不定线性系统的鲁棒故障检测

ROBUST FAILURE DETECTION FOR UNCERTAIN LINEAR SYSTEM

1.引言 Beard等提出的检测滤波器是一种基于模型的故障检测与识别(FDI)方法,然而,检测滤波器要求精确地已知系统的动力学方程 X=AX+Bu,Y=CX (1) 其中,X∈R^n,Y∈R^m,u∈R^r分别为系统状态输出和控制;A、B、C为具有相应维数已知常阵。由于系统的复杂性以及系统外界干扰的不确定性,获取系统精确的数学模型非常困难。

The failure detection of dynamical systems is dealt with. There are a lot of approaches for failure detection of dynamical systems, but most of them assume that the model of the system is an exact model without any error It has been proved in prac-tice that this assumption is unrealistic. So a robust failure detection filter is proposed to solve the failure detection of dynamical systems with model error. The basic idea :is that the control make the model error unimportant if there is no failure, and the failure detection filter: gives the failure information if there is a failure. The computer simulation results have proved that the new approach in this paper is effective for robust failure detection of dynamical systems.