Nonlinear Systems Fault Diagnosis Based On High Gain Adaptive Observer

With the increasing complexity of equipment in modern society,the cost of equipment becomes more and more expensive because of the improvement of equipment precision.Because the system fails,it will have a huge impact on personnel safety and equipment damage,the research on equipment fault diagnosis has theoretical and practical value.In reality,most of the phenomena are presented in a nonlinear state.When the equipment becomes more and more complex,the nonlinear degree of the system may become higher and higher,and the high complexity of the equipment also brings the great challenge of reducing the accuracy of the system modeling,which the system modeling is always imprecise and showing a high degree of uncertainty.Based on the above reasons,the fault diagnosis of the system becomes more complex and difficult.The fault diagnosis of nonlinear systems with uncertainties has become one of the research hotspot.In this paper,the problem of fault diagnosis for a class of nonlinear systems with uncertainties is studied,the uncertainty of this kind of system lies in that some parameters of the system are unknown,and the unknown parameters are presented as constants,These unknown parameters can be estimated by some online adaptive algorithms,Therefore,if the system is in normal state,the parameters are stable,When the system fails,it can be considered that some parameters have changed,Accordingly,the estimated parameters will also change,Therefore,the research of fault diagnosis based on estimated parameters is of great significance.In this paper,the problem of fault diagnosis for a class of nonlinear systems with unknown parameters is studied.This paper will introduce mainly the contents from the following aspects:The nonlinear systems with unknown parameters and its high gain adaptive observer and is studied.There are some methods for stability analysis by constructing equivalent multi delay systems.Some methods use Lyapunov's second method for stability analysis.In this paper,the stability analysis is based on the second Lyapunov method.The Lyapunov candidate function will be proved to be stable and convergent through the partial signal boundary of the error system.The signal boundary conditions are obtained by analyzing the observability of the system.Other fault diagnosis methods based on adaptive parameter estimation are studied and taken as a control experiment.The advantages and disadvantages of this method and other adaptive parameter estimation are analyzed,according to the performance of fault diagnosis and the observer construction.A control experiment was set up.One is to construct adaptive parameter estimation of equivalent multi delay system according to the actual system.The other is the adaptive parameter estimation method based on the actual system.A dynamic model of a two link robot is built by Lagrangian mechanics,and the relationship between the torque and the angle of the manipulator is obtained.The dynamic model is transformed into the corresponding state space equation,and the corresponding parameters of the robot are set,according to the state space equation of the robot.A high gain adaptive observer is designed to reconstruct the fault state,according to the adaptive parameter estimation,fault detection and fault separation are carried out.