基于最优频率边界经验小波变换的工作模态分析  

Operational Modal Analysis Based on Optimal Frequency Boundary Empirical Wavelet Transform

作　　者：王秋萍 唐剑 秦世强[1] WANG Qiu-ping;TANG Jian;QIN Shi-qiang(School of Civil Engineering and Architecture,Wuhan University of Technology,Wuhan 430070,China)

机构地区：[1]武汉理工大学土木工程与建筑学院,武汉430070

出　　处：《武汉理工大学学报》2023年第5期95-101,共7页Journal of Wuhan University of Technology

基　　金：国家自然科学基金(51608408).

摘　　要：为提升经验小波变换频谱分割精度,提出一种基于最优频率边界的频谱分割方法,并结合随机减量技术、Hilbert变换和局部最小二乘法,识别结构模态参数。该方法以傅里叶谱最大幅值处频率为中心点,建立最小带宽到最大带宽的一系列频带,并利用经验小波变换提取不同频带的子分量及对应残量,将皮尔逊相关系数的最小值对应的带宽定义为最优频率边界,进行频谱分割。通过对比仿真信号下最优频率边界、局部极大值和尺度空间法频谱分割结果,并结合一座人行斜拉桥工作模态分析,考察了基于最优频率边界的经验小波变换仿真信号分解和实际结构模态参数识别效果。结果表明:相比于局部极大值和尺度空间法,最优频率边界能够避免噪声和频谱泄露的影响,获得更为准确的频谱分割结果,提升经验小波变换分解的子分量信噪比。To improve the accuracy of spectrum segmentation of empirical wavelet transform,a spectrum segmentation method based on optimal frequency boundaries was proposed.And the optimal frequency boundary empirical wavelet transform was combined with random decrement technique,Hilbert transform and local least square method to identify structural modal parameters.The method established a series of frequency bands from the minimum bandwidth to maximum bandwidth centered on the frequency with the maximum amplitude of the Fourier spectrum.And it used empirical wavelet transform to extract subcomponents and corresponding residues of different frequency bands,and defined the bandwidth corresponding to the minimum value of the Pearson correlation coefficient as the optimal frequency boundary for spectral segmentation.The effectiveness of the empirical wavelet transform signal decomposition and the identification of actual structural modal parameters based on the optimal frequency boundary were examined by comparing the spectrum segmentation results of the optimal frequency boundary,the locmax,and the scalespace method of the simulated signal,as well as the operational modal analysis of a pedestrian cable-stayed bridge.The results show that,compared with locmax and scalespace method,the optimal frequency boundary can avoid the influence of noise and spectral leakage,obtain more accurate spectrum segmentation results,and improve the signal-to-noise ratio of subcomponents decomposed by empirical wavelet transform.

关 键 词：桥梁工程 工作模态分析 经验小波变换 频谱分割 最优频率边界 

分 类 号：TU311.3[建筑科学—结构工程]