Research On Fault Detection Methods For Networked LPV Systems Based On LMI Technique

With the development of network communication technology and the deepening of the integration of network communication technology and industrial production,the application of networked control systems become more universal and intelligent.As a representative class of linear time-varying systems,LPV systems are used to describe the nonlinear and time-varying characteristics of many practical systems.With the widespread application of network technology in such practical systems,it is of great significance to study LPV systems in a network environment.However,with the introduction of the network environment,the complexity of the system continues to increase.Due to the impact of network bandwidth and network stability,data transmission is inevitably affected by data loss,network-induced delays and quantization error,and other issues,resulting in fault problems that affect normal production operations.At the same time,due to the increasing demand for the reliability of actual industrial systems,timely discovery of fault information is very important.Fault detection technology,as a means of ensuring the safe operation of a system,is an important basis for judging whether a fault occurs and ensuring the stable operation of the system,and has become a hot topic of research in recent years.Therefore,considering the factors of non-ideal network environment,this paper studies the fault detection methods for LPV systems.The main research contents are as follows:Firstly,for the LPV systems with measurement data loss and network-induced delay,the design of L₁ and L₂-L∞fault detection filters is studied.The Bernoulli random distribution sequence is used to describe the random loss of measurement data.The sufficient conditions for robust L₁ and L₂-L∞fault detection filters are obtained by using Lyapunov stability theory and LMI technology.The coupling between the system matrix and the Lyapunov function matrix is eliminated by the projection theorem,and the infinite dimensional PLMIs are converted into finite dimensional LMIs by using the approximate basis function and grid technology.Finally,the required feasible solution is obtained.Secondly,considering the existence of probability interval delay in the system and use Bernoulli random distribution sequence to divide the time delay by interval probability.At the same time,due to the limited network communication resources,with using the event-triggered mechanism to reduce the amount of data transmission,so as to improve the efficiency of data transmission,by constructing time-delay correlation and parameter dependent Lyapunov functional,the fault detection system is asymptotically stable and satisfies the corresponding suppression level under two types of noise disturbances.Finally,according to Lyapunov stability theory,the design method of robust L₁ and L₂-L∞fault detection filters with low conservation is given.Thirdly,considering the switched LPV systems with sensor saturation constraints,the fault detection problem under the influence of continuous peak bounded strong noise interference is studied.Based on the ADT method and Lyapunov stability theory,the delay-dependent and parameter-dependent L₁ performance criterion is obtained,which makes the fault detection system asymptotically stable and meets the expected performance index.Finally,the numerical simulation verifies that the two subsystems can achieve the expected fault detection level under the condition of low switching rate.Finally,in order to reduce the waste of network resources,an event triggering mechanism is adopted to improve the efficiency of information transmission.Considering the random occurrence of faults in the system,the L₂-L∞fault detection problem of LPV systems with random faults is studied.The fault weighted model is established to improve the fault detection performance.The threshold logic method is used to identify the fault signals in the system,and the parameter-dependent fault detection filter is selected to improve the detection accuracy.The sufficient conditions for the existence of the fault detection filter are obtained through the Lyapunov stability theory and meet the L₂-L∞performance index.The numerical simulation compares the optimal solutions under different fault occurrence probabilities.