最新刊期
卷 56 ， 期 1 ， 2024
- 摘要：In the 3C (computer, communications, and consumer electronics) industry, there are strict requirements for robots' safety, interaction, accuracy, and flexibility. To solve the problem of compliant interactive control with cooperative robots, the zero-force control and collision detection methods are studied in this paper. Firstly, a general inverse kinematics (Newton–MP) algorithm is established to analyze the redundant cooperative robots, in which the inverse kinematics problem is transformed into an iterative solution of the Newton–MP method. Secondly, for the zero-force control problem of cooperative robots, the friction force is considered to formulate a complete dynamic equation. Meanwhile, a complete dynamic equation is constructed based on the acceleration cubic friction model, in which the genetic algorithm is applied to identify multi-parameters of friction models. Furthermore, a collision detection method is proposed based on a One-class convolution neural network and an un-collision dataset is built to achieve the detection task. The pseudo-negative Gaussian data is incorporated into the One-class convolutional neural networks to optimize the feature space, and the binary cross-entropy loss serves as the loss function to train the network. The One-class convolutional neural network-based collision detection method has the ability to compensate the dynamic influence of model uncertainty, which solves the problem of inaccurate modeling of traditional collision detection methods. Finally, the experimental results demonstrate that the proposed Newton–MP method achieves desired performance, i.e., 0.00013 mm absolute error. In addition, compared with the ideal friction model, the velocity-fitted cubic friction model is a more preferred solution for zero force control. By analyzing the collision detection method of the external moment observer and the One-class convolution neural network, it can be proved that the One-class convolution neural network can accurately detect the abnormal collision of the cooperative robots in a model-free manner.关键词：dynamics;collaborative robot;One-class CNN;friction parameter identification
- 339
- |
- 21
- |
- 0
发布时间：2024-03-14 - 摘要：Influenced by wind, waves, and surges, offshore crane ships can suffer from serious changes in ship attitude, resulting in changes in the posture of the crane and cargo, which has potential safety hazards to cargo and personnel. The stability control of the wave compensation platform can effectively reduce the motion impact on the safety, stability and accuracy of offshore operations, which is extremely important for the precise loading of offshore equipment on floating cranes. Aiming at the difficulty in modeling and inaccurate control caused by the hysteretic nonlinearity of the compensation platform, an active wave compensation strategy is proposed based on PI modeling and backstepping sliding mode control in this work. Firstly, the hysteresis effect curve of the compensation system is obtained through experiments, the PI hysteresis model is established by analyzing the system hysteresis loop. The parameters of the PI hysteresis model are identified by the recursive least square method, which supports the formulation of the system model. In succession, the backstepping control compensation method is designed based on Lyapunov stability, and the initial control speed is accelerated by combining the sliding mode control law. Finally, the backstepping sliding control method is applied to the compensation system. The performance of the proposed method is validated by simulating the response under regular and irregular waves in MATLAB. The resulting control strategy is further applied to perform compensation movement by driving the electric cylinder (based on the motion control card Control the servo motor). The real-time data of the system movement is collected through the sensor, which is fed back to the control strategy to form a closed loop, aiming to confirm the performance of the compensation platform under the both the regular and irregular waves. The experimental results show that the established PI hysteresis model of the Stewart platform achieve desired accuracy, and the backstepping terminal sliding mode control algorithm is able to compensate the wave motion in the actual control of the Stewart platform with high-performance. Compared with PID, backstepping method, reinforcement learning and other control methods, the proposed method can provide high performance to meet the practical requirements.关键词：active wave compensation;PI hysteresis model;backstepping terminal sliding mode control;Stewart floating platform
- 246
- |
- 26
- |
- 0
发布时间：2024-03-14 - 摘要：With the continuous promotion of the 14th Five-Year Plan for offshore wind power, the demand for megawatt-level high-power offshore fans in deep far sea areas has increased. However, in the process of offshore work such as lifting and installing the fans, the continuous fluctuation impacts of complex waves on the ship lead to a significant decline in the accuracy and efficiency, and even cause significant losses to personnel safety and property. Effective wave compensation for engineering ships in the complex sea conditions can provide a stable working environment to ensure accurate and efficient tasks. Therefore, a wave compensation control strategy based on the improved sparrow search PID algorithm was proposed in this paper and applied to the Stewart compensation platform. Firstly, based on Stewart compensation platform, the kinetic and inverse kinematics models of the wave compensation system were established, and an iterative solution algorithm of the positive solution model was designed. Then, the PID (Proportion Integration Differentiation) was used for wave compensation control since it is mature and easy for hardware implement. To get the suitable three parameters of PID for superior performances, the sparrow search algorithm was used to optimize the parameters. Whereafter, the Circle Chaotic Mapping method was used to initialize the values distribution to solve the problem of uneven initialization. The Cauchy Mutation and Reverse Learning methods were used to improve the global optimization ability of the algorithm. Finally, the motion of an engineering ship in class 4-6 sea state was input to the system, and the model was built with MATLAB/Simulink and carried out on the Stewart hardware platform to verify the effects of the compensation control method. In view of the above research, this paper mainly contains the following aspects: 1) Building the simulation model of the wave compensation system. The SimMechanics tool in Simulink was used to establish a dynamic modle according to the mechanical structure and transmission mode of Stewart platform. The upper and lower layers were used as wave compensation system and ship motion simulation system respectively. The pose analysis and homogeneous coordinate transformation were carried out, and the forward and inverse kinematics models were established. In the meantime, for the continuous ship trajectory, the iterative solution algorithm of the forward solution model was designed and verified by simulation.2) Establishing the optimal control method of the wave compensation system. According to the model characteristics and control requirements of wave compensation, the PID control method was applied. The wave compensation system was controlled by PID based on the difference between the axis length reference displacement and the actual displacement detected by the encoder. PID parameters were optimized by the sparrow search algorithm, while corresponding improvements were made to address the shortcomings of the basic sparrow search algorithm. 3) Simulation verifications based on wave compensation system. The AQWA software was used to generate ship motion data with different wave heights and periods. The ablation experiments of the improved sparrow search algorithm were carried out in MATLAB/Simulink software under the sea state of 4-6 PM spectrum and the gravity waves under the sea state of 6 at 90°and 180°wave direction angles. It was verified that Circle Cauchy Reverse Sparrow Search Algorithm (CCRSSA) had better effect on PID control parameter optimization. Then comparing with Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and other algorithms, the results showed that the improved sparrow search algorithm had good optimization control effect and convergence speed under different sea conditions of 4-6 levels. Finally, compared with reinforcement learning control method, it was proved that the improved PID control could achieve better compensation for ship motion and had the advantage of real-time response. 