Robust Control Of Networked Systems And Research Of Swarm Behavior

This thesis investigates some stability and robust control problems of networked control systems (NCSs), and some behavior of swarm under digital communication network.Firstly, the stability of networked system based on data packed is studied. For discrete networked system with input and output data packed, we present a design of a controller, and analysis the stability of the corresponding closed-loop systems. It was proved that the design of the observer and the state feedback controller of the augmented system satisfied separation principle. The corresponding controller is given in terms of linear matrix inequalities.Secondly, for discrete networked system with uncertainties, subject to the dropout of transmitted data we concern with the robust H_∞ control using state feedback and estimated state feedback based on an observer, respectively. By means of linear matrix inequality approach, the robust H_∞ control law is designed. The controllers guarantee the robust stochastic stability of the corresponding closed-loop systems with prescribed performance level. The effectiveness of the proposed design approach are illustrated by numerical example.Thirdly, the robust stability problem for discrete system with stochastic uncertainties and nonlinear perturbations is studied. Under the assumption of data dropout being an identically independently distributed process, the networked control system is modelled as a stochastic switching uncertain system. Based on a stochastic Lyapunov function method, sufficient conditions which ensure the robust asymptotic stability in the mean square are obtained.Fourthly, we investigate a formation control for dynamical agents in digital communication network. Each agent in the network is described by a continuous-time dynamics and the communication network is described by a undirect graph. When there is no time-delay, we give the sufficient conditions of agents achieving consensus stability using some properties of graph theory and Jury criterion. Under the assumption of time-delay during data transmission, we give the sufficient condition of agents achieving consensus stability using approach of linear matrix inequalities. Numerical simulations are provided to demonstrate the effectiveness of our results.Finally, collective behavior of two-dimension multi-agent system is studied. We also assume that they are communicated through digital networked. Based on the properties of Laplacian associated with the graph and Jury criterion, we give the conditions of consensus stability for the cases of no leader and existing a virtual immobile leader, respectively. Numerical simulations are provided for the different cases.