Research On Fault Detection For Discrete-time Networked Control Systems Based On Event-triggered Mechanism

In the industrial production process,discrete-time networked control systems are widely used in aerospace,remote monitoring,communication and transportation.Networked control systems connect the controlled objects,sensors,controllers,actuators and other components of the system through a shared network,which also are able to carry out missions over long distances.In addition,this kind of networked control systems has the advantages of low cost,flexible and stable structure and easy installation.However,communication resources of network systems are limited.Subjecting to the networked bandwidth,data may occur time delays and information loss thus triggering systems faults and affecting normal industrial activities.With the increasing demand for the reliability and safety of control systems in the industrial production process,the occurrence of any systems equipment failures may cause production stagnation or life safety problems.Fault detection can not only detect and identify faults as early as possible,but also take corresponding measures before the occurrence of faults,so as to minimize performance degradation and avoid dangerous situations.In the process of network communication,the traditional periodic sampling data may not be transferred over the network to filter.On the premise of guaranteeing the desired performance,the event-triggered strategy can be designed to reduce the amount of transmitted data to reduce the network load.Therefore,it has great significance to research the fault detection problems of discrete-time networked control systems based on the event-triggered mechanism.The following contents are mainly studied:Firstly,for a class of discrete-time uncertain distributed delays networked control systems with transmission delay and random packet dropouts,the problem of fault detection based on event-triggered mechanism is studied.Event-triggered strategy is used to improve the transmission efficiency of information and reduce the usage of the network bandwidth.Norm bounded uncertainty and Bernoulli distribution random variables are introduced to describe the uncertainty of the system model and the random packet dropouts in network information transmission.The sufficient conditions for the asymptotic stability of the fault detection system and the existence of the fault detection H∞ filter are obtained by using Lyapunov stability theory.Then,the parameters of the fault detection filter and the event-triggered weight matrix are obtained by using the convex optimization algorithm.The effectiveness of the proposed method is verified by numerical simulation.Secondly,since communication resources are limited in actual engineering,the controlled object will also be nonlinear.Therefore,the common nonlinear T-S fuzzy model is used to study fault detection based on event-triggered mechanism.On the one hand,considering the distributed delays and random packet dropouts in the system,the random packet dropouts is described by using the random white noise sequence which meets the Bernoulli distribution,and the fault detection dynamic system is obtained by using the parallel distribution compensation algorithm and event triggered strategy.According to Lyapunov stability theory,sufficient asymptotic stability conditions for the fault detection dynamic system is derived.The fault detection problem is transformed into a convex optimization problem by using LMI technology,and the effectiveness of the design scheme is verified by numerical simulation.On the other hand,the problem of fault detection for discrete-time T-S fuzzy systems with random mixed time delays is studied under the event-triggered strategy.Multiple independent random variables subject to Bernoulli distribution are used to control the random occurrence of multiple discrete time delays,infinite distributed delays and successive packet dropouts.Then a fuzzy fault detection filter based on event-triggered mechanism is designed to construct the fault detection system,which ensures the mean-square asymptotic stability of the fault detection system and reduces the error between the residual signal and the fault signal as much as possible.Thirdly,most physical systems are affected by random parameter changes.Using the dynamic modeling characteristics of Markov chain,Markov jump systems can simulate multiple actual systems and describe the random jump of systems parameters.Therefore,the problem of fault detection for a class of Markov jump systems with discrete distributed timedelay is investigated.For different jump modes,an event-triggered communication scheme is constructed to select the output signal to be transmitted.By analyzing the networked time delays,event-triggered mechanism,transmission time delays and distributed delays are unified into a discrete-time Markov jump fault detection system with distributed delays based on eventtriggered mechanism.Based on time-delay systems analysis method,the random stability criterion of a discrete-time Markov jump fault detection system is given,and the corresponding fault detection filter and event-triggered parameters are obtained.The numerical simulation shows that the proposed scheme is effective and the event-triggered communication scheme can greatly save the limited network resources.Finally,the dynamic output feedback control can improve stability,reduce the control cost,and effectively meet the H∞ performance indicators of closed-loop control systems.Therefore,based on a two-terminal event-triggered mechanism,the dynamic output feedback control problem of discrete-time network control systems is studied,and a kind of sensor to the controller of the channel and the controller to the actuator two-terminal event-driven mechanism are put forward to reduce the network load.A dynamic output feedback controller based on event trigger mechanism is designed,and the closed-loop model is constructed.According to Lyapunov stability theory and LMI method,the criterion to guarantee the mean square asymptotic stability of the system and to satisfy the performance are obtained,and the solution methods of the controller and the event triggering parameters are given.Finally,the effectiveness of the designed controller is verified by numerical simulation.On this basis,the proposed event-triggered condition is adopted,and two independent event triggers are used to reduce the information transmitted to the network channel.The problem of fault detection filter and controller simultaneous design for the discrete-time networked control systems based on event-triggered mechanism is studied.An integrated module containing fault detection filter and controller is designed,and two independent event-triggered conditions are used to construct event triggers respectively,which reduces the measurement data between sensor and integrated module,and the transmission of control data between integrated module and the system,thus improving the efficiency of information transmission.Considering the mixed time delays and random packet dropouts in the system,the Lyapunov stability theory is used to ensure the mean square asymptotic stability of the model and achieve the desired performance of fault detection and control.The effectiveness of the proposed method is verified by a simulation example.