4) Test verification based on the hardware platform. The comparison test between CCRSSA and PSO was carried out with hardware equipment. The length of the electric cylinder was collected by the enconder, and then the positive solution of the collected electric cylinder length data was calculated. The compensation results of three degrees of freedom showed that the compensation control results of CCRSSA under different sea conditions had advantages, with adaptability and generalization. Finally, compensation error and efficiency in different ship motion values were calculated. Although the compensation efficiency was decreased slightly with the increase of sea state grade, it was still above 95%. It could meet the needs of high precision wave compensation and meet the requirements of actual offshore operations. This paper mainly studies the compensation control of multi-degree-of-freedom wobble caused by irregular and regular wave movement of engineering ships in deep sea areas under class 4～6 sea conditions. From the aspects of forward and inverse kinematics analysis, modeling, control and parameter optimization, some achievements have been made. However, there are still some domains that have not been explored and studied, and the studies are needed to supplement in follows: 1) In practical applications, the sea state environment is complex and changing, and the invariable compensation control system model is not suitable for the changing environment. Therefore, it will be more helpful to study adaptive models under the changeable environment and carry out the hardware online test with virtual simulation platform. 2) The model established in this paper is relatively complex and mainly used for the design of offline control schemes, which cannot apply some deep learning methods to online interactive learning. Therefore, it is urgent to study models that can be applied to deep learning and online control.关键词：Wave compensation platform;three-degree-of-freedom compensation;proportional-integral differential controller;improved sparrow search alg
- 256
- |
- 16
- |
- 0
发布时间：2024-03-14 -
Empirical Knowledge-guided Neural Network Method for Mechanical Performance Analysis of RC Columns AI导读
摘要：The mechanical performance analysis of reinforced concrete (RC) columns using only experimental or numerical methods usually faces challenges in balancing computational accuracy and efficiency, while purely data-driven methods suffered from poor interpretability and over-dependence on available data samples. To address this issue, an empirical knowledge guided neural network (KGNN)-based RC column analysis by investigating the fusion mechanism of empirical knowledge, test data and machine learning methods. A test database is firstly built based on 761 quasi-static test specimens. In succession, the influence rules of primary characteristics of RC columns on their mechanical properties are analyzed based on empirical knowledge to formulate mathematical representations. Finally, the test data and empirical knowledge were implemented into the artificial neural network to develop high performance, explainable, generalizable KGNN model with only minor training samples. The result comparisons of the proposed KGNN method and the pure data-driven neural network (BPNN) demonstrate that although the BPNN slightly over the KGNN in terms of the load-carrying capacity prediction accuracy, with mean square error and correlation coefficient of 0.070 and 0.978 comparing to 0.108 and 0.942 of the KGNN. However, the results of the BPNN are not consistent with the empirical knowledge and further causes overfitting problem since it fails to capture the relationship between the characteristics and mechanical properties of RC columns. Fortunately, the KGNN method can not only quickly and accurately provide the mechanical properties of RC columns, but also present a higher consistency with the empirical knowledge with greater reliability and practicality. Through this work, the neural network-based methods integrating experimental data and empirical knowledge are expected to provide a novel analysis approach for RC structures.关键词：RC column;physics and data driven neural network;empirical knowledge;mechanical properties;test database- 274
- |
- 24
- |
- 0
发布时间：2024-03-14 - 摘要：The knowledge-aware recommendation (KGR) domain commonly suffers from the problem of supervised signal sparsity, and contrast learning methods are increasingly studied to address this issue. However, existing contrast learning-based KGR models still have the following limitations. First, existing methods failed to suppress the interference information of unnecessary neighbouring nodes in the knowledge graph because graph convolution is used to directly aggregate all neighbouring nodes; Second, focusing only on the global information would lead to ignoring the fine-grained local features, causing over-smooth issues. In this work, a Knowledge-aware Recommender System with Cross-Views Contrastive Learning (KRSCCL) is proposed to address the aforementioned issues. In the KRSCCL, a relational graph attention network is proposed to construct a global view, including user, item and entity nodes. A lightweight graph convolutional network is designed to construct a local view, including user and item nodes, in which local features are emphasized to effectively mitigate the over-smooth problem. Finally, the contrastive learning mechanism is performed between intra- and inter-graph node pairs of the two views to fully extract KG signals and further optimize the user and item representations. Experimental results on three public datasets from different domains demonstrate that the proposed KRSCCL achieves expected performance improvement on all the three datasets over selective baselines, F1 score improvement on Movielens-1M, Last.FM and Book-crossing are 2.0%, 0.3% and 5.1%, respectively. Most importantly, the relational graph attention network can effectively exclude the noise during the feature aggregation of complex networks, the local views can optimize the generation of the node representation and alleviate the over-smooth problem.关键词：knowledge-aware recommendation;contrastive learning;relational graph attention;recommender system
- 257
- |
- 15
- |
- 0
发布时间：2024-03-14
INTELLIGENCE INTERDISCIPLINARY SCIENCE AND ENGINEERING
-
Adaptive Robust Prediction of Groundwater Level Based on Fusion Attention Mechanism LSTM Network AI导读
摘要：Groundwater level is an important factor affecting groundwater infiltration of sewage pipe network in dry weather. Accurate prediction of groundwater level can effectively improve the accuracy of groundwater infiltration estimation in dry weather, and assist in optimizing pipe network disease control and maintenance strategies. Aiming at the problems of low accuracy, low sensitivity, and weak generalization ability in the current urban complex hydrological prediction, a new robust adaptive water level prediction algorithm was proposed in this paper. First, a prior processing was carried out on the hydrological data, which solved the problems of large time span, high noise, missing and abnormal, and non-stationary data. Secondly, in view of the influence difference of input features on predictive indicators, a new spatial variable attention mechanism was proposed in model training stage, which can quickly identify key variables associated with water levels and assign different influence weights to input features. Furthermore, in view of the influence difference of various sequence lengths on the prediction effect, an adaptive temporal attention mechanism was also designed to adaptively find out the hidden state of the encoder related to the predictors of different sequence lengths, so as to capture time dependencies. On this basis, with the context vector as the input, an LSTM hydrological prediction algorithm integrating attention mechanism was proposed. Finally, the effectiveness of the proposed algorithm was verified by the hydrological data of Petrignano, Italy. The prediction performance was compared with GRU, Elman, LSTM, VA–LSTM and S–LSTM methods. The results showed that the proposed STA–LSTM network based on the fusion attention mechanism has a better prediction effect than other algorithms when faced with complex, large-scale, and noisy data, indicating the strong adaptability and robustness of the algorithm. The research results of the paper provide a reference for the reasonable adjustment and timely control of municipal drainage strategies.关键词：groundwater level prediction;spatial-temporal attention mechanism;LSTM;adaptive prediction;robust prediction- 304
