Control And Filtering Of Discrete-Time Markov Jump System With Event-Triggered Scheme

Due to the rapid development of communication technology and computer network in recent years,the traditional network control system has gradually formed a network control system(NCS),which is a closed loop control system that transmits data through the network.It has low cost,easy to install and maintain,interactive and good,so the network control system has been widely studied by scholars.As a result,NCS has been widely studied by scholars.However,the limited network bandwidth also brings various problems to NCS,such as delay,packet loss and so on,which usually affect the stability of the system.In order to solve this problem,scholars have proposed an event triggering mechanism,which can filter out some"unnecessary" data and effectively reduce the burden of network transmission.At the same time,the introduction of quantizer can further improve the data transmission capacity.In practice,there are many dynamic systems with catastrophic characteristics,such as biochemical systems,power systems,economic systems and NCS.Markov jump systems can be used to simulate these systems,so it is more practical to treat discrete Markov jump systems as controlled objects.Not only the traditional control system is gradually networked,but also the neural network tends to be networked.Recently,there are many research results on filtering problem of time-delay neural networks,but to our best knowledge,there are few studies on discrete Markov jump time-delayed neural networks based on event-triggered mechanism.This paper mainly studies the control problem of discrete Markov jump system and filtering of neural networks based on event-triggered mechanism.The main work and research results are as follows:(1)Firstly,the research background,purpose and significance of the network control system are introduced.The shortcomings of the network control system and the research situation of these problems of home and abroad are analyzed.Through the analysis of the current research results,the characteristics of each solution are summarized.(2)For a class of discrete Markov jump systems,in order to reduce the occupation of network resources,an event generator is introduced at the sampler end.On this basis,the pair quantizer is introduced to the state feedback channel and the control inputted channel respectively to further improve the data transmission efficiency.Since inevitable time-delay existed in the network,a unified closed-loop control system model is established by using the time-delay analysis method.According to Lyapunov stability analysis and Jenson inequality,the stability and performance index of the system are analyzed,and the corresponding output feedback controller is designed(3)For the discrete Markov jump delayed neural network,the filtering problem under event-triggered mechanism is studied.Considering network delay,neuron delay and signal quantization,the system is established as a discrete Markov jump delayed neural network filtering error system with time-delay,quantization and event-triggered mechanism.By using Lyapunov stability analysis,and linear matrix inequality(LMI),the sufficient conditions are obtained to guarantee the stability and H? performance index of the augmented system,and the H? filter is designed.