and the volume compensator structure ensured that the system did not come into contact with seawater at non-target depths during the sampling process, fully achieving contamination-free and pressure-stable operation. In addition, the in situ fixed preservation structure injected RNAlater preservation solution into the samples immediately after filtration to effectively prevent RNA degradation during recovery and improve overall sample quality.ConclusionsThis study develops a set of in situ microbial multi-stage filtration sampling devices for operation across the entire sea depth. The system primarily utilizes a multi-stage membrane structure and an in situ preservation structure to enable the collection of enriched, high-quality microbial samples. The plunger pump, deep-sea oil-filled motors, RNAlater syringe pumps, and other functional components are designed and developed to meet the actual working requirements. High-voltage equipment is employed to test the oil-filled motors, and current values are obtained under different pressures, providing a foundation for structural adjustment and improvement. In addition, high-pressure testing of the oil-filled motor is conducted using specialized high-pressure equipment, and the current values under varying pressures are recorded to support further optimization of the sampling device. The device was successfully deployed twice in the Mariana Trench at depths exceeding 10 000 m and effectively obtained abyssal microbial samples, establishing a reliable basis for building an abyssal microbial gene pool and supporting potential application research. Future improvements of the sampling system can focus on structural optimization and weight reduction.关键词：full-sea-depth;multistage filtration;microorganism;in situ sampling;Mariana Trench268|391|0更新时间：2026-01-29
- “在清洁能源领域，专家提出新型洋流能涡激振动能量俘获系统，揭示了其非线性动力学行为，为水下微型无线传感器供电提供新方案。”摘要：ObjectiveDue to the development of clean and renewable energy, the harvesting of ocean current energy, particularly vortex-induced vibration energy suitable for low flow rates, continues to emerge as a research hotspot for powering microwireless underwater sensors. However, linear systems face limitations such as low power output and a narrow energy harvesting range, which prevent them from meeting the required electrical demand. In contrast, nonlinear systems possess a lower energy barrier, allowing them to produce significant motion under weaker excitation and to operate effectively across a broader flow velocity range. Therefore, this study proposes a nonlinear vortex-induced vibration energy harvesting system designed to generate and supply greater power for underwater devices. The dynamic characteristics and power generation performance of the proposed system are investigated.MethodsThe Van der Pol equation was utilized to describe the coupling effect between the fluid and the blunt body, and a nonlinear restoring force was generated by the geometric nonlinearity caused by the oblique arrangement of linear springs. An underwater vortex-induced vibration power generation system incorporating a nonlinear restoring force was proposed. The system converted the lateral reciprocating motion of the blunt body into the unidirectional rotational motion of the generator through unidirectional bearings, gear rack mechanisms, speed-increasing boxes, and rotor generators. The fluid-solid-electrical coupling dynamic equation of the system was established, and the static equilibrium point bifurcation and different steady-state motion intervals of the blunt body's nonlinear vibration were obtained using nonlinear vibration theory. First, stability analysis was conducted on the system, and the parameter ranges under mono-stable, bi-stable, and tri-stable conditions were determined. The amplitude ratios of the system under the three steady states were compared, and it was found that the system has a strong power generation advantage in tri-stable motion. Then, using numerical simulation and the fourth-order Runge-Kutta method, the effects of fluid parameters (water flow velocity u) and structural parameters (mass ratio m*, damping ratio δ) on the bifurcation characteristics of the system were studied in detail through time-domain diagrams, phase diagrams, bifurcation diagrams, and Poincaré cross-sections of blunt body vibration. Then, the influence of unidirectional bearings on system speed was analyzed, followed by the analysis of the effects of mass ratio and damping ratio on power generation.Results and DiscussionsThe variation in the stable motion characteristics of the bluff body with parameters a and b was obtained based on the potential energy function of the system. The barrier heights of the potential energy function differed under various stable motion conditions. In comparison, the potential well depths of the potential energy function under tri-stable motion were lower, making cross-well motion more likely to occur. Based on the amplitude response analysis, the tri-stable energy harvesting system exhibited a wider operating range and better dynamic output performance than the bi-stable and mono-stable systems. It was found that the motion state of the system changed with different flow velocities by analyzing the influence of ocean current velocity on the vibration characteristics of the blunt body. Minor periodic motion occurred within wells, while large-scale chaotic motion appeared between wells, and periodic motions of 8, 4, 2, and 1 were observed within the wells. It was found that the bluff body underwent large-scale inter-well motion in the resonance intervals of 0.540 m/s
- “在航道运输领域，专家提出了基于注意力机制的LSTM多因素滨海航道水深预测模型，为提升航道运输智能化和数字化管理水平提供技术支撑。”摘要：ObjectiveAccurate prediction of water depth in coastal waterways is essential for ensuring the safety and efficiency of construction and transportation activities, particularly in environments characterized by complex and dynamic hydrological conditions. The Pinglu Canal, an inland waterway affected by tidal influences, serves as an example where traditional prediction models fail to accurately forecast water depth under complex hydrological conditions. This study presents a water depth prediction model based on an attention-enhanced long short-term memory (LSTM) network, which is integrated into a decision-support platform for real-time channel transportation management.MethodsFirst, key hydrological factors, including upstream discharge, daily rainfall, tidal current velocity, and tidal level, were incorporated to construct the LSTM-based coastal waterway depth prediction model. The raw hydrological data were preprocessed to address issues such as missing values, noise, and irregular time intervals, ensuring suitability for time-series modeling. Then, the LSTM model was utilized to capture the long-term temporal dependencies within the data, allowing the model to account for complex interactions among various hydrological variables over time. An attention mechanism was introduced to optimize the model's performance. This mechanism improved the model architecture by enabling it to dynamically adjust the weight of each feature at every time step, prioritizing the most relevant factors based on the current data. The attention mechanism enhanced both the accuracy and stability of the model, particularly for ultra-long-term water depth forecasting under dynamic and complex coastal hydrological conditions. Finally, the optimized model was embedded into a transportation decision-support platform, allowing for real-time water depth prediction, dynamic correction of forecasts based on new data, and navigable time window evaluation. The model's effectiveness was validated through comparative analyses with existing prediction models and field measurements, demonstrating superior accuracy and reliability in predicting waterway depth.Results and DiscussionsThe results demonstrated that at two monitoring points, 5 km from the coast (Point 1) and 30 km inland (Point 2), the traditional LSTM model exhibited larger prediction errors, especially in long-term forecasts. The MAE ranged from 0.07 m to 1.08 m for short-term predictions and from 0.12 m to 1.74 m for long-term forecasts. The model also tended to overestimate water depth. In contrast, the attention mechanism-based LSTM model