- |
- 42
- |
- 0
发布时间：2024-03-14 - 摘要：Because the data in the big data environment presents the characteristics of dynamic updating, incremental attribute reduction is attracting increasing research attention in the field of rough set theory. As a common information system, the incomplete hybrid ordered information systems (IHOIS) is still without the study of the incremental attribute reduction. To address this issue, an incremental attribute reduction algorithm for object updates is proposed for IHOIS in this paper. Firstly, a neighborhood tolerance dominance relation is proposed to build a new neighborhood dominance rough set based on binary relation. In succession, a neighborhood dominance conditional entropy is defined and further serves as a heuristic function to design a non-incremental attribute reduction algorithm for IHOIS. Then, the neighborhood tolerance dominance relation and neighborhood dominance conditional entropy were reconstructed in the form of matrices. In response to the dynamic updates of the IHOIS, matrix-based calculation strategies are applied to study the incremental updates of neighborhood dominance conditional entropy with both the increasing and decreasing of the information system objects. Finally, the update mechanism of neighborhood dominance conditional entropy is utilized to develop the incremental update algorithms for attribute reduction for both the increasing and decreasing of the IHOIS objects. The experimental results show that compared with the non-incremental algorithm, the incremental algorithm reduces the number of attributes by 3.6% on average, improves the classification accuracy by 2.4% on average, and improves the efficiency of attribute reduction by about 10 times on average. Compared with other incremental algorithms, the proposed incremental algorithm reduces the number of attributes by 9.0% on average, improves the classification accuracy by 2.1% on average, and increases the average efficiency of attribute reduction by 94%. Based on the reported results, it can be concluded that the proposed incremental algorithms have higher performance in both performance and efficiency of the attribute reduction task.关键词：ordered information system;incomplete hybrid;dominance rough set;attribute reduction;incremental;conditional entropy
- 246
- |
- 17
- |
- 0
发布时间：2024-03-14 - 摘要：The event coreference resolution (ECR) is mainly to determine whether different event mentions refer to the same event. ECR not only effectively alleviates the problem of information redundancy in event extraction tasks, but also provides an effective way for event completion. Although many scholars have conducted extensive research on ECR using deep learning methods and achieved significant achievements, there are still issues in most ECR models, such as insufficient explicit information representation, noise introduced by arguments, and sparse distribution of coreference events. Aiming at the above problems, an end-to-end ECR method using explicit argument information and event chain reconstruction was proposed. First, an event extraction model called OneIE was used to extract event triggers and arguments. Then, a Transformer encoder is used to express the context of the event mentions, and the confidence score was introduced into the argument information coding to mitigate the error transmission. Meanwhile, the information of the argument in the horizontal and vertical directions of the trigger was decomposed by the gating mechanism, and the noise of the argument was filtered by fusing the information of the directions according to the correlation coefficient of the argument and the trigger. Afterwards, the coreference score of the event pairs was calculated by the feed forward network. Finally, to verify the validity of the event mentions, the event chains were reconstructed to correct the deviation of the model caused by the sparse event coreference. In order to verify the effectiveness of our method, the proposed model is trained and tested on the public dataset ACE2005. The experimental results showed that our model in end-to-end ECR task is 5.67% and 6.24% higher than the other models in the scores of CoNLL and AVG on average.关键词：event coreference resolution;NLP;pre-trained language model
- 203
- |
- 2
- |
- 0
发布时间：2024-03-14
ARTIFICIAL INTELLIGENCE
- 摘要：With the rapid development of IoT technology and artificial intelligence technology, vehicle edge computing has attracted more and more attention. Effectively utilizing the various communication, computational and caching resources in the vicinity of vehicles, and employing edge computing system models to migrate computational tasks closer to the vehicles, have become a hotspot in current Internet of Vehicles research. Due to the limited computational resources of in-vehicle devices, the computational demands of vehicle users cannot be met without making full use of the computational resources available in the vicinity of vehicles. Aiming to minimize the computational latency of vehicular tasks, a collaborative offloading mechanism for computational tasks in vehicle edge computing was investigated in this paper. Firstly, a three-layer architecture for task collaborative offloading was designed considering the computational resources of parked vehicles in the vicinity of vehicles as well as the computational resources of roadside units, which was comprised with three tiers: cloud server layer, roadside unit collaboration cluster layer, and the parked vehicle collaboration cluster layer. By means of collaborative offloading between the roadside unit collaboration cluster and the parked vehicle collaboration cluster, the free computational resources of system were fully leveraged, which further enhanced resource utilization. Then, in order to segment roadside units into collaboration clusters, a roadside unit collaboration cluster partitioning algorithm based on k-means clustering algorithm was proposed. A distributed iterative optimization approach with block-coordinate upper-bound minimization was utilized to design a task collaborative offloading algorithm for offloading the computation of terminal vehicle users' tasks. Finally, by comparing with other algorithm schemes through experiments, the algorithm proposed in this paper has better performance in terms of system latency and system throughput according to the stimulation result. Specifically, the system latency was reduced by 23% and the system throughput was increased by 28%.关键词：edge computing;vehicle edge computing;cooperative offloading;k-means;compute offloading
- 71
- |
- 16
- |
- 0
发布时间：2024-03-14 - 摘要：In applications of harsh outdoor environments, unmanned aerial vehicles (UAVs), known for their flexibility and convenience, were utilized to assist in carrying user tasks to edge servers through wireless data transmission. However, it was found that UAV flight platforms struggled to provide long-duration task offloading services, significantly limiting their application prospects. This study investigated how to effectively integrate UAV task offloading and charging scheduling in a mobile edge computing environment. Firstly, a new application model was constructed, which cohesively managed UAV task offloading scheduling and its own charging needs, incorporating several wireless charging platforms into the UAV-assisted task offloading application scenario. These platforms enabled UAVs to autonomously recharge during task execution, providing automated charging services without the need for human intervention. UAVs independently decided whether to proceed to the nearest charging node for power replenishment based on their current power level and upcoming task offloading plans. However, opting to recharge at a charging station not only incurred additional time and energy consumption from cruising altitude to the charging station but also required consideration of the time cost during the charging process and its impact on overall task scheduling. When UAVs decided to recharge, additional time and effort were needed to descend from cruising altitude to the charging node. Secondly, the value of user tasks and UAV charging needs were considered to optimize the benefits of