Intermediate Observer Based Fault Estimation And Fault-Tolerant Control For Fuzzy Stochastic Systems

With the continuous development of technology,industrial systems are becoming increasingly complex,and with the increase of working hours and the changing working environment,systems are prone to failures.At the same time,the requirements for control accuracy in industrial production are also continuously strict.Faults can affect control accuracy and even cause serious harm to the systems.Fault estimation and fault-tolerant control are important means to deal with faults,so fault estimation and fault-tolerant control of systems have received extensive attention from relevant scholars.On the other hand,actual systems are generally nonlinear systems,and fuzzy models can approximate nonlinear systems with a set of linear systems,which provides convenience for the analysis and design of nonlinear systems.At the same time,stochastic noise is very common in practical systems,which can affect system performance.Based on this,this paper mainly studies the problems of fault estimation and fault-tolerant control for fuzzy stochastic systems,mainly including the following three parts:In the first part,based on the full-order intermediate observers,the problems of fault estimation and fault-tolerant control for fuzzy stochastic systems with actuator faults and disturbance are studied.Firstly,the improved intermediate observers are devised to estimate the faults.The difference from the classical intermediate observer is that this method introduces the output estimation error feedback into the observer design,which increases the design freedom.Then a fault-tolerant controller based on the information estimated by the observer is designed.Lastly,a simulation is provided to show the reliability of the method.In the second part,based on the reduced-order intermediate observers,the problems of fault estimation and fault-tolerant control for fuzzy stochastic systems with actuator faults,sensor faults and exogenous disturbance are studied.Firstly,the reduced-order intermediate observers and disturbance observers are proposed to reconstruct the faults and disturbance.Then,according to the estimated information,the observer based fault-tolerant controllers are devised.Compared with the full-order observer design method,the designed reduced-order observers are with lower dimensions and less cost.And the obtained linear matrix inequality(LMI)conditions can ensure that the whole closed-loop systems are mean-square exponentially stable.Finally,two examples are provided to show the applicability of the method.In the third part,based on the dynamic intermediate observers,the problems of fault estimation and fault-tolerant control for switched fuzzy stochastic systems with actuator faults,sensor faults and disturbance are studied.Firstly,the dynamic intermediate observers are devised to reconstruct the faults,where the estimation error feedback terms of the output and its derivative are considered simultaneously,which can reduce the effect of the fast time-varying actuator fault.Then,based on the piecewise Lyapunov function and the average dwell time,the integrated design technique of fault estimation and fault-tolerant control is considered,this technique can ensure that the systems are mean-square exponential stability with the weighted H_∞performance index and avoid the influence of fault estimation error on the closed-loop system.The single-step LMI conditions are obtained to calculate the observer and controller gain matrices,simultaneously,which overcomes the coupling of the observer and controller gains.Finally,two simulations are offered to illustrate the advantage and the reliability of the method.