consistently kept the MAE below 0.15 m, even under sudden rainfall or upstream discharge events, showing enhanced accuracy in both short-term fluctuations and long-term trends. The model's performance across seasonal variations further highlighted its robustness. During the dry season, the MAE was reduced by 64.68%, and during the wet season, it decreased by 72.36%. The RMSE was also reduced by 67.51% and 73.39% in the respective seasons, while the R² coefficient improved by 2.18% and 5.60%. This demonstrated the model's adaptability to both stable and volatile water conditions. The attention mechanism-based LSTM model significantly outperformed traditional LSTM models in predicting waterway depth. Compared to the traditional model, the MAE of was reduced by 65.00%～72.00%, and the R² coefficient increased by 2.20%～5.60%, demonstrating superior predictive accuracy and stability. This improvement was particularly evident under complex hydrological conditions, where the model effectively captured non-linear and dynamic relationships among key features such as tidal flow speed, daily rainfall, and tidal water level. In addition, when compared to a single feature vector model, the three-feature vector combination (daily rainfall, tidal flow speed, and tidal water level) resulted in an MAE of no more than 0.15 m and an R² coefficient of no less than 0.99, substantially improving the model's accuracy and stability for predicting waterway depth under complex coastal hydrological conditions. Finally, when integrated into the waterway transportation decision-support platform, the model's capabilities, such as real-time water depth prediction, dynamic correction, and navigable time window evaluation, substantially enhanced the platform's effectiveness. This integrated system provided reliable and accurate information for waterway transportation management, contributing to safer and more efficient navigation in complex coastal environments.ConclusionsThis study presents an attention mechanism-based LSTM model for predicting waterway depth in complex coastal environments. The proposed model substantially improves prediction accuracy and stability, especially under dynamic hydrological conditions. The model effectively adapts to diverse waterway conditions by incorporating key hydrological features, ensuring higher precision in depth forecasting. When implemented within a waterway transportation decision-support platform, the model enables real-time prediction, dynamic correction, and navigable time window assessment, enhancing the intelligence and digital management of waterway transportation systems. This research provides reliable technical support for future engineering applications.关键词：Pinglu Canal;Waterway construction and transportation;Waterway depth prediction;attention mechanism;LSTM model;Waterway transportation decision support platform72|341|0更新时间：2026-01-29
- “在海上风电基础领域，专家通过模型试验和数值模拟，揭示了高桩承台基础受力特性和荷载分配机制，为服役期间安全性评价和基础设计优化提供理论依据。”摘要：ObjectiveThis study aims to conduct an in-depth investigation into the load transfer mechanisms and distribution behavior of high-pile cap foundations under cyclic loading conditions. A systematic analysis is conducted to clarify the evolution of stiffness in the sand‒foundation interaction system, with particular emphasis on the response to repeated external loads. The findings provide a theoretical foundation for assessing the operational safety of offshore wind turbine foundations supported by high-pile caps during their service life. In addition, this study provides practical insights and technical references for the design, optimization, and risk mitigation of such foundation systems in future offshore wind energy projects.MethodsGeotechnical centrifuge modeling was employed to simulate the performance of high pile cap foundations under both monotonic and unidirectional multi-stage cyclic horizontal loading. The tests were conducted at a scaling ratio of 1:100 using dry sand to replicate the seabed substrate. A strictly controlled loading rate of 0.06 mm/s was applied to maintain quasi-static conditions. Displacement responses were recorded at a sampling frequency of 10 Hz to capture detailed deformation behavior. The experimental setup incorporated a horizontal loading mechanism, laser displacement sensors, and axial force transducers installed around the pile cap. In addition, strain gauges were mounted along the lengths of the piles to measure internal force variations. This instrumentation allowed for a comprehensive assessment of the foundation's mechanical performance, including stiffness degradation under high-amplitude cyclic loading, accumulation of permanent displacements, and development of tilt. Complementing the physical experiments, a finite element model was developed to numerically simulate the tests. The model was calibrated using experimental data to ensure accuracy. Key parameters such as axial force and bending moment distributions along the pile shafts were extracted from the simulation results. The numerical analysis served to validate trends observed in the centrifuge tests and provided additional insight into the underlying load distribution mechanisms that governed axial and flexural behavior in the pile group.Results and Discussions1) Under monotonically increasing horizontal load applied to the cap, the foundation displacement initially exhibited a linear relationship with the applied load. Beyond a load level of 0.46Fu, the load displacement curve demonstrated pronounced nonlinearity, indicating the onset of plastic deformation within the surrounding soil mass. The bearing capacity of each individual pile within the group was progressively mobilized, culminating in an ultimate horizontal load capacity of 17.88 kN for the entire foundation system. 2) The axial force measured in the piles decreased significantly with increasing embedment depth. This attenuation was particularly evident along the upper portion of the piles, with axial forces approaching zero at a depth of approximately 200 mm. In contrast, the bending moment increased initially with depth, reached a maximum value at around 30 mm below the mudline, and then decreased gradually. Among the instrumented piles, Piles 1#, 2#, and 8# were subjected to net tension forces, while the remaining piles experienced compression. All piles exhibited bending moments in the same direction, indicating consistent group behavior. When symmetric horizontal loading was applied with respect to the cap central axis, both axial forces and bending moments displayed symmetric distribution patterns across the pile group. 3) Under cyclic loading amplitudes of 0.2Fu, 0.4Fu, and 0.6Fu, the cumulative displacement of the foundation increased steadily with the number of loading cycles. However, when the cyclic load amplitude was subsequently reduced to 0.2Fu, a slight recovery in cumulative displacement was observed. The unloading stiffness of the pile cap remained largely stable throughout the tests at 0.2Fu and 0.4Fu. In contrast, under 0.6Fu loading, significant particle rearrangement occurred in the sand surrounding the piles, which led to a reduction in soil density and stiffness. Subsequent reduction of the load amplitude to 0.2Fu resulted in a self-healing effect, where the foundation stiffness showed partial recovery. 4) Similar to the accumulation of displacement, the tilt angle of the high pile cap increased gradually under cyclic loading at all three amplitude levels. The largest increase in tilt occurred during the initial cycles following an increase in load amplitude. Under constant-amplitude unidirectional cyclic loading, the tilt angle exhibited a steady increase with cycle count. When the load amplitude was reduced to 0.2Fu, the sand surrounding the piles underwent additional compaction under sustained low-amplitude cyclic loading, resulting in a slight increase in overall soil stiffness and a corresponding reduction in cap tilt by approximately 0.28°.ConclusionsThis study integrates centrifuge physical modeling and finite element numerical simulations to examine the behavior of high-pile cap foundations subjected to horizontal monotonic and cyclic loading conditions. The research clarifies the load transfer trajectories and distribution mechanisms among piles within the group. In addition, the study quantitatively evaluates the influence of cycle number and load amplitude on horizontal displacement, soil stiffness, cumulative deformation, and tilt. Based on the results, the following engineering recommendations are proposed: 1) Project construction efficiency can be improved by optimizing the bearing capacity through rational reduction in pile length or pile number without compromising overall safety. 2) The design phase should give particular attention to the bearing capacity of piles oriented in the predominant wind and wave directions. 3) Although the number of loading cycles affects the unloading stiffness of the pile cap, the influence of load amplitude is more significant; therefore, foundation stiffness variations under extreme environmental conditions require careful consideration.关键词：high-rise cap foundation;centrifugal model tests;lateral monotonic and cyclic loading;load transfer and distribution;cumulative displacement59|38|0更新时间：2026-01-29