UAV-assisted user device task offloading under conditions sensitive to delay and energy constraints. This involved not only optimizing the UAV’s flight path and task allocation but also its charging schedule, ensuring sufficient charging and efficient operation while executing tasks. Such a cooperative scheduling strategy enabled UAVs to maximize the processing of user tasks while maintaining necessary operational energy, thereby enhancing the performance of the entire mobile edge computing system. Finally, a deep reinforcement learning algorithm was employed, and the deep Q network (DQN) was fine-tuned to form the Fixed DQN algorithm, effectively addressing the large-scale state-action search space issue within the model. This approach capably handled complex decision-making problems and facilitated effective learning and optimization across a wide state space. With the deep learning framework, the algorithm processed high-dimensional input data and made accurate offloading and charging decisions in various dynamic environments. This was significantly important for improving the efficiency and effectiveness of UAV task offloading and charging scheduling. The design of the algorithm comprehensively considered the following key aspects: Initially, the state space and action space of the algorithm were defined, ensuring that the agent could accurately perceive the environment and make effective decisions. Subsequently, the composition of the reward function was detailed, guiding the agent to progress towards the desired goal during training. Solely using the maximization of task offloading benefits as a constraint was found to prevent the agent from meeting the condition of serving each user at least once. Therefore, a method of minor learning goal constraints was proposed in the study. Specifically, the task offloading rewards accumulated by the agent in the phase of not completing minor learning goals were not directly awarded to prevent deviation from the path to achieving these goals. Afterwards, an experience replay mechanism was introduced, which improved learning efficiency and reduced correlations between samples by storing and reusing past experiences. Additionally, two asynchronously updated neural networks were employed to stabilize the learning process. Based on this, the hyperparameters of the Fixed DQN algorithm were meticulously optimized to further enhance the algorithm’s performance. Most current research was based on the assumption that UAVs possess certain task processing capabilities. However, a different assumption was adopted in the paper, where the primary role of UAVs was only to carry tasks, not directly participate in task processing. Additionally, the autonomous charging needs of UAVs were also considered. This assumption is closer to actual application scenarios, where UAVs are primarily used for data collection and transmission, rather than data processing. The limitations of UAV endurance and the need for charging during task execution were also taken into account. In the study, 11 nodes were set up within a circular area with a radius of 3000 meters as a test environment to verify the feasibility of the Fixed DQN algorithm. To comprehensively evaluate the performance of the proposed Fixed DQN algorithm, extensive experiments were subsequently conducted under various conditions, including different numbers of user nodes, charging nodes, and varying lengths of service time. For comparative analysis, the experiments also included comparisons with ant colony algorithms, genetic algorithms, and DQN algorithms. In this way, the effectiveness of the Fixed DQN algorithm in different scenarios, especially in complex and dynamically changing environments, were deeply explored. The experimental results showed that under all test conditions, the Fixed DQN algorithm significantly outperforms the ant colony algorithm, genetic algorithm, and DQN algorithm, particularly in scenarios with an increased number of nodes and extended service times. Furthermore, the performance improvement of Fixed DQN over DQN highlights the effectiveness of deep reinforcement learning in parameter tuning. These findings confirms the efficiency of the Fixed DQN algorithm and the importance of parameter tuning strategies in addressing UAV task offloading and charging scheduling issues.关键词：edge computing;UAV;task offloading;reinforcement learning;charging scheduling
- 33
- |
- 5
- |
- 0
发布时间：2024-03-14 - 摘要：The widespread adoption of wireless networks has led to a significant increase in the deployment of Access Points (APs) for Received Signal Strength (RSS) fingerprint localization. This surge has introduced redundancy, negatively impacting localization and increasing computational costs. While traditional AP filtering mitigates redundancy to some extent, an Effective Access Point Set Construction method (EID) based on information distinctiveness was proposed in this paper. Firstly, EID was utilized to evaluate APs by assessing their spatial resolution through information distinctiveness. Moreover, an incremental clustering algorithm was designed, which was able to form sets of different categories according to localization abilities of APs. Finally, an AP effective set selection strategy based on the maximum point set distance principle was proposed in the paper, resulting in suitable AP sets. Extensive experiments validated the performance of the proposed EID in real-world scenarios. EID was also compared with existing AP selection methods including the Group Discrimination-Based (GDB) algorithm, Software Defined Network (SDN) algorithm, and Nonuniform Quantization RSSI Entropy (NQRE) algorithm. Experimental results showed that EID demonstrates a significant improvement in localization accuracy by 18.7%, 11.2%, and 14.6% with enhanced stability, achieving a localization error below 1.2 m in 95% of cases, even with a 40% reduction in AP quantity.关键词：AP selection;information discrimination;effective set;indoor localization
- 17
- |
- 2
- |
- 0
发布时间：2024-03-14
INTERNET OF THINGS
- 摘要：In September 5th 2022, an earthquake measured Ms 6.8 occurred in Hailuogou scenic spot, Luding County, Sichuan Province, and its intensity at epicenter is Ⅸ. Compared with Wenchuan earthquake (Ms 8.0) in 2008 and Lushan earthquake (Ms 7.0) in 2013, its magnitude is smaller. However due to the mountainous topography, the earthquake-triggered geological disasters are highly developed, obstructed traffic lines and caused 118 persons dead or missing. In order to investigate the earthquake-triggered disasters and loss, a 5-days field visit is taken to the affected area of Luding County and Shimian County. Based on the collected data and regional geological information, the analysis is conducted to formulate the development of geological disaster, the damage of civil structures and human fatalities. The major findings can be summarized as follow. 1) The hanging-wall effects of geological disasters are not obvious, which mainly occur in the areas with earthquake intensity of more than Ⅶ, and are small in volume and can be categorized mainly as rock fall and gravel/sand/debris slide. The rock planar slide and soil slump can be occasionally witnessed. 2) The landslide types are controlled by geological conditions. The upstream Detuo town in Luding County belongs to the Dadu river valley which is hot and arid, where the slope mass is highly loosened but slightly weathered, and the disaster type is mainly rock fall. However, downstream of the town, the valley is humid, the slope mass is both highly loosened and weathered, and gravel/sand/debris slide is the major failure type. 3) Geological disasters appear mainly in the areas of mountain ridge, intersection of gentle and steep slope, the walls of an erosion gully and highway lines. Engineering disturbance is an enhancing factor for the dense distribution of disasters and human fatalities. 4) Damage of civil structures mainly occurred in the areas with intensity of Ⅸ and is most serious in Moxi town. The construction quality of rural self-built houses and old buildings is the main problem affecting the anti-seismic capacity of the structures. The seismic site effect should be considered during post-quake reconstruction. 5) In the region of alpine and deep valley, it is effective to reinforce highway slope and to keep a building site away from the affected zone of possible rock falls, thus to reduce disasters and human fatality.关键词：Luding earthquake;geological disaster;structure damage;geological condition;site effect;control measure
