航空学报 > 2002, Vol. 23 Issue (6): 520-524

基于LMI的一类非线性离散动态系统的鲁棒故障诊断

陈玉东, 翁正新, 施颂椒   

  1. 上海交通大学自动化研究所 上海 200030
  • 收稿日期:2001-07-30 修回日期:2002-03-01 出版日期:2002-12-25 发布日期:2002-12-25

ROBUST FAULT DIAGNOSIS FOR A CLASS OF NONLINEAR DISCRETE DYNAMIC SYSTEMS BASED ON LMI

CHEN Yu-dong,WENG Zheng-xin,SHI Song-jiao   

  1. Research Institute of Automation, Shanghai Jiaotong University, Shanghai 200030, China
  • Received:2001-07-30 Revised:2002-03-01 Online:2002-12-25 Published:2002-12-25

摘要:

针对一类非线性不确定离散动态系统 ,提出了一种新型鲁棒故障诊断方法。该方法不但能够对被诊断系统进行故障检测 ,而且同时能够实现故障的分离和辨识。它首先通过构造一个辅助系统 ,将故障的辨识问题转化为min max问题 ,并且通过巧妙设计辅助系统的输出增益矩阵 ,使得辅助系统与被诊断系统的状态差值方程和输出差值方程稳定 ,然后将min max问题转化成LMI问题 ,最后通过求解LMI问题来实现故障诊断。分析了故障诊断方法的鲁棒性、灵敏度和故障的辨识误差。仿真结果验证了该方法的有效性。

关键词: 故障诊断, 故障辨识, 非线性系统, 离散动态系统, 线性矩阵不等式, 鲁棒性, 灵敏度

Abstract:

A new method of robust fault diagnosis for a class of nonlinear discrete dynamic systems with uncertainties is proposed. It can not only detect faults, but also isolate and identify faults. Firstly, it converts the problem of fault identification into a min max problem by constructing an auxiliary system, and designs the gain matrix in the auxiliary system appropriately such that the state difference equation and the output difference equation of the auxiliary system and the systems to be diagnosed are both stable, then converts the min max problem into a LMI problem, and finally realizes the fault diagnosis by solving the LMI problem. The analyses for robustness, sensitivity and error of fault identification are also considered. Simulation indicates its effectiveness.

Key words: fault diagnosis, fault identification, nonlinear systems, discrete dynamic sysuems, Linear Matsix Inequality (LMI), robustness, sensitivity