- “全球能源需求激增，海上漂浮式光伏系统发展迅速，专家探讨其技术挑战与未来方向，为能源转型提供新思路。”摘要：Significance With the increasing global energy demand, floating photovoltaic (FPV) systems have continued to advance rapidly over the past decade, driven by their superior energy output and efficiency. Among these systems, offshore FPV installations stand out for their wider deployment potential and access to abundant solar resources, establishing them as a crucial component of future energy solutions. However, the harsh marine environment presents stringent requirements for the structural design, installation, and operational strategies of offshore FPV platforms. This study provides comprehensive guidance and technical references to support the design and engineering practices of future offshore FPV systems.ProgressThe development status and structure of offshore floating PV were briefly reviewed, and the challenges of offshore FPV development were analyzed. The research progress of key issues mainly focused on the design analysis, construction, operation, and maintenance technology of offshore FPV platforms. 1) Offshore FPV platforms consisted of components such as buoyancy units, supporting structures, connection mechanisms, and mooring systems. The material selection and configuration of these components significantly affected the platform's stability, durability, and cost-effectiveness. Each material was carefully evaluated for its strength, corrosion resistance, and economic feasibility under specific marine conditions. 2) Offshore FPV generally consisted of an array of multiple floating bodies connected by components, and the structural integrity of the platform was influenced by various factors such as monomer strength and connector design. Using a rigid module and flexible connector (RMFC) model, the array coupling dynamic response analysis and flexible connector design were conducted to examine the system's performance. 3) Offshore FPV was mainly deployed in near-shore shallow waters where small tidal level variations caused significant fluctuations in water levels, leading to drastic changes in mooring system tension. In addition, non-uniform seabed topography required asymmetric mooring system designs. 4) The size of offshore FPV arrays reached hundreds or even thousands of meters, and shallow water effects combined with seabed topography variations resulted in inhomogeneous wave fields. These conditions demanded higher technical requirements for hydrodynamic performance optimization, connection safety improvement, and mooring safety enhancement through coupled dynamic response analysis using inhomogeneous wave field simulation methods. 5) The construction and installation methods currently used for offshore oil and gas platforms and wind power systems, such as lifting and floatover techniques, were not fully applicable to offshore FPV platforms. Existing offshore construction methods required significant adaptation to accommodate the lightweight structures and dynamic surface characteristics of FPV platforms, which necessitated the development of installation technologies specifically designed for their unique structural forms and service environments. 6) The marine environment posed numerous challenges to operating and maintaining offshore PV systems. With reference to the risk management and emergency response mechanisms of offshore oil and gas platforms, as well as the intelligent monitoring and predictive maintenance technologies of wind power platforms, a series of operation and maintenance strategies, including real-time monitoring, preventive maintenance, cleaning and biological attachment prevention, equipment replacement, and maintenance, were implemented and optimized to ensure the long-term stable operation of the system.Conclusions and ProspectsOffshore FPV demonstrates significant development potential due to its vast spatial capacity, high energy quality, and hydrocooling effect. However, the harsh marine environment imposes greater demands on the structural design and engineering practices of offshore FPV platforms. This condition indicates that offshore FPV will encounter more complex wind and wave conditions, as well as higher transportation and maintenance costs. Therefore, ensuring the long-term stable operation of offshore FPV platforms in deep-sea environments, minimizing construction and maintenance costs, and enhancing the platforms' power generation efficiency have become essential trends in the advancement of offshore PV key technologies. In addition, collaborative development serves as an effective approach to improving the efficiency of comprehensive marine energy utilization. These trends collectively highlight the strong developmental value of offshore FPV platforms as a promising clean energy solution for the future.关键词：FPV system;development status;technical key issues;future development trend193|741|0更新时间：2026-01-29
SUSTAINABLE DEVELOPMENT AND UTILIZATION OF MARINE RESOURCES
- “在道路安全领域，研究人员基于YOLOv7-tiny提出了MFF-YOLO算法，有效平衡了检测精度与运算复杂度，为移动终端设备上的道路缺陷检测提供参考。”摘要：ObjectiveRoad damage detection serves as the premise and foundation of road maintenance and is essential for road safety and timely repair. In real-world scenarios, mobile terminal devices are more practical for detection tasks due to constraints in working environments. However, the limited computational power of these devices makes it challenging to apply detection algorithms with high computational complexity. This study proposes MFF‒YOLO, a lightweight road damage detection algorithm that features multiscale feature fusion based on YOLOv7-tiny to address the issue of existing road damage detection algorithms struggling to balance detection accuracy and computational efficiency, which hinders their deployment on mobile devices.MethodsFirstly, road damage often occurred in complex backgrounds with high levels of noise, so it was crucial to extract road damage features accurately. The multiscale feature extraction block (MFFBlock) and the downsampling block (DSB) were designed to improve the feature extraction capability of the algorithm. MFFBlock employed a multiscale feature extraction and fusion strategy that integrated a multi-branch structure with various types of convolution to produce outputs with different receptive fields, extracting both global and local feature information. DSB combined max pooling with convolution at a stride of 2 to maximize the retention of effective feature information and to ensure computational stability during downsampling. Based on MFFBlock and DSB, an efficient multiscale feature extraction backbone network (MFEnet) was constructed, utilizing MFFBlock of different sizes at various stages to enhance the algorithm’s multiscale feature representation ability. Secondly, after extracting road damage features, it was equally crucial to efficiently fuse these features in the neck network. The Slim-Neck design paradigm was used in the neck feature fusion network, utilizing GSConv and VoV‒GSCSPC modules to aggregate features while making the network more lightweight without losing important information. In addition, a novel feature selective fusion structure (FSF‒PAFPN) was proposed. Building on PAFPN, FSF‒PAFPN introduced a feature selective fusion mechanism (FSF), which used a channel attention (CA) module to