- 203
- |
- 3
- |
- 0
发布时间：2024-03-14 -
A Method for Estimating Human Loss of Life in Dam-Breach Flood Considering Human-Fluid Interaction AI导读
摘要：Dam-breach floods pose a significant threat to human life downstream. Conducting a quantitative risk assessment of life loss caused by dam-breach has important practical implications for emergency response and disaster relief efforts. This study presents the HURAM2.0 model, an extension of the existing HURAM1.0 model of life loss Bayesian network, which considers the interaction between the human body and water flow. It first assesses the stability of individuals in floodwaters and then determines the risk of drowning. By employing Monte Carlo simulation, the model integrates the effects of water depth and flow velocity on life loss. The HURAM2.0 model is applied to analyze the life loss caused by the breach of the Tangjiashan landslide dam. The main conclusions are as follows: The HURAM2.0 model established a quantitative relationship between flow velocity and loss of life, thereby providing a more accurate description of the stability and survival capabilities of the human body in water flow. In comparison to the HURAM1.0 model, the HURAM2.0 model demonstrated greater accuracy in predicting the loss of life under conditions of strong flood intensity. Additionally, within the HURAM2.0 model, apart from water depth and flood severity which remain relatively consistent, the sensitivity of the other variables increases. Specifically, the number of floors in residential buildings, shelter provided by buildings, and duration of the breach variables were found to exhibit a respective increase in their impact on model calculations by 142%, 95%, and 93%. This enhancement strengthens the model's interpretability under low, medium, and high flood intensity, providing a favorable position for Bayesian inversion. In the risk analysis of the Tangjiashan landslide dam, the HURAM2.0 model differentiates the loss of life under different flow velocity conditions, which aligns more closely with reality. The risk and mortality rate were high before the spillway was excavated. The risk was significantly reduced and the number of deaths greatly decreased after on-site surveys and spillway excavation. It is recommended to use early warning systems and evacuation measures to mitigate the risk of life loss.关键词：dam breach;physical modeling of human stability;risk of loss of life;Bayesian network;landslide dam breach floods- 201
- |
- 1
- |
- 0
发布时间：2024-03-14 - 摘要：According to the characteristics of progressive landslide creep, the existing calculation model of anti-skid pile gives less consideration to the interaction process of progressive landslide pile-anchor-soil. In response to this issue, the deformation characteristics of progressive landslides and the dynamic working process of prestressed anchor anti-skid piles from no-load to full load are combined. Using Winkler’s elastic foundation beam theory, the anchor cable resistance pile-anchor deformation coordination conditions were optimized. Considering the gradual change of soil pressure behind the piles, a calculation model for pile anchor soil collaborative work under different working conditions was constructed. The calculation method for the initial tension value during the construction stage of prestressed anchor anti-skid piles was derived. Combined with engineering examples and numerical simulation results of finite element software, to obtain high-precision calculation values, MATLAB software was used to compile programs for calculation analysis. The results show that: ① The stage deformation characteristics of "stability-peristalsis-instability" of progressive landslide determine the three-stage process of "no-load-load-full load" dynamic internal force deformation process. The possible soil pressure model after each stage pile is specified, the design of the full load stage does not conform to the actual mechanism of progressive landslide. This further proves the importance of the prestressed tension value design of the anchor cable in the construction stage; ② Progressive landslide thrust from the initial no-load stage which has not been generated to the full load stage. Both phases are not independent of each other, the two interact through pile-anchor deformation, which shows that the prestress tension value of the anti-skid pile in the construction stage is affected by the internal force and deformation of pile anchor in the full load stage; ③ According to the dynamic internal force deformation process of the progressive landslide anchor cable resistance pile, the internal force balance of the bearing section and the anchorage section of the anti-skid pile is ensured to be equal, the design principle of full pile bending moment balance in full load stage is optimized, and the static soil pressure as the design standard of the prestressed initial tension value in the construction stage is established. Comparison shows that this calculation method outperforms the remaining four existing algorithms, at the same time, the finite element numerical simulation results and the proposed algorithm fit is high. The rationality and reliability of this calculation method are further verified. The research results can provide technical guidance for the design, calculation and construction of prestressed anchor cable anti-skid pile in progressive landslide.关键词：progressive landslide;anchor cable prestress;coordinated deformation of pile anchor;soil pressure state;anchor cable design tension
- 208
- |
- 4
- |
- 0
发布时间：2024-03-14 - 摘要：Susceptibility zoning is the basic step of regional geological hazard risk assessment, and a reasonable classification method for the susceptibility levels is of significance for obtaining effective regional landslide susceptibility maps. However, few studies have compared the advantages and disadvantages of the susceptibility classification method, especially the failure to link landslide with the predicted susceptibility index. Yanchang County in Shaanxi Province was taken as the study area, and three machine learning models were applied to calculate the landslide susceptibility index of the region, which were classification and regression tree (C&RT), random forest (RF), and radial basis function (RBF) models. Then four GIS-based classification methods (natural breaks, equal interval, quantile, geometrical interval) were used to classify susceptibility levels and to generate landslide susceptibility maps. Aiming at the nonlinearity correlation between landslide inventory distribution and susceptibility index was not considered in the classification of susceptibility, a frequency ratio threshold method was proposed to classify susceptibility levels innovatively. The results showed that although the accuracies of the three models expressed by the receiver operating characteristic (ROC) curve were all greater than 0.75, large differences in landslide distribution patterns among different landslide susceptibility maps were observed. Different classification methods have a comparative analysis role in the final landslide susceptibility mapping. The landslide susceptibility maps using geometrical interval and quantile methods identified more landslides in very high susceptibility areas, while the total area of very high and high susceptibility areas was too high. Moreover, the density of landslides from equal interval and frequency ratio threshold methods was larger. That means the landslides identified are more concentrated. This paper innovatively proposed the frequency ratio threshold method for landslide susceptibility classification, which could provide a new idea for accurate zoning of susceptibility, provide scientific reference for project site selection and land use planning in areas with poor slope stability, and improve the ability of geological safety assessment and emergency management.关键词：landslide susceptibility;classification method of susceptibility levels;frequency ratio threshold method;machine learning
- 235
- |
- 4
- |
- 0
