非平稳随机激励下线性系统模态识别的小波方法

Method of modal identification based on wavelet transform in LTI system under non-stationary random excitation

通过对非平稳随机激励下的线性时不变系统的响应x(t)进行小波变换.利用各级小波系数的正交性,对其中的某些平稳小波系数进行合成,得到满足结构动力学方程的新平稳信号y(b),用其代替原始的结构响应x(t),结合其他各种平稳随机激励下的模态识别的方法,进行非平稳随机激励下系统参数的识别.仿真结果表明,该方法可以在一定程度上消除非平稳随机激励所引起的响应信号的非平稳性.

Modal parameters＇ identification of linear system to non-stationary random excitation has been a hot research topic in recent years. A new synthesis stationary signal y （b） based on wavelet transform of original response x（t） and wavelet coefficients＇ orthogonality is presented in this paper. The new signal, which satisfies structural motion equation, instead of original response, is used in identifying the modal parameters of non-stationary linear system, combined with the method of modal identification under stationary random excitation. The numerical results show that the method can remove the non-stationary of output signal under non-stationary random excitation to a certain extent.