基于计算机视觉的桥梁拉索索力测量

Cable Force Measurement of Bridge Cables Based on Computer Vision

针对传统桥梁索力测量成本高,传感器布置困难等缺点,以及在环境激励下一般的目标跟踪算法难以准确测量拉索微小振动的问题,提出一种基于相位运动放大算法和菱形搜索的亚像素模板匹配算法的索力测量方法。首先,利用智能手机采集拉索振动视频,经过相位运动放大算法处理;其次,通过菱形搜索和曲面拟合的亚像素模板匹配算法提取拉索位移时程;最后,通过傅里叶变换获取拉索的振动频率,再利用索力-频率的关系计算拉索索力。通过实验室斜拉桥模型和户外人行天桥测试验证该方法的可行性。结果表明:测得拉索的一阶、二阶振动频率与加速度响应对比误差分别在2.4%和2.7%以内。实验室拉索模型试验索力计算结果与张拉力相比表明:随着张拉力提升,抗弯刚度影响再减小,理论计算值与实际值偏差逐渐减小,该方法可以成为索力测量的一种新手段。

Aiming at the shortcomings of high cost and difficult sensor layout of traditional bridge cable force measurement,as well as the difficulty of general target tracking algorithms to accurately measure the small vibration of cables under environmental excitation,a cable force measurement method based on phase motion amplification algorithm and diamond research sub-pixel template matching algorithm is proposed.Firstly,the vibration video of the cable is collected by a smartphone,and the phase motion amplification algorithm is processed.Secondly,the cable displacement time-history is extracted by the sub-pixel template matching algorithm of diamond search and surface fitting.Finally,the vibration frequency of the cable is obtained by Fourier transform,and then the cable force is calculated by using the relationship between cable force and frequency.The method is verified by the laboratory cable-stayed bridge model and the outdoor footbridge cable test,and the results show that the errors of the first-order and second-order vibration frequency and acceleration response are within 2.4% and 2.7%,respectively.Compared with the tensile force,the calculation results of the cable force in the laboratory cable model test show that with the increase of tensile force,the influence of bending stiffness decreases,and the deviation between the theoretical calculated value and the actual value gradually decreases,which can become a new method for cable force measurement.