Event-triggered Intelligent Control For Several Types Of Unknown Discrete-time Nonlinear System

With the increasing development of science and technology,the digitization and networking of industrial systems is an inevitable trend,and the research on the control of networked systems is a basic requirement to promote the development of industrial technology.Traditional control design methodology is difficult to get rid of the dependence on accurate mathematical model of the target system,so it is relatively weak in solving the control problem of industrial systems with high uncertainty and nonlinearity.Intelligent control is discipline formed by the intersection of artificial intelligence,automatic control,computer science and other disciplines.Its emergence provides an effective way to solve the control problems of uncertain nonlinear systems.Strict-feedback nonlinear systems have received extensive attention for they can model many practical systems,and a large number of intelligent control schemes for such systems have been proposed.However,the existing research results are mainly for continuous strictfeedback nonlinear system,and the digital control urges the research on intelligent control of discrete strict-feedback nonlinear systems.On the other hand,integrating communication network into control-loop can bring much convenience to the installation and maintenance of the system,but it also brings new challenges to the control design of the system.Traditional timetriggered control scheme requires periodic communication among the components.If this kind of scheme is applied to a networked control system,it will cause unnecessary waste of resources,even lead to network congestion,thereby affecting the stability of the system.Compared with time-triggered control,the event-triggered control can greatly save network bandwidth by transmitting the data only when it is needed.Consider the networking and digitization development direction of industrial systems,it is very necessary to study event-triggered intelligent control of discrete strict-feedback nonlinear systems.Although some scholars have done a little research on event-triggering intelligent control of uncertain discrete strict-feedback nonlinear systems,there are still many problems worth studying.Based on a few existing results,this paper studies the event-triggering intelligent control of uncertain discrete strict-feedback systems,and improves some deficiencies in the existing literature.In order to avoid the non-causal problem in the control design of discrete strict-feedback system by directly using backstepping method,this paper proposes a novel event-triggering backstepping control design framework.Combining this framework with adaptive neural network,adaptive fuzzy logic system,reinforcement learning and other technologies,several control schemes are proposed for discrete strict-feedback single-input single-output,multi-input multi-output and multi-agent systems with limited network resources,input saturation,unmeasured state and other engineering problems.These solutions can greatly reduce the network data transmission from the sensor to the controller channel while ensuring the control performance of the system.The development of this thesis can not only further improve the existing networked system control theory,but also provide theoretical guidance for the construction of networked control systems in practical engineering.