基于LMD的非线性结构系统识别

Nonlinear structural system identification based on LMD

提出了一个基于局部均值分解(local mean decomposition,LMD)识别的非线性多自由度结构系统的新方法。首先运用LMD将非线性结构的自由响应信号自适应地分解为有限个乘积函数之和,然后由每个乘积函数分量的包络函数和纯调频函数求取相应的瞬时幅值和瞬时频率,根据自由响应信号瞬时特征间的关系,进而得到非线性结构的骨架曲线。算例仿真结果验证了该方法的有效性和可行性。

In this paper, a nonlinear multi - degree of freedom structure system identification novel method based on local mean decomposition （LMD） was proposed. First, freedom response signal of nonlinear structure was adap- tively decomposed into a finite number of product function （PF） by LMD. Corresponding instantaneous amplitude and instantaneous frequency were found through each PF component envelope function and pure FM function. Ac- cording to the relationship between instantaneous characteristics of free response signal, the nonlinear structural skeleton curve was obtained. Simulation results verify the effectiveness and feasibility of the new method.