发布时间：2024-03-14 - 摘要：The coupling of loose deposits and slurry is easy to trigger geological disasters. The shear strength deterioration of fine components under high water content is one of the most important factors. Based on the ring-shear test, the experimental study on fine components that are taken from a loose deposits in Three Gorges Reservoir area is carried out in the water content range of solid-liquid phase change. The following conclusions can be obtained: The coarse and fine component binary characteristics is significant. The saturated water content ωsr of fine component in undisturbed condition is 23.6%, ωL and ωP are 27.2 and 18.1 respectively, and the tests conducted in range of ω=23.6%~29.0% could cover the whole phase transformation process; Ring-shear test could be successfully carried out in water content conditions of 25.0%, 26.0% and 27.0%, which are all in soft plastic status. When ω=25.0%, the sample just enters soft plastic state. Compared with ωsr=23.6%, the peak and residual internal friction angle are reduced by 23.5% and 18.6% respectively, while cohesion is reduced by more than 80%, which can be regarded as basic loss; When ω exceeds ωL and achieves 28.0% and 29.0%, ring-shear test can only be carried out at σ=0 kPa, and the shear strength is extremely low, while the physical meaning of this shear strength is closer to viscosity of very viscous mud; Referencing to the water content range involved in this study, cohesion c is basically lost when samples enters soft plastic state, while internal friction angle is characterized by gradual decrease with the increase of ω in soft plastic status, but suddenly loses when ω achieves to ωL. The deterioration process of both shear strength parameters are asynchronous. The stress analysis and generalized calculation show that before phase transition, rapidly increase of K0 for samples in soft plastic status effectively alleviates the significant drop of
$\varphi $ 关键词：loose deposits;fines component;solid-liquid phase change;ring shear test;shear strength characteristics- 192
- |
- 2
- |
- 0
发布时间：2024-03-14 - 摘要：How to select non-landslide samples for landslide susceptibility prediction (LSP) modeling is an important uncertainty affecting the LSP results. To study the influence of different non-landslide sample selection methods on LSP modeling, five sampling methods were proposed (Randomly selected from the whole area, from the specific attribute area with a slope lower than 5°, from the area outside buffer zone which is 300 m from each landslide, selected by information value method, selected by Semi-supervised machine learning) with the same number of landslide grid units, and coupled with Random Forest (RF) to construct random selection-RF, low-slope RF, buffer-based RF, IV-RF, and semi-supervised RF models for LSP. Taking Nankang County of Jiangxi province as the study area, a total of 19 environmental factors such as elevation, slope, population density, and road density were acquired, and 233 landslide inventories were obtained. The landslide inventory was divided into 2598 grids as landslide samples to construct the input-output of the above-coupled model. Then, the prediction accuracy and the distribution characteristics of predicted landslide susceptibility indexes were used to analyze the LSP modeling uncertainty. To further solve the problem of unreasonable distribution of landslide susceptibility indexes predicted by the coupled model, a sample set with a 1∶2 ratio of landslide to non-landslide was used for LSP, and the condition of the sample set with equal proportion was compared in semi-supervised RF. Results showed that: 1) The prediction accuracy of models such as low-slope RF, buffer-based RF, IV-RF, and semi-supervised RF was substantially better than that of the random selection-RF model, suggesting that accurate selection of non-landslide samples was critical for LSP. 2) The modeling performance of the semi-supervised RF was optimal, which predicted the distribution characteristics of landslide susceptibility indexes more accurately and reliably at landslide∶non-landslide = 1∶2 than at 1∶1. It is necessary to explore the ratio of landslide to non-landslide samples in depth in future studies.关键词：landslide susceptibility prediction;non-landslide samples selection;semi-supervised machine learning;information value;random forest
- 200
- |
- 4
- |
- 0
发布时间：2024-03-14
MECHANISM OF LANDSLIDE–DAMMED LAKE AND ITS CONTROL
- 摘要：In the background of China's dual goals of achieving carbon neutrality and fundamental environmental improvement, Chongqing, as an important strategic pivot point in the development of western China, has the responsibility and obligation to take the lead in reducing pollution and carbon, and help build a national ecological civilization. The emission processes of CO2 and air pollutants are complex and involve many elements, thus it is necessary to clarify the key factors influencing the emission of both, which can provide scientific reference for the realization of synergistic control of pollution reduction and carbon reduction in Chongqing. The extended LMDI–CO2 model and LMDI-AP model based on the log-mean division index (LMDI) model were constructed, and the CO2 emissions and industrial SO2 emissions in Chongqing from 2011 to 2020 were used as objects for empirical analysis. And then the effects of factors on the changes in CO2 and industrial SO2 emissions in Chongqing were evaluated quantitatively. Meanwhile, the influence mechanisms of the factors were unearthed according to the historical development of Chongqing’s economy, society, and energy environment, and the key factors of synergistic emission reduction are summarized to explore the synergistic control mechanism of pollution reduction and carbon reduction in Chongqing. The results showed that from 2011 to 2020, the economic development effect was the primary factor driving the increase of CO2 emissions in Chongqing, with a contribution rate of 245.59%. The energy intensity effect contributed the most to CO2 emission reduction, with a contribution rate of –210.51%. In addition, the optimization of industrial structure also contributed to CO2 emission reduction. With the development of the economy, the expansion of population, and the improvement of living standards, the pressure of CO2 and pollutant emission reduction in Chongqing will increase, and the “high carbonization” energy structure will restrict both emission reduction. Improving energy efficiency and optimizing industrial structure played a key role in achieving “double reduction”. Therefore, energy intensity, industrial structure, and energy structure were the three key factors affecting synergistic emission reduction. Reducing energy intensity, upgrading industrial structure, and optimizing energy structure will be important ways to achieve synergistic control of pollution reduction and carbon reduction in Chongqing in the future.关键词：carbon dioxide;air pollution;factors;cooperative;LMDI
- 242
- |
- 58
- |
- 0
发布时间：2024-03-14 - 摘要：To evaluate the CO2 leakage risk through wellbores in CO2 geological storage sites, a self-developed numerical simulation code—WellRisk is introduced. The WellRisk could quantitatively evaluate the CO2 wellbore leakage risk, and it was applied to an actual CO2 geological storage site—Shenhua Ordos saline aquifer CO2 storage demonstration project. The leakage of CO2 along the injection wells and monitoring wells at the site was quantitatively evaluated by WellRisk, and the results were compared against the simulation results of NRAP–IAM–CS (The National Risk Assessment Partnership–Integrated Assessment Model–Carbon Storage) developed by the US Department of Energy, demonstrating a reliable quantitative assessment of CO2 leakage risk by WellRisk. This paper defines the wellbore leakage coefficient (the ratio of the effective sectional area of the wellbore where leakage occurs to the total sectional area of the wellbore) and takes the wellbore leakage coefficient as an important parameter to quantitatively characterize the quality of wellbore cementing. When the leakage coefficient is 10–6, the total CO2 leakage of the Shenhua storage site in 1 000 years is 720 tons, accounting for 0.24% of the total injection, which is less than the IPCC CO2 leakage risk threshold of 1%. Therefore, wellbores with leakage coefficients of 10–6 or lower are low-risk leakage wells with good cementing, corresponding to wellbore permeability below 10–12 m2 in the NRAP–IAM–CS software. Wellbores with leakage coefficients of 10–5 or higher are high-risk leakage wells with poor cementing. The results of WellRisk and NRAP–IAM–CS both show that when the quality of wellbore cementing in injection wells and monitoring wells is good, there is almost no risk of CO2 leakage at the Shenhua CO2 storage site.关键词：geologic CO2 storage (GCS);CO2 leakage;wellbore;risk assessment;cementing
- 193
- |
- 2
- |
- 0
