小波分析在桥上移动荷载识别中的应用

Application of wavelet analysis in identification of moving loads on a bridge

利用小波分析技术和三次样条函数对测点的位移响应进行曲线拟合,根据拟合结果微分求得速度、加速度响应。利用梁体有限元振动方程结合模态迭加法推导出的移动荷载识别公式进行桥上移动荷载识别。对接触力的识别结果,利用小波分析技术将其在时域内分解为低频、高频部分,发现对轴重和沉浮运动引起的动荷载识别较好,识别的主要误差在于点头运动引起的动荷载。进一步的分析表明,虚假点头运动是引起误差的根源。指出在识别方法中剔除虚假点头运动将是提高识别精度的有效途径。计算机仿真计算表明,该方法具有较好的识别效果。

In this paper,we use the wavelet analysis technique and cubic spline function to fit the displacement response. And we get the velocity and acceleration response based on the curvefitting result. Finally, we find the loadtime curves from the formulas of moving loads identification according to response based on the finite element vibration equtaion of bridge and the modal superposition principle. Using the wavelet analysis to decompose the identification results into low and high frequency loads in the time domain, we find that the results of axles weight and loads induced by vertical motion are well in line with real loads and errors accur mainly loads induced by pitch motion. Further study shows that the false pitch motion is the essential problem to cause errors. So we point out that eliminating the influence of the false pitch motion will be the effective way to improve the accuracy of the identification results. This is the very valuable identification method as shown by emulation calculation of computers.