selectively fuse shallow and deep features across layers at the same resolution level, achieving simple yet effective multiscale feature fusion. Finally, the K-Means algorithm was utilized to cluster the RDD2022 dataset, obtaining anchors that were more consistent with the shape characteristics of road damage objects, reducing the training difficulty, and improving detection accuracy. The RDD2022 road damage dataset was used for algorithm training and verification and consisted of a total of 23 767 damage images. These images, which included four typical road damage types: longitudinal cracks (D00), transverse cracks (D10), alligator cracks (D20), and potholes (D40), were divided into training, validation, and test sets in a ratio of 8:1:1.Results and DiscussionsThe results from the ablation experiment showed that using K-Means to re-cluster the RDD2022 dataset for generating anchors improved mAP@0.5 by 0.5%. In addition, implementing MFEnet as the backbone feature extraction network increased mAP@0.5 by 0.6%, while its parameters and FLOPs were 5.59×106 and 12.3×109, reduced by 7.1% and 6.8%, respectively. Adopting the Slim-Neck design paradigm enhanced mAP@0.5 by 0.5%, while reducing parameters and FLOPs by 20.0% and 21.1%, to 4.47×106 and 9.7×109, respectively. On this basis, adopting the FSF-PAFPN feature fusion structure further improved mAP@0.5 by 0.7%, with parameters and FLOPs increased to 4.51×106 and 9.8×109, respectively. The results of the algorithm performance comparison based on RDD2022 showed that, in terms of accuracy, MFF‒YOLO outperformed all other evaluation metrics except for recall (R) and average precision (AP) in the D40 category. The AP for the D00, D10, and D20 categories reached 60.8%, 58.9%, and 68.6%, respectively, with a precision (P) of 64.7%. It achieved the highest mAP@0.5 of 60.1%, an improvement of 2.3 percentage points compared to YOLOv7-tiny. In terms of computational complexity, MFF‒YOLO also excelled, with parameters and FLOPs of 4.51×106 and 9.8×109, respectively, values that were only slightly higher than YOLOv8n, YOLO-LWNet-s, and LE-YOLOv5. Compared to YOLOv7-tiny, these metrics were reduced by 25.1% and 25.8%, respectively. In addition, MFF‒YOLO reached a detection speed of 81 frames per second, maintaining an impressive real-time detection capability. Comparing the actual detection effects of YOLOv7-tiny, YOLO-LWNet-s, LE-YOLOv5, and MFF‒YOLO algorithms indicates that the detection performance of MFF‒YOLO exceeded that of the other algorithms. MFF‒YOLO accurately located road damage objects with high confidence and demonstrated strong performance even when the features were not obvious or when the background was complex.ConclusionsThe results demonstrated that the MFENet proposed in this study can effectively enhance the network’s multiscale feature extraction capability while reducing computational complexity. The Slim-Neck design paradigm ensured the aggregation of features while maintaining a lightweight network structure. The FSF‒PAFPN structure achieved more efficient multiscale feature fusion and improved the algorithm’s ability to characterize road damage features. Accordingly, MFF‒YOLO significantly improved detection accuracy while reducing computational complexity. Considering accuracy, computational complexity, and detection speed, MFF‒YOLO achieved a balanced performance and is more suitable for road damage detection on mobile devices, providing a valuable reference for mobile terminal road damage detection.关键词：road damage;object detection;YOLOv7-tiny;lightweight382|778|0更新时间：2026-01-29
- “在大规模随机粒子近邻搜索领域，专家提出了MAD-index-sort缓存优化策略，通过自动改变排序方向，性能最高可提升30.3%。”摘要：ObjectiveWhen utilizing a kd-tree for large-scale random particle nearest neighbor search, particles with closely aligned index values within the computational domain can become spatially distant, resulting in significant variations in the kd-tree search path over a short time frame. Therefore, this divergence reduces the efficiency of accessing node data and ultimately impedes the effectiveness of kd-tree nearest neighbor searches. This inefficiency becomes particularly evident in non-uniform or randomized particle distributions, where traditional cache optimization strategies show limited effectiveness. Maximum dispersion dimensionality reduction in principal component analysis (PCA) is introduced to address this issue. It employs the mean absolute difference (MAD) as the dispersion measure and proposes a novel cache optimization strategy named MAD-index-sort. MAD-index-sort improves cache utilization and enhances the overall performance of kd-tree-based nearest neighbor searches by dynamically reordering particle index values based on spatial distribution.MethodsThe MAD-index-sort strategy introduced in this study utilized the dimensionality reduction principles found in PCA to achieve optimal data reordering. Specifically, the method employed MAD as a dispersion measure, which served as the key metric for determining the most appropriate dimension along which the particle data should be reordered. By calculating the dimension with the highest MAD value, identifying the axis along which particle distribution exhibited the greatest variance, the algorithm ensured that particles with closer spatial proximity were assigned similar index values, thus optimizing data access during kd-tree searches. This reordering process, in turn, enabled more efficient cache usage because it minimized cache misses and increased the likelihood of retrieving relevant data from faster memory. The proposed approach was then integrated with the automatic termination criterion for the kd-tree framework, which further streamlined the nearest neighbor search process by automatically halting the search once optimal criteria were met, reducing unnecessary computations. A series of rigorous experimental trials was conducted using various particle distribution models, including uniform, random, and clustered configurations to comprehensively evaluate the effectiveness of this approach. Key performance metrics, such as cache miss rates, search path divergence, and total computational time, were carefully monitored and compared against baseline methods, including Z-index-sort, which was traditionally employed for uniform grid methods, and the standard unsorted kd-tree approach.Results and DiscussionsThe experimental results revealed that the MAD-index-sort strategy consistently surpassed both traditional unsorted kd-tree methods and alternative sorting strategies such as Z-index-sort across various particle distribution scenarios. Specifically, MAD-index-sort improved search efficiency by dynamically adjusting to particle dispersion characteristics, which led to a marked reduction in cache misses and overall computational time. When applied to randomized particle distributions with high spatial divergence, MAD-index-sort significantly decreased the cache miss rate. Compared to the unsorted kd-tree, the cache miss rate was reduced by 24.2%, while compared to Z-index-sort, the reduction was 18.6%. In addition, the improved performance was particularly evident in computational fluid dynamics (CFD) simulations, where particle positions frequently shifted irregularly due to external forces. In this context, MAD-index-sort dynamically adapted to the shifting particle arrangement, reduced cache miss rates by 20%, and simultaneously cut the total computational time by 15% compared to the unsorted kd-tree method. The algorithm was versatile enough to handle both static and dynamic conditions, making it highly reliable for real-time applications where particle distributions tended to change unpredictably. In addition, the results indicated that MAD-index-sort performed exceptionally well in scenarios involving highly dispersed particle clusters. In these cases, the search time was shortened by 28.5% compared to the unsorted kd-tree and by 22% compared to Z-index-sort. This efficiency was attributed to MAD-index-sort's ability to reorder particles based on the dimension with the greatest variance, ensuring that spatially close particles were assigned similar index values. Therefore, the cache access pattern was optimized, which led to fewer cache misses and improved search times. In addition, MAD-index-sort