###
工程科学与技术:2022,54(6):135-145
←前一篇   |   后一篇→
本文二维码信息
码上扫一扫!
基于激光条带追踪的桥梁振动非接触测量方法
(1.重庆交通大学 土木工程学院,重庆 400074;2.深圳大学 土木与交通工程学院,广东 深圳 518060;3.中铁二十二局集团第五工程有限公司,重庆 400711)
Non-contact Measurement Method of Bridge Vibration Based on Laser Strip Tracking
(1.School of Civil Eng., Chongqing Jiaotong Univ., Chongqing 400074, China;2.School of Civil and Transportation Eng., Shenzhen Univ., Shenzhen 518060, China;3.The 5th Eng. Co., Ltd. of China Railway 22nd Bureau Group, Chongqing 400711, China)
摘要
图/表
参考文献
相似文献
本文已被:浏览 96次   下载 22
投稿时间:2021-09-16    修订日期:2022-04-06
中文摘要: 针对计算机视觉的结构位移测量方法存在对靶点要求高及光照依赖性强的问题,提出一种通过竖直投射激光条带于桥梁结构表面,追踪光带中心以分析桥梁振动的新方法。通过大功率线激光器投射激光线至梁底面或顶面,倾斜拍摄激光线连续影像,利用改进灰度重心法亚像素追踪图像上激光条带中心线位置的改变,几何换算得到桥梁振动位移时程曲线,进而分析获得桥梁结构的动力参数。对简支梁模型桥进行锤击强迫振动室内试验,利用SONY 4K摄像机进行影像采集,利用LVDT位移传感器和加速度传感器进行同步数据采集。分析结果表明:激光条带横截面光强近似服从高斯分布,通过改进灰度重心法可得到激光条带中心亚像素坐标,追踪连续影像条带中心位置改变,进而得到其振动位移时程曲线;以LVDT采集数据为参考,SONY摄像机所得动位移测量结果相对误差极值为3.17%;以加速度传感器为基准对模态分析结果进行比较,摄像机对前两阶频率识别的相对误差极值为1.81%,表明动位移测量与低阶模态参数识别都取得了理想的效果;不同于加速度传感器所得模态振型空间精度受限于传感器数量,本文方法所得模态振型空间精度可达像素级,且几乎无环境光照改变干扰试验结果。该方法为非接触测量识别桥梁自振特性提供了一条经济高效的思路。
Abstract:Vision-based structural displacement monitoring methods have the disadvantages of strong dependence on target and light conditions. A new method to analyze the bridge vibration by projecting a laser strip on the surface of the bridge and tracking the center of the light band was proposed to overcome the problems. Specifically, the laser line was projected to the bottom or top surface of the beam through the high-power line laser and the continuous images of the laser line were taken by the tilted camera. Through subpixel tracking of the change of the laser stripe center line position in the image, the bridge displacement time histories can be obtained by geometric conversion, and then the dynamic parameters can be obtained. Under indoor experimental conditions, the hammer impact vibration test of a simply supported beam model bridge was carried out. The Sony 4K camera was used as the image acquisition equipment, and the linear variable differential transformer (LVDT) displacement sensors and seven accelerometers were used as the traditional data acquisition equipment. The results demonstrate that the intensity of the laser stripe cross section approximately obeys the Gaussian distribution, and the sub-pixel coordinates of its center can be obtained by the improved gray center method. The time histories of the center can be acquired by tracking the change of its coordinates. Compared with the LVDT data, the relative error extreme value of dynamic displacement measurement results obtained by the Sony camera was 3.17%. The relative error extreme value of the first two frequency identification by the Sony camera was 1.81%. Both the dynamic displacement measurement and the low-order modal parameter identification had achieved ideal results. Different from the fact that the spatial accuracy of modal shapes obtained by accelerometers was limited by the number of sensors, the spatial accuracy of modal shapes obtained by this method could reach pixel level, and there was almost no disturbance caused by the interference of ambient light. The proposed method offered a high-efficiency and cost-effective way for non-contact displacement measurement and modal identification of bridges.
文章编号:202100936     中图分类号:U446.3    文献标志码:
基金项目:国家自然科学基金项目(51708068);重庆交通大学研究生科创基金项目(2021B0004)
作者简介:第一作者:唐亮(1983-),男,副教授,博士.研究方向:桥梁健康监测与损伤识别.E-mail:tangliang@cqjtu.edu.cn;通信作者:吴桐,E-mail:wutong@mails.cqjtu.edu.cn
引用文本:
唐亮,吴桐,杜鹏,刘年,周志祥,戚小磊.基于激光条带追踪的桥梁振动非接触测量方法[J].工程科学与技术,2022,54(6):135-145.
TANG Liang,WU Tong,DU Peng,LIU Nian,ZHOU Zhixiang,QI Xiaolei.Non-contact Measurement Method of Bridge Vibration Based on Laser Strip Tracking[J].Advanced Engineering Sciences,2022,54(6):135-145.