不規則入力を受ける非線形系のスペクトル解析および同定

概要

論文の詳細を見る
The purpose of this work is to identify the characteristics of a nonlinear vibration system subjected to a random input. In general, a statistical technique of a higher-order spectrum could be used for non-Gaussian data analysis. The dependencies between two frequency components are expressed. And the authors tried to understand the characteristics of a nonlinear system through bispectra. In this identification method, smoothing with a bispectral window is needed. The design of this bispectral window is dealed with an optimization problem here. Statistical characteristics for the design of bispectral window are bias and variance. The factor of the bias for this window is equal to that of the squared variance. Here, a numerical simulation was executed for an asymmetrical nonlinear vibration system. Moreover, the results of an experiment for one degree of freedom models with a cell sponge are examined and discussed.
一般社団法人日本機械学会の論文
1996-12-25