Research On Fault Detection Methods Of Networked Control Systems With Mixed Delays

As industrial production gradually enters a new era of networked intelligence,network control systems have good application prospects in the fields of aerospace,remote monitoring,communications and transportation,etc.due to their low cost,flexible structure and easy installation.Compared with the conventional point-to-point control system,this closed-loop control system that can be networked and distributed can perform many tasks in a long distance by connecting the network space and the physical space.The network signals are transmitted by the shared network?and wired?.However,due to the limitation of the network channel bandwidth,the network itself will inevitably cause data loss,delay and other problems that can cause system failure.These will have a far-reaching effect on the stability of the system Complex impact.With the increasing size and complexity of industrial systems required in the 21st century,researchers have also given higher hope for the safety and reliability they show.Any component failure can cause production units.Very serious economic loss.Fault detection is not only an inspection method for judging whether the system is faulty,but also an important basis for establishing an early warning mechanism to ensure the safe operation of the system.Therefore,studying the problem of fault detection in networked systems in the field of automatic control has important theoretical significance and practical application value.This paper selects different system performance indicators for two types of external noise interference,energy-bounded and continuous peak-bounded,and gradually explores the fault detection problems of continuous linear networked systems and nonlinear networked systems,while considering the mixed delay,Data loss and sensor saturation and other constraints,and then select H_?and L₁two performance indicators to analyze the stability of the system,and finally through MATLAB simulation to prove the feasibility and effectiveness of the method designed in this paper.The main research contents are as follows:First,study the problem of network system fault detection with time-varying delay.Solve the system fault detection problem by solving the observer-based H_? fault detection filter,construct a closed-loop control system and analyze the system stability using Lyapunov stability theory,and filter the existence of H_?fault detection based on the linear matrix inequality technique The filter provides sufficient conditions,and then uses the convex optimization method to obtain the parameter matrix of the designed fault detection filter,and the effectiveness of the method is verified by simulation.On the other hand,the fault detection filter based on the filter form is used to generate the residual signal,and the Lyapunov stability theory and linear matrix inequality method are used to meet the L₁ performance index while maintaining the asymptotic stability required by the system.The simulation verifies the feasibility of the design method.Secondly,considering the probability interval delay in the controlled object,further study the fault detection problem of networked system.Assuming that the time-varying delay has certain random characteristics in the network environment,the Bernoulli random variable is introduced to represent the probability of delay distribution at different intervals,and the probability distribution information is used to create a new closed-loop control system model.Using methods such as Schur's complement lemma and integral inequality,the fault detection system meets a certain level of H_?interference suppression.The residual evaluation function value is used to determine whether the system fails at the instant of failure.On the other hand,for the external interfering signal is a continuous interval bounded probability interval time-delay system,the problem of L₁ fault detection with sensor saturation constraints is studied.By constructing the Lyapunov-Krasovskii function with time-delay correlation,the performance criterion that makes the system mean square asymptotically stable and has L₁noise suppression level is established.The low conservative fault detection filter of the system is designed using linear matrix inequality technology.Numerical simulation shows that the proposed method can detect the fault signal accurately and quickly.Thirdly,for the nonlinear system with distributed time delay described based on the T-S fuzzy model,the H_?fault detection problem of the system is studied.First,based on the principle of parallel distribution compensation and Lyapunov theory,by constructing the Lyapunov-Krasovskii function with time-delay correlation and basis dependence,the designed fault detection filter guarantees the asymptotic stability of the system,and performs congruent transformation and variable substitution to establish the external Noise interference is the H_?performance criterion when the energy is bounded,and the matrix parameters of the target fault detection filter are obtained.MATLAB simulation shows the feasibility of the designed method.Finally,for the network control system with random measurement data loss,random delay and distributed delay,the problem of L1 fault detection whose external interference signal is continuous peak and bounded is studied.Design a fault detection filter that generates a residual signal,use Bernoulli random distribution to describe the random measurement data loss and random delay of the system itself,and construct a delay-dependent Lyapunov-Krasovskii functional to make the fault detection system asymptotically stable,The L₁ suppression level is met under continuous strong noise interference,and the L₁ filter parameters for low-conservative fault detection are gained in the form of linear matrix inequalities.The data simulation shows that the method can detect the fault signal effectively and sensitively.