demonstrated a significant performance boost, achieving search times 32.8% faster than unsorted kd-tree methods, with a 30% reduction in cache miss rates. These findings indicated that MAD-index-sort was effective in handling irregular particle distributions and further demonstrated its versatility and adaptability. Further analysis revealed that MAD-index-sort consistently improved the cache hit rates across all tested particle distribution models. For example, in scenarios characterized by extreme spatial variance, the strategy attained an average 27.3% reduction in computational time, coupled with a 17% improvement in cache hit rates, compared to traditional methods. This broad applicability of the MAD-index-sort strategy highlighted its effectiveness in a wide range of computational tasks that required efficient data access, particularly in systems where memory hierarchy played a crucial role in performance. Accordingly, the experimental results validated the efficiency and flexibility of the MAD-index-sort strategy, showing that it significantly reduced cache misses and improved search performance across a wide variety of particle distribution scenarios. MAD-index-sort optimized cache utilization and minimized computational overhead by dynamically adjusting the index sorting process based on particle dispersion, making it a practical tool for kd-tree-based nearest neighbor searches in different environments.ConclusionsThe introduction of the MAD-index-sort strategy represents a meaningful enhancement in kd-tree-based nearest neighbor search optimization, particularly for scenarios involving random or highly dispersed particle distributions. The strategy improves cache efficiency, reduces computational overhead, and enhances overall search performance by incorporating PCA-inspired dimensionality reduction and using the MAD metric to dynamically reorder particle indices. The experimental results indicate that MAD-index-sort can improve search efficiency by up to 30.3% compared to unsorted kd-tree searches, making it a highly adaptable and versatile tool for a wide range of computational applications. The implications of this research extend beyond kd-tree nearest neighbor searches, as the underlying principles of dynamic index reordering and cache optimization can be applied to other computational problems where data access efficiency is critical. Future work can investigate integrating MAD-index-sort with advanced cache management techniques, such as predictive caching algorithms or hybrid data structures, to enhance performance. In addition, further research on reducing the computational complexity of distance calculations, particularly in high-dimensional spaces, can yield greater efficiency gains and make kd-tree-based methods more suitable for large-scale, real-time applications.关键词：kd-tree;particle nearest neighbor search;cache optimization;particle index value sorting176|359|0更新时间：2026-01-29
- “在高速列车轴箱轴承故障诊断领域，研究者提出了ITR-Net深度迁移学习方法，有效提升了不同工况下轴承故障诊断的准确率，为迁移学习在轴箱轴承故障诊断应用提供了新途径。”摘要：ObjectiveEfficiently assessing the health status of axlebox bearings in high-speed trains is crucial for maintaining reliable train operation. Current deep learning-based bearing fault diagnosis faces two significant challenges: it requires many labeled actual fault samples, and the training and test sets need to satisfy independent and identically distributed conditions. Transfer learning relaxes the limitations of these issues for intelligent bearing fault diagnosis, and it utilizes transferable knowledge learned from existing labeled datasets to accomplish tasks within different but similar unlabeled datasets. However, the current transfer learning model based on a single source domain suffers from underutilization of labeled data, reduced transfer diagnosis accuracy, and potential negative transfer when the dataset distribution varies significantly. This study proposes ITR-Net (Inception Transformer and ResNet), a multi-source domain deep transfer learning method that integrates IFormer (Inception Transformer) and ResNet for high-speed train axlebox bearing fault diagnosis research.MethodsThe method selected supervised labeled data under various operating conditions in the multi-source domain, and first obtained the time-frequency spectrograms of the one-dimensional vibration signals of the bearings as the model input by using the continuous wavelet transform based on the Morlet wavelet basis. The main structure of the proposed network framework consisted of three parts, namely the common feature extractor, the specific feature extractor, and the specific classifier. The common feature extractor adopted the IFormer network, which used the classical structure of the convolutional neural network (CNN) with depth-wise separable convolution (DWConv) and maximum pooling to capture the local information of the input data. It employed the multi-head self-attention (MSA) mechanism in the Transformer network to capture the global information of the input data, so the IFormer network mined more comprehensive feature information. The common feature extractor was utilized to extract domain-invariant features in different source and target domains. The specific feature extractor adopted the classical convolutional neural network ResNet, which efficiently extracted the feature information of the input patch while effectively avoiding gradient disappearance or gradient explosion that can have occurred with the increase of network depth. The specific classifier was utilized to output the classification results for different source domains and the target domain, which facilitated subsequent metrics to measure the distance between the different predicted labels output. In applying the transfer strategy, the study optimized the multi-kernel maximum mean difference (MK-MMD) after the common feature extractor to align the overall distributions of the source and target domains; optimized the local maximum mean difference (LMMD) after the specific feature extractor to enable the model to extract fine-grained information from the input features; optimized the cross-entropy loss (CEloss) after the specific classifier to improve the model classification accuracy on the source domains; and optimized the mean-squared error (MSE) loss after the specific classifier to reduce the differences between the predicted labels of the target domain output by different classifiers.Results and DiscussionsSix multi-source domain transfer tasks were set using the Integrated High-Speed Train Bearing Experiment Station and the Integrated Power Transmission Fault Diagnosis Experiment Station datasets to demonstrate the effectiveness of the proposed method. Analyzing the results of multi-source domain transfer and single-source domain transfer showed that the effect of multi-source domain transfer was significantly better than that of single-source domain transfer. Comparing the proposed method ITR-Net with other popular transfer learning methods, namely deep adaptive networks (DAN), joint adaptation network (JAN), correlation alignment (CORAL), domain adversarial neural network (DANN), and Multi-feature spatial adaptation networks (MFSAN), the proposed method achieved an average transfer accuracy of 96.66% in six transfer tasks, while the comparative methods achieved 87.24%, 88.30%, 92.45%, 94.11%, and 93.35%, respectively. This result demonstrated the superiority of the proposed method. The t-distribution stochastic neighbor embedding (t-SNE) visualized the feature clustering of the target domain features extracted from the six migration tasks. It was observed that the target domain features in the proposed method achieved more distinct clustering based on different bearing fault types, and the overall clustering of the unsupervised target domain features under the same fault types was improved, which proved the method's effectiveness. In the ablation experiments, the average transfer accuracies of using MK-MMD, LMMD, and MSE alone were 92.30%, 93.19% and 93.18%, respectively; when MK-MMD and LMMD were utilized together, the average migration accuracy reached 95.26%; when the complete loss function was applied, the average accuracy reached the maximum of 98.63%. The ablation results proved that the adaptive migration strategy constructed using MK-MMD, LMMD, and MSE further enhanced the degree of domain feature alignment among multi-source domains, as well as between individual source domains and the target domain, resulting in the best migration learning effect.ConclusionsThe results showed that the proposed method can fully utilize the data information of multiple source domains, and the transfer using multiple source domains can effectively improve the diagnosis performance of faults in the target domain. The distributions of the source domains and the target domains can be aligned, and the ablation experiments confirmed the effects of different loss functions on the transfer performance of the network models by applying the MK-MMD, LMMD, CELoss, and MSE loss functions to construct the transfer strategy at different network stage positions. The results provide a new approach for applying transfer learning to axlebox bearing fault diagnosis.关键词：axlebox bearing;transfer learning;fault diagnosis;domain adaptation;feature learning449|583|0更新时间：2026-01-29
