Event-triggered Control Of Networked Switched Systems With Input Constraint

Switched systems are normally composed by a series of continuous or discrete subsystems and the rules governing these subsystems.For complex systems and multi-objective control problems,switched systems can make the system description more complete and the control design more flexible.On the other hand,with the rapid development of sensor,signal processing and communication technology,the studies of networked control systems have gained increasing attention.Recently,a new type of control mechanism,event-triggered control,has become a hot research topic.It can not only ensure control performance but also significantly reduce the communication frequency and unnecessary data sampling,thereby saving system resources.From the above,the event-triggered control of networked switched systems with input constraint is studied in this thesis.This thesis focuses on two aspects.On the one hand,the thesis carries out in-depth research on switched systems under different switching laws.First,the switched systems with the average dwell time switching law are taken as the controlled object.Then,the controller internal uncertainty and the system external disturbance are considered to study the non-fragile H_?control problem.On this basis,a new mixed time and state switching law is introduced,and the H_?control performance of switched systems subject to network transmission delay is studied.Furthermore,the networked switched systems with actuator saturation and network attack are considered,and a switching signal based on event-triggering strategy and network attack is constructed.Finally,the system stability is analyzed based on the maximum estimation of attraction domain.On the other hand,the continuous and discrete event-triggering strategies are respectively introduced into switched systems,and the Zeno behaviour caused by continuous event-triggering strategy is analyzed.Moreover,the improved event-triggering strategies based on adaptive threshold parameter and waiting time T are proposed,which can adjust the triggering frequency according to system performance and avoid Zeno-like behavior.In particular,the coupling phenomenon between switching instants and data updating instants is clarified in the system performance analysis.In addition,the effectiveness of each proposed method is verified by simulations.