非线性广义系统传感器故障估计方法

The method of sensor fault estimation for nonlinear descriptor systems

随着工业控制系统对可靠性的要求日益提高,故障诊断显得越来越重要。当系统存在干扰且干扰上界未知时,有效地对系统中的传感器故障进行估计,需要进一步的研究。为此,针对传感器故障为常值突变故障或缓慢变化故障的情况,研究了一类含外部干扰的非线性广义系统的传感器故障估计问题,提出了一种基于增广状态观测器的传感器故障估计方法。通过状态重构将传感器故障作为增广状态的一部分,构造出一个新的广义系统并设计增广状态观测器,使其能够同时对原广义系统中的状态和传感器故障进行估计;考虑到系统中的干扰项,根据李雅普诺夫稳定理论并结合线性矩阵不等式(LMI)方法,分两种情形给出了估计误差系统渐近收敛的充分条件,保证了可以对原系统中的故障进行渐近估计;通过仿真算例验证了增广观测器对传感器故障估计的有效性,表明了所提方法的可行性。

With the increasing demand for reliability of industrial control systems,fault diagnosis becomes more and more important.When there are interference in the system and unknown upper bound of disturbance,the effective estimation of sensor faults in the system needs further study.A novel sensor fault estimation method based on augmented state observer is proposed for a class of nonlinear descriptor systems with external disturbances in view of the fact that the sensor fault is a constant abrupt fault or a slow change fault.Firstly,by considering sensor fault as a part of augmented state through state reconstruction,we constructed a new generalized system and designed an augmented state observer in order to estimate the state and sensor fault of the original generalized system simultaneously.Secondly,considering the disturbances in the system,we applied the Lyapunov stability theory and the linear matrix inequality(LMI)method and gave the sufficient conditions for asymptotic convergence of estimation errors in two cases,which ensures asymptotic estimation of faults in the original system.Finally,we verified the effectiveness of the augmented observer for sensor fault estimation with a simulation example,which shows the feasibility of the proposed method.