- “在河湖管理领域，专家提出了基于改进YOLO v5s的河道巡查图像识别模型，有效提高了河湖问题精准识别能力。”摘要：ObjectiveUtilizing advanced methods to rapidly and accurately identify issues such as river and lake disorder and water quality represents a key approach to improving the efficiency of river and lake management and the river chief system. This approach aims to effectively address problems related to limited observation range, low efficiency, and delayed response to river-related issues that occur with traditional manual patrols employing unmanned aerial vehicle (UAV) patrols integrated with image recognition technology. However, current challenges persist, including complex river surface environments, irregular shapes, and diverse forms of floating debris in rivers of varying sizes, widespread distribution of algal blooms, and concealed or easily obstructed locations of illegal sand mining activities. These factors make it difficult for traditional image recognition algorithms to accurately detect such issues. Therefore, this study proposes a river patrol image recognition algorithm based on an improved YOLO v5s, referred to as the YOLO v5s‒CDF algorithm.MethodsThe YOLO v5s‒CDF algorithm introduced several key improvements to enhance the detection performance of the original YOLO v5s model. First, the FocalNext module replaced the C3 module in the backbone network, incorporating depth-wise separable convolutions and dilated convolutions to improve feature extraction capabilities for small objects. The depth-wise separable convolution applied different convolutional kernels to each input channel, extracting important features from multiple channels, while the dilated convolutions increased the receptive field without adding parameters or computational cost, capturing broader contextual information. Second, the Context Aggregation attention mechanism was added between the neck and head structures to adjust the weights of input data, enabling the model to focus on key image information. This attention mechanism combined channel attention and spatial attention to refine the feature representation, enhancing the model's ability to capture critical details. Lastly, the Decoupled Head replaced the original coupled detection head, separating feature extraction and task prediction to accelerate network convergence and further enhance small object detection. The dataset used in this study consisted of aerial images captured by DJI Air 2s drones, covering various rivers in Hebei Province, China. The images were annotated using the Labelimg tool and divided into training and validation sets in an 8 ∶ 2 ratio. Data augmentation techniques, such as flipping, color transformation, and affine transformation, were applied to the training set to improve the model's robustness and generalization ability. The study evaluated the model's performance using metrics such as precision, recall, F1-score, Average Precision (AP), and mean Average Precision (mAP). The number of parameters and floating-point operations per second (FLOPs) were utilized to represent the model's complexity.Results and DiscussionsThe experimental results demonstrated that the YOLO v5s‒CDF model achieved a mean Average Precision (mAP) of 86.7%, surpassing the original YOLO v5s model by 4.1%. The improved model exhibited significant enhancements in both precision and recall. The precision increased from 85.9% in the original YOLO v5s model to 87.0% in the YOLO v5s‒CDF model, while the recall improved from 76.4% to 80.8%, indicating a substantial reduction in missed detections. When compared to other models, such as YOLO v7-tiny and YOLO X-s, the YOLO v5s‒CDF model outperformed them with mAP improvements of 20.4 percentage points and 8.3 percentage points, respectively. Among all target categories, the detection performance for river garbage showed the most significant improvement, with the average precision increasing by 6.1 percentage points, rising from 74.8% to 80.9%, and the precision also increasing by 2.3 percentage points, from 88% to 90.3%. In addition, the recall demonstrated an improvement of 7.3 percentage points, increasing from 64.0% to 71.3%, highlighting the YOLO v5s‒CDF model's remarkable enhancement in detecting small targets. For targets related to suspected illegal sand mining activities, the average precision increased from 90.2% to 92.2%; however, the precision experienced a decline from 88.7% to 86.7%, while the recall improved from 87.2% to 89.8%. Regarding green algae targets, the average precision rose from 82.8% to 86.9%, accompanied by an increase in precision from 81.0% to 84.0% and an improvement in recall from 77.9% to 81.3%. These improvements demonstrated the YOLO v5s‒CDF model's effectiveness in detecting several river issues across different object categories. The visual analysis of detection results further confirmed the superior performance of the YOLO v5s‒CDF model. When compared to the original YOLO v5s model, the improved model exhibited enhanced robustness and generalization capabilities in detecting river issues under complex environments, such as varied illumination conditions, water surface fluctuations, and reflections. The YOLO v5s‒CDF model successfully identified and localized a higher number of objects, particularly small and irregularly shaped floating garbage, which were often missed by the original model. The visual results aligned with the quantitative improvements observed in precision and recall. Ablation studies revealed the individual contributions of each introduced module, with the FocalNext module improving mAP by 1.8 percentage points, the Decoupled Head by 2.4 percentage points, and the combination of both modules by 3.7 percentage points. The addition of the Context Aggregation attention mechanism on top of the improvements from the FocalNext module and the Decoupled Head further improved the mAP by 4.1 percentage points compared to the original YOLO v5s model, while only slightly increasing the model's parameters and computational complexity. However, the model's performance in detecting submerged objects and illegal sand mining activities occurring at night required further improvement. Future research will focus on enhancing the model's applicability and accuracy under various environmental conditions, such as incorporating river topography changes as detection targets using 3D drone imagery and improving the collection of submerged object data.ConclusionsThe YOLO v5s‒CDF algorithm presents a viable technical approach to addressing challenges in river and lake supervision by integrating drone-based remote sensing technology with advanced image recognition methods. The enhanced model, which incorporates the FocalNext module, Context Aggregation attention mechanism, and Decoupled Head, demonstrates superior accuracy and robustness in detecting river debris, algal blooms, and potential illegal sand mining activities. It highlights the broad application potential of UAV remote sensing and target detection technologies in ecological environment management. In addition, integrating the YOLO v5s‒CDF algorithm with complementary technologies, such as water quality sensors, hydrological models, and geographic information systems (GIS), can provide a comprehensive framework for river health evaluation and management.关键词：YOLO v5s;small object detection;river garbage;FocalNext module;attention mechanism375|380|0更新时间：2026-01-29
