故障诊断中的混沌参数分析

Chaotic Parameters Analysis in Fault Diagnoses

在建立的模拟实际故障的实验装置上采集模拟旋转机械振动故障的信号,运用相空间重构理论,对实测故障信号进行时间序列重构。为使重构相空间能充分地反映系统运动特征,必须恰当地选取时间延迟与嵌入维数。应用互信息函数法确定出不同故障时间序列时间延迟,应用改进伪近邻法确定最小嵌入维数。在此基础上计算关联维数和李亚谱诺夫指数,以此二参数作为特征量,有利于分析识别故障信号,增强可靠性。为复杂旋转机械故障诊断提供一种识别方法。

The simulating fault vibration signals were collected on the established experimental equipment.Using phase space reconstruction theory,the time-series measured fault signal was reconstructed.In order to make the reconstructed phase space adequately reflect the motion characteristics of the system,delay time and embedding dimension must be properly selected.The delay times of the time series of different fault were determined with the method of mutual information function and the minimum embedding dimension was determined with the method of improved false nearest.On this basis,the correlation dimension and Lyapunov exponent were calculated.Taking the two chaotic parameters as the characteristic quantities,it is favorable for recognizing fault signal,enhancing reliability.It provides an identification method for the complex rotating machinery fault diagnosis.