Dynamics Analysis And Control Of Stochastic Hybrid Systems Based On Discrete Time Observations

In the past two decades,due to the continuous development of science and technology and the increasing improvement of computer technology,the stochastic hybrid system has been widely concerned by a large number of researchers in the field of control.At the same time,most systems such as network system,biological system,financial system and power system can be described by stochastic hybrid system,so stochastic hybrid system has a wide range of applicability.This paper takes the stochastic hybrid network dynamics system as the main research object,and comprehensively considers the consistency of the network system with markov switching,the consistency of the second-order multi-agent network system and the consistency of the multi-agent system with leader.The main work is as follows:1.Firstly,the dynamic behavior of a stochastic hybrid system with markov switching is considered.Secondly,the feedback control problem of a stochastic network system with markov switching based on discrete time observations is studied,in which the consistency protocol is designed for some parts.Then,by selecting an appropriate Lyapunov function and combining It(?) formula,linear matrix inequality,lipschitz condition and basic inequality,the mean-square asymptotic consistency condition of a stochastic hybrid network system with markovswitching,the mean-square asymptotic consistencyconditionis obtained based on discrete time observation.Finally,numerical modelthrough Matlab,and the effectiveness of the control input protocol and the correctness of the theoretical results are further verified,giving the simulation resultsat the same time.2.This paper mainly studies the distributed feedback control of second-order multi-agent network system based on discrete time observation.Firstly,a complex stochastic network multi-agent system is constructed.According to the complexity of the interaction of each node in the speed and position,the feedback control in discrete time is designed.Secondly,the stochastic stability theory,lipschitz condition,basic inequality and It(?) formula are used to calculate the minimal generator of the second-order multi-agent system,so as to discuss the stability of the multi-agent system.Then,the mean-square asymptotic consistency condition of the second-order multi-agent stochastic network system based on discrete time observations are obtained.Finally,the simulation and theoretical results are given through Matlab experiments,and the validity of the protocol is verified.3.On the basis of the second-order multi-agent system model based on discrete time observation,the consistency of the second-order multi-agent network system with leader-follower is discussed.By using the same method,the mean-square consistency condition and the theoretical result of the leader information received by the node information of the follower are obtained.