INTELLIGENCE INTERDISCIPLINARY SCIENCE AND ENGINEERING
- “在超临界CO2相变致裂技术领域，专家通过实验研究了致裂管释压特性，为技术参数取值和破岩效果分析提供理论依据。”摘要：ObjectiveAs the core components responsible for energy release in supercritical CO2 phase transition fracturing, the fracturing tube faces challenges related to successful activation and the calculation of the pressure generated after activation. Successful activation of the fracturing tube is essential for effective energy release. After activation, the magnitude of the generated pressure is a key factor for evaluating the fracturing effect. The depressurization characteristics of the fracturing tube are mainly influenced by the activator quality, CO2 filling pressure, and energy release plate thickness. Current research primarily focuses on qualitative analyses of influencing factors, and no unified expression exists for calculating the peak pressure in the fracturing tube. Therefore, further investigation of the energy release characteristics of the fracturing tube is crucial for improving the efficiency and safety of supercritical CO2 phase transition fracturing.MethodsAn independently built experimental system was utilized to conduct cavity pressure tests under different activator qualityes (70, 90, and 120 g), CO2 filling pressures (7.5, 9.5, and 11.5 MPa), and energy release plate thicknesses (1.9, 2.6, and 3.4 mm) to determine the dynamic pressure response inside the fracturing tube. Correlation analysis methods were utilized to examine the relationships among activator qualityquality, CO2 filling pressure, energy release plate thickness, and peak pressure inside the fracturing tube. The effects of these parameters on the pressure rise time inside the fracturing tube were analyzed based on the pressure-volume-temperature characteristics of the gas. The calculation method for peak pressure was discussed in conjunction with the stress state and failure mode of the energy release plate. A quantified method for determining activator quality, CO2 filling pressure, and energy release plate thickness was proposed by comparing and analyzing the experimental results of this study with existing results reported in the literature.Results and DiscussionsUnder different experimental conditions, the peak pressure inside the fracturing tube ranged from 77.64 to 140.94 MPa, the pressure rise time ranged from 7 to 36 ms, and the time required for pressure decay from the peak value to atmospheric pressure was relatively short, ranging from 5 to 15 ms. The analysis indicated that the loading rate during supercritical CO2 phase transition fracturing for rock fracturing was approximately 104 MPa/s. The pressure response curve inside the fracturing tube consisted of a pressure rise stage before reaching the peak pressure and a release stage after the peak pressure. During the pressure rise stage, the pressure inside the fracturing tube initially increases slowly and then surges linearly with time. During the release stage, the pressure inside the fracturing tube initially drops rapidly and then gradually decreases to atmospheric pressure over time. The correlation between peak pressure and both activator quality and CO2 filling pressure was weak, whereas a strong correlation existed between peak pressure and energy release plate thickness. When the activator quality and energy release plate thickness were constant, higher initial CO2 filling pressure resulted in a shorter pressure rise time and a greater final pressure increase. When CO2 filling pressure and energy release plate thickness were constant, faster activator combustion led to a shorter pressure rise time, while greater activator quality produced a higher pressure increase. Increasing energy release plate thickness resulted in higher peak pressure and longer pressure rise time. The destruction pressure of the energy release plate calculated using the punching shear method exhibited the smallest error relative to the experimental results, at 2.11%, whereas the welding flat cover method produced the largest error, at -190.12%, and the tensile failure methods yielded errors of 18.87% and -19.75%, respectively. These discrepancies primarily resulted from differences in the applicable conditions of the calculation methods. The dominant failure mode of the energy release plate was punching shear, indicating that the punching shear method better represented actual conditions. Comparison of the experimental results of this study with existing literature demonstrated that, within the commonly used range of energy release plate thicknesses (1.9～6.0 mm), the CO2 filling pressure ranged from 7.5 to 14.4 MPa, with an average value of 10.18 MPa, while the Kc values mainly ranged from 150.0 to 258.0, with an average value of 203.5.ConclusionsThe results indicate that the dynamic pressure response curve within the fracturing tube primarily consists of an exponential ascent stage before peak pressure and a rapid release stage after peak pressure. The duration of pressure increase in the fracturing tube is jointly influenced by the CO2 filling pressure, activator quality, and energy release plate thickness. The pressure rising capacity is jointly controlled by the activator quality and the CO2 filling pressure, whereas the ultimate peak pressure is predominantly governed by the thickness of the energy release plate. The maximum pressure during fracturing tube excitation can be characterized by the failure pressure of the energy release sheet calculated using the punching shear failure formula. The initial CO2 filling pressure in the fracturing tube should not be less than 7.5 MPa, and the activator consumption coefficient Kc is recommended to be 200. These findings provide a theoretical basis for parameter optimization of supercritical CO2 phase transition fracturing technology and for the evaluation of rock-breaking effectiveness.关键词：supercritical CO2;rock blasting;fracturing pipe;pressure release characteristics261|328|0更新时间：2026-01-29
- “在页岩储层断层研究领域，专家建立了三维精细化重构方法，深入探讨了川南页岩气储层断层几何特征，为页岩气开发提供新认识。”摘要：ObjectiveThe three-dimensional refinement and characterization of faults in shale reservoirs are closely associated with the investigation of shale reservoir accumulation and preservation conditions and play a critical role in revealing the mechanisms governing engineering casing deformation.MethodsThis study presented a method for precise fault geometric characterization through fine-scale reconstruction and analysis of fault data points, with the following key components: 1) fault plane equation fitting using singular value decomposition (SVD), which was evaluated by the R-squared coefficient of determination; 2) geometric parameter calculations, including fault dip angle, dip direction, and relief degree (RDFS), which was defined as the average distance from scattered points to the fitted plane; 3) boundary determination through the Graham scan and Alpha Shape algorithms for convex and non-convex boundaries, respectively, with the Alpha Shape algorithm being recommended for more precise boundary delineation; 4) dimensional characterization using minimum bounding rectangles to quantify fault length, width, and height. The integrated methodology enabled comprehensive geometric analysis of 3D fault surfaces and provided robust technical support for shale gas exploration and development in the southern Sichuan Basin.Results and discussionsIn the shale gas development area of the southern Sichuan Basin, a total of 86 faults were identified and were primarily categorized into two dominant groups, with one set dipping southeast (SE) and the other northwest (NW). These faults predominantly exhibited moderate-to-low dip angles, while high-angle faults were relatively scarce. Four major faults adjacent to shale gas production zones demonstrated distinct geometric evolution, as they initiated from gently dipping detachment layers at the shale base, traversed through reservoir formations with progressively increasing dip angles, and ultimately developed steeply dipping segments within the Permian Maokou Formation. The geometric characteristics revealed an inverse correlation between fault scale and dip