基于观测器的非线性系统传感器故障估计方法  

Sensor Fault Estimation Based Observer for the Nonlinear Systems

作　　者：孙延修[1] 杜莹[1] 黎虹[1] SUN Yan-xiu;DU Ying;LI Hong(Department of Basic Course,Shenyang Institute of Technology,Fushun 113122,China)

机构地区：[1]沈阳工学院基础课部,辽宁抚顺113122

出　　处：《火力与指挥控制》2020年第12期43-47,52,共6页Fire Control & Command Control

基　　金：辽宁省教育厅科学研究基金资助项目(L201906)。

摘　　要：现代工业系统对安全性与可靠性的要求越来越高,故障诊断技术逐渐成为研究的热点之一。针对一类Lipschiz非线性系统,提出一种基于增广状态观测器的传感器故障估计方法。将传感器故障作为增广状态的一部分,构造出一个广义系统,通过对广义系统进行等价变换,进而设计出增广状态观测器,使其能同时对原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,利用Lyapunov函数和线性矩阵不等式(LMI)方法,得到了增广状态观测器存在的两个充分条件,充分条件既可以保证观测误差系统的收敛速度,同时也可通过求解最优问题,减少系统干扰对传感器故障估计的影响;为保证增广状态观测器对系统传感器故障的鲁棒估计,通过仿真算例证实了增广状态观测器误差系统为渐近收敛,验证了传感器故障估计方法的有效性。Modern industrial systems require more and more safety and reliability,and fault diagnosis technology has gradually become one of the research hotspots.A sensor fault estimation method based on augmented state observer is proposed for a class of Lipschiz nonlinear systems.Firstly,a generalized system is constructed by taking sensor fault as part of augmented state,an augmented state observer is designed by equivalenting transformation of the generalized system,and which can estimate the state and sensor fault of the original system simultaneously.Secondly,considering the disturbances in the system,two sufficient conditions for the existence of the augmented state observer are obtained by using Lyapunov function and the method of linear matrix inequality(LMI),the sufficient conditions can guarantee the convergence rate of the observation error system and reduce the influence of system disturbance on sensor fault estimation by solving the optimal problem.Finally,in order to guarantee the robust estimation of sensor faults,the simulation example proves that the error of the augmented state observer is asymptotically convergent,and the effectiveness of the sensor fault estimation method is verified.

关 键 词：非线性系统 故障估计 线性矩阵不等式 增广状态观测器 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]