基于经验小波变换的高墩大跨铁路连续梁桥模态参数识别

Modal Parameter Identification of High-pier Long-span Railway Bridge Based on Empirical Wavelet Transform

为了解高墩大跨铁路连续梁桥动力特性,对其进行环境振动试验,分别使用经验小波变换、经验模态分解,以及随机子空间识别三种方法识别其模态参数.介绍了松头江特大桥工程概况、环境振动试验概况;利用经验小波和经验模态分解对测试到的主桥横向、竖向加速度时程信号进行分解,获得单分量信号响应;结合随机减量技术和Hilbert变换识别桥梁模态参数,并利用随机子空间识别结果作为比对校核.结果表明:该桥前五阶自振模态中有三阶是横向振动,高墩大跨连续梁桥横向振动效应明显.三种方法均能较准确识别桥梁模态参数;相比经验模态分解,经验小波变换能够避免模态混叠和模态遗漏;相比随机子空间识别,能避免稳定图中的虚假模态.

Based on the environmental vibration test of high-pier and long-span railway continuous beam bridge,the modal parameters were identified by empirical wavelet transform,empirical modal decomposition and random subspace identification.Empirical wavelet and empirical mode decomposition were used to decompose the measured lateral and vertical acceleration time history signals of the main bridge,and the single-component signal response was obtained.The modal parameters of bridges were identified by combining random decrement technique with Hilbert transform,and the results of random subspace identification were used as comparison check.The results show that three of the first five natural vibration modes of the bridge are transverse vibration,and the transverse vibration effect of the long-span continuous beam bridge with high piers is obvious.All three methods can accurately identify the modal parameters of bridges.Compared with empirical mode decomposition,empirical wavelet transform can avoid modal aliasing and modal omission.Compared with random subspace identification,it can avoid false modes in the stable graph.