angle. Larger-scale faults typically displayed gentler dips with lower angles and greater curvature, whereas smaller-scale faults tended to exhibit steeper dips with higher angles and smoother geometries. Regional fault lengths reached up to 42.84 km, although most measured less than 10 km. Both fault length and fault area distributions followed power-law patterns with long-tail characteristics. Vertically, faults generally extended approximately 2 km, while horizontal spans ranged from a typical 2 km to a maximum of 5 km. This distribution pattern reflected the self-similarity and multi-scale characteristics of fault systems and indicated scale-invariant growth mechanisms during tectonic deformation. The observed geometric relationships provided critical constraints for understanding fault connectivity and fluid migration pathways in shale reservoirs.ConclusionsThe reconstruction results of the planar fault are beneficial for optimizing horizontal well technology in shale gas operations. These results also contribute to a clearer understanding of fault slip behavior and the mechanisms of crustal deformation and are significant for predicting future fault activity and its developmental trends. In addition, the findings support an improved understanding of fault development patterns across different spatial scales.关键词：southern Sichuan shale gas;fault distribution;fault reconstruction;geometric characteristics;plane faults67|240|0更新时间：2026-01-29
- “氢能作为清洁能源，对能源转型至关重要。专家提出耦合和解耦瞬态仿真模型，有效模拟天然气管网能量计量和气体组分追踪，提高仿真效率，为天然气管网运行提供参考。”摘要：ObjectiveConsidering the cleanliness and high efficiency of hydrogen energy, the development of hydrogen energy is of great significance for raising the energy transition. Because the gas energy of hydrogen-blended natural gas varies due to differences in gas components, the measurement approach for natural gas is gradually shifting from the traditional volume-based measurement mode to the energy-based measurement mode to ensure fair natural gas transfer. Transient operation simulation that considers gas component tracking represents an effective method for achieving virtual real-time metering of natural gas components and gas energy, particularly for stations without gas chromatographs. The transient simulation results of gas components and other operating parameters provide valuable references for pipeline scheduling and risk assessment.MethodsFirstly, a coupling method that considered the spatiotemporal evolution of state variables and gas components during the transient flow of gas networks was proposed to simulate the time-varying operating variables and mole fraction of each component of natural gas at each discretized node in the gas pipeline network. The pipeline governing equations consisted of the continuity equation, the momentum equation, the energy equation, and the convection-diffusion equation. The BWRS EoS and enthalpy equation were applied at each discretized node. Equations of variable relations were added at the connection nodes of multiple devices to ensure the united operation of the entire gas pipeline network within the model. The coupling model considered the component tracking part and the flow parameter calculation part as an integrated whole, whereas the decoupling model divided the entire process into the gas equation of state parameter calculation part, the flow parameter calculation part, and the gas component tracking part. The parameters of the gas equation of state were calculated based on the gas components at different nodes, which were given by the initial conditions at the current time step based on the parameter calculation equations. Based on the calculated parameters, the flow parameters were obtained by solving the transient simulation model, which consisted of the pipeline governing model, equations of variable relations at the connection nodes, boundary conditions, and initial conditions. Then, the flow parameters were substituted into the component tracking model to calculate the gas components, and this model consisted of the convection-diffusion equation, boundary conditions, and initial conditions. If the pre- and post-iteration component deviations at each node met the specified accuracy requirements, the iteration ended at the current time step; otherwise, the components were updated using the calculated values, which were then used as the initial conditions to recalculate the parameters of the gas equation of state, and the above process was repeated until the component deviation between two consecutive iterations satisfied the error criteria.Results and DiscussionsThe simulation results of the proposed coupling model and the software TGNET were compared, and the average absolute deviations of pressure and temperature at delivery Node 3# were 0.086 MPa and 0.45 K, respectively, while the average absolute deviations of methane and hydrogen mole fractions were both 0.098%. For the decoupling model, to obtain simulation results that were independent of the discrete grid, 30 s and 60 s were selected as temporal steps, and 1 km, 2.5 km, and 5 km were selected as spatial steps for the simulations. The simulation results were compared under different combinations of temporal and spatial steps. The average absolute deviations of pressure and temperature at Node 1# were 0.072 MPa and 0.44 K, respectively, while the average absolute deviations of methane and hydrogen mole fractions were both 0.10%. The average relative deviation between the methane mole fraction calculated by the decoupling model under different combinations of spatiotemporal step sizes and the methane mole fraction simulated by commercial software was analyzed, together with the computation time required under each combination. When the temporal step size was constant, the average relative deviation decreased as the spatial step size decreased. Similarly, when the spatial step size was constant, the average relative deviation decreased as the temporal step size decreased. Considering both the computation time of the decoupling model and the average relative deviation relative to commercial software, 2.5 km and 30 s were selected as the spatial and temporal steps to simulate subsequent transient scenarios. For a simulation duration of 1 h at this spatiotemporal step size, the computation times of the coupling model and the decoupling model were 4 383 s and 1 050 s, respectively, and the computation time of the decoupling model was reduced by 76.04%. Then, the model results under three different boundary condition combinations were investigated. Under boundary condition combination A, the average absolute deviations of pressure, temperature, methane mole fraction, and hydrogen mole fraction at delivery Node 3# were 0.078 MPa, 0.39 K, 0.21%, and 0.21%, respectively, while the corresponding values at Node 1# were 0.067 MPa, 0.43 K, 0.14%, and 0.14%. Under boundary condition combination B, the average absolute deviations of pressure, temperature, methane mole fraction, and hydrogen mole fraction at delivery Node 3# were 0.078 MPa, 0.41 K, 0.21%, and 0.21%, respectively, while the corresponding values at Node 1# were 0.068 MPa, 0.44 K, 0.13%, and 0.13%. Boundary condition combination C cannot be supported by the commercial software TGNET.ConclusionsThe proposed coupling and decoupling models demonstrate high accuracy in flow parameter calculation and gas component tracking, reaching a performance level comparable to that of commercial software. These models are not only applicable to linear natural gas pipelines but also exhibit high accuracy in simulating multi-branch natural gas pipeline systems. At a specific spatiotemporal step size, the computation time of the decoupling model is 76.04% shorter than that of the coupling model, indicating superior simulation efficiency. In addition, the decoupling model can accurately capture the spatiotemporal evolution patterns of state variables and gas components under various combinations of boundary conditions, demonstrating strong adaptability to different boundary condition scenarios.关键词：Hydrogen blended natural gas pipeline networks;transient simulation;components tracking;decoupling model;coupling model601|495|0更新时间：2026-01-29
CARBON NEUTRALITY AND CLEAN ENERGY
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0