发布时间：2024-03-14 - 摘要：Objective: Accelerating the construction of modern energy systems plays an important role in building energy power, guaranteeing national energy security, helping to realize the goal of carbon peaking and carbon neutrality, and supporting the high-quality development of the economy and society. However, there is no quantifiable methodology for assessing the effectiveness of the construction of modern energy systems. Carrying out quantitative research on the construction process of modern energy systems is of great significance for understanding the construction process of China’s modern energy systems, providing quantitative guidance for the construction of the country’s modern energy systems, and helping China’s energy revolution, “dual-carbon” strategy, and even the construction of new energy systems. Methods: The study adopts a systematic analysis method to construct a modern energy systems evaluation index system with four dimensions, namely, clean, low-carbon, safe, and efficient, quantitatively assesses the historical progress of China’s modern energy systems construction and conducts comparative analyses with 76 countries around the world, to locate the level of China’s modern energy systems construction from different perspectives. Results and Discussion: In the past ten years, China’s modern energy comprehensive index has shown a continuous upward trend, with an average annual growth rate of 2.41%, and reached the world average level in 2020, ranking 38th among 76 major economies in the world, indicating that China’s efforts to promote the energy revolution have been effective, and the level of construction of the modern energy systems has been steadily improved. Further analysis shows that China’s energy structure indicators, carbon emission series indicators, and energy intensity indicators are key indicators affecting the level of China’s comprehensive index. China’s efficiency index ranking (42nd) is significantly behind developed economies, from the point of view of specific high efficiency indicators, energy intensity is still high, mainly because the proportion of high energy-consuming industries is still too high, industrial structure adjustment is not yet in place, the level of technology is not enough energy-saving and efficient, coupled with a lot of waste, etc., indicating that there is a huge space for China’s energy efficiency to improve, and on the other hand, it also indicates that more attention should be paid to scientific and technological inputs; China’s safety index ranking (16th) is relatively advanced, proving that China’s coal in the energy security of the “ballast” position, but for the same reason led to China’s low carbon index is relatively backward (56th); Over the past decade, China has achieved remarkable results in the clean utilization of energy, but the overall ranking is low (48th). The clean and efficient utilization of coal still needs to be strengthened, the realization of the clean and efficient utilization of coal is energy clean, especially ultra-low emissions and economically viable carbon capture, utilization, and storage, which is the endpoint of the clean and efficient utilization of coal, China’s clean development and utilization of energy is still a long way to go. Conclusions: In general, in the construction of modern energy systems, the gap between China and developed countries is gradually narrowing, but there are still some problems in the construction of China’s modern energy systems. Therefore, the future to accelerate the construction of a modern energy system based on renewable energy, supplemented by fossil energy; Secondly, new energy and low-carbon technology to reach the volume is still time-consuming, for a long time, coal to support China’s economic and social development faster, is China’s energy security and guarantee the ballast and stabilizers, to energy security is always in the first place. Clean and efficient use of coal still needs to be strengthened, and the realization of clean and efficient use of coal is energy clean; finally, the key to energy transformation is to optimize the energy structure, to multi-energy complementation, integration of heat, electricity, gas, cold, and enhance the flexibility of thermal power units; in the end-consumption aspects of energy efficiency, reduce energy intensity, the implementation of electric energy substitution, and the formation of an energy consumption pattern centered on electricity. In the face of changes in the global energy pattern, our country to ensure energy security and economic and social development needs as the basic premise, put forward “accelerate the planning and construction of a new energy system”, which is a new concept and new judgment of our energy industry. The connotation of the new energy system has changed according to the requirements of the new era and new development.关键词：modern energy system index;clean;low-carbon;safe;efficient
- 212
- |
- 2
- |
- 0
发布时间：2024-03-14
CARBON NEUTRALITY AND GREEN ENERGY
- 摘要：Flow measurement is an important means for irrigation, efficient allocation of water resources and water diversion management. In the open channel, the overfall is usually applied as a facility connecting the upstream and downstream. When the overfall is not affected by the flow of downstream, the overfall is classified as a free one, which can be used as a discharge measurement by a single measurement of depth at the end of the channel. At this condition, the flow pattern changes from a subcritical flow into a supercritical one, and the pressure distribution in the end section deviates greatly from the hydrostatic pressure distribution with a skewed shape, which is the key to derive the discharge formula by energy equation and momentum equation. In this study, the characteristics of sectional pressure distribution of free overfall in a rectangular open channel were studied by model test and numerical calculation, and then the water depth-flow relationship formula of open channel with free overfall was established. It was found that the pressure of the end section was conformed to the self-similar distribution law, as the distribution of dimensionless pressure values tended to be the same curve, in which the maximum value was 0.238he (here he was water depth of end section) and it occurred at 0.209he distancing from the end fall. The above findings are independent of the flow discharge, bottom slope and side wall roughness. Based on these findings, a simple formula combining a power function and wake function was proposed to describe the pressure distribution at the end section. Then using the momentum equation and continuity equation, a quasi-theoretical end-depth–discharge relationship was proposed. With the known end depth, bed slope and Manning’s coefficient, the discharge could be direct solved by the proposed equations. The proposed equation was valid over the entire practical ranges of Q=5～100 L/s, S=–0.011 2～0.053 4, n=0.093 0～0.019 3, and they showed excellent agreement with the available experimental data in this paper and from others, with the precision of ±5%. The proposed formula for discharge is a useful computational tool for estimation of discharge in rough bed rectangular open channels. The results of this paper provide theoretical and technical supports for the open channel discharge measurement using free over flow, which is important to the application of free overfall as a discharge measure device.关键词：rectangular channel;free overfall;pressure distribution;end-depth−discharge relationship;flow measurement
- 61
- |
- 19
- |
- 0
发布时间：2024-03-14 - 摘要：The evolution process of dike-break flood in the urban neighborhood equipped with pressurized stormwater pipe networks was very complex. In general, the buildings and stormwater pipe networks in urban neighborhood has profoundly changed the evolution characteristics of flood, thus affecting the hydraulic characteristics of the drainage function of the neighborhood. Based on the generalized sink model established in typical urban neighborhood, including houses, sidewalks and stromwater pipe networks, we measured the block water depth, stromwater pipe network flow and pressure during the evolution of typical urban neighborhood for different dike-break flood, so as to explore the impact of different storage heights, forebay inlet flow on the fluctuation of neighborhood water level and the discharge flow of pressurized stormwater pipe networks. Generally speaking, the dike-break flood will form a large area of hydraulic jump under the action of the side wall of the flume and the house. Furthermore, there are thin layered sheet flows near buildings. With the process of flood evolution, the hydraulic jump area first migrates horizontally and merges at the sidewalk. Then, the hydraulic jump is gradually dissipated in the longitudinal direction of the block road until it disappears. After 100 seconds, the water level in the neighborhood is basically stable, showing a trend of "first increasing, then decreasing" from upstream to downstream. In fact, during the evolution of the neighborhood, compared with the flow of different forebay inlets, the impact of different storage heights on the discharge of the pressurized stromwater pipe network is more significant. The peak flow at the gutter inlet of the block at the maximum storage height is about 1.5 times of the maximum forebay inlet flow. In this study, on the basis of combining the water level of the block and the pressure of the pressurized stormwater pipe network, the Euler number is adopted to calibrate the discharge coefficient of the short pipe submerged outflow formula. Finally, the error of the corrected theoretical calculation value is reduced from 32% to 2%, perfectly reflecting the discharge capacity of the street gutter inlet. In general, the research results not only provide detailed verification data for mathematical models, but also provide theoretical guidance for urban flood control.关键词：dike-break flood;pressurized stormwater pipe networks;gutter inlet;generalized sink model
- 72
- |
- 21
- |
- 0
发布时间：2024-03-14 - 摘要：In alpine canyon, flash floods and debris flows are common. Usually, they are accompanied by strong gully erosion, and consequently create significant sediment accumulation in the gully mouth, which easily causes casualties and property losses in the region. It is difficult to track the movement process of gully bed sediment washed by flash floods and debris flows through field surveys, model experiments, and other traditional methods in alpine canyon areas due to high mountains and deep valleys. Therefore, it is impossible to provide information regarding the accumulation forms of sediment carried by mountain floods and debris flows as well as the distribution properties of sediment particle size in the accumulation body. To investigate the sediment accumulation morphology and particle size characteristics of mountain flood and debris flow, the generalized mountain gully area was taken as the research object. The coupled numerical methods of computational fluid dynamics (CFD) and discrete element method (DEM) were applied to simulation of sediment deposition process of mountain flood debris flow with different rheological characteristics. The effect of the volume concentration of mountain flood and debris flow on the accumulation process of sediment particles of different sizes in the accumulation area is analyzed in detail. As the volume concentration of mountain flood and debris flow increases, sediment accumulation rate first increases and then decreases. Furthermore, excessive volume concentration causes sediment to accumulate in a different form. The variations in sediment accumulation distance with time can be categorized into four stages: rapid increase, initial deceleration, recovery of growth and stable development. The variation of particle dispersion with movement time is of a power function, and the development rate is closely correlated with sediment particle size and volume concentration. There is a strong correlation between the degree of dispersion of sediment particles with time and the size and volume concentration of sediment particles. A power function with three parameters can be used to predict this process. Increased sediment particle size leads to an increase in the dispersion of sediment particles in the accumulation body. However, the increase in mountain floods and debris flows volume concentration causes the dispersion of sediment particles to increase at first and then decrease. The findings in this study can serve as a scientific basis for further understanding of mountain floods and debris flows.关键词：alpine canyon area;flash flood gully;sediment accumulation;accumulation morphology;CFD–DEM
- 92
- |
- 23
- |
- 0
发布时间：2024-03-14
HYDRAULIC ENGINEERING
- 摘要：Radio telescope is widely used in the fields of radio astronomy and navigation. The electromagnetic performance of radio telescope is heavily affected by the structural thermal deformation due to solar heating as the working frequency and aperture increases. To understand the solar thermal behaviors of the 110 m telescope to be built in Xinjiang, we established a thermal-mechanical coupling model to simulate the temperature/displacement fields at different solar times and wind speeds, and concluded the spatial and temporal characteristics. Then we evaluated the reflector precision using the best fit parabolic method and revealed the law and mechanism of antenna thermal deformation by the trend in sub-reflector’s position compensation. Results show that large wind speed leads to uniform expansion and thus high reflector accuracy when the solar heating causes thermal deformation error. The decrease in reflector accuracy is largely attributed to the non-uniformly deformation of antenna structure caused by non-uniform temperature filed. Larger temperature difference results in more non-uniformly deformation and thus lower reflector accuracy. For the antennas with the same attitude but different wind speeds, they have similar reflector error distributions that can change with the position of sunlight direct incident point. However, the magnitude of deformation varies with the sunlight intensity and speed. The reflector precision increases with the increasing wind speed and has a similar law of variation for various attitudes. The thermal error can be greatly compensated by the sub-reflector repositioning. The approach and results of this paper will provide valuable guidelines for the design and construction of large radio telescope as well as its thermal error control.关键词：radio telescope;thermal error;thermal deformation;reflector accuracy;thermal error compensation
- 30
- |
- 1
- |
- 0
发布时间：2024-03-14 - 摘要：Life stage identification accuracy of space rolling bearings is low due to the large difference of sample distribution, the small number of available training samples, and the unequal number of samples at different life stages under variable working conditions. Therefore, this paper proposes a novel unsupervised transfer learning method called class-contrast cluster-allocation heterogeneous transfer learning (CAHTL). This method firstly transfers a small number of labeled samples under historical working conditions and the unlabeled samples (i.e., the testing samples) under the current working conditions into a public feature space through heterogeneous transfer learning, so as to minimize the distribution difference between the samples under different working conditions. After that, the positive and negative samples of sample features in target domain are constructed by using cluster points in source domain to achieve the number redistribution of two domain samples, and then contrastive learning on the positive and negative samples in two domains is carried out to make a better classification characteristic on the testing samples. Then, the classification of testing samples is completed by calculating the similarities between the testing samples and the cluster points without parameter learning, which can prevent the large difference in identification accuracy of samples at different life stages in the case of unequal samples and the over fitting of CAHTL in the case of few labeled training samples. Finally, the stochastic gradient descent and momentum renewal are used to asynchronously update the CAHTL parameters to maintain the consistency of sample features and improve the convergence speed of CAHTL. CAHTL can use few and unequal training samples at known life stages under historical working conditions of space rolling bearings to identify the life stages of testing samples under the current working conditions with high accuracy. The effectiveness of the proposed CAHTL is verified by an experiment of identifying the life stage of a space rolling bearing.关键词：transfer learning;contrastive learning;momentum renewal;space rolling bearings;life stage identification
- 102
- |
- 25
- |
- 0
发布时间：2024-03-14
MECHANICAL ENGINEERING
- 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