Research On Adaptive Neural Networks Event-triggered Control For Uncertain Systems

With the continuous development of computer,communication technology,network control system is more and more popular.However,while it brings new development opportunities in the field of control,because of the limited bandwidth,various problems can occur when information is being transmitted at very high frequencies,which will affect the performance of the system and even the stability of the system.In order to deal with this problem,reducing the transmission frequency is an effective method,As a typical non-periodic control method,event trigger control has attracted people's attention.This paper makes a further study on the problem of event trigger control for a class of uncertain systems.Due to the uncertainty of the system,the existing methods of event trigger control cannot be used directly.Neural networks are widely used in uncertain systems because of their strong learning ability and nonlinear approximation ability.Therefore,this paper firstly adopts the traditional adaptive neural network algorithm to ensure the performance of the system,and then introduces the event triggered mechanism to reduce the signal transmission frequency and communication amount.Compared with the traditional adaptive neural network controller,the transmission and updating of data are determined by events rather than time,which can reduce the communication frequency and improve the stability of the network system.Two switching event triggering mechanisms are proposed for continuous strict feedback systems based on state feedback.The switching strategy can avoid Zeno phenomenon and has reasonable triggering frequency,while the adaptive neural network control algorithm guarantees good performance of the system.For discrete strict feedback non-linear systems,adaptive neural network event triggering control based on state feedback and output feedback is studied respectively.Firstly,the system is transformed into a series of predictive systems to avoid causal contradiction.On the basis of adaptive neural network control,a switched event triggering mechanism is proposed,which can reduce communication frequency on the basis of guaranteeing good performance of the system.In this paper,Lyapunov function is used for segment discussion,rigorous mathematical formula is used for proof,and matlab simulation software is used to verify the effectiveness of the proposed algorithm in both theoretical and practical aspects.The research content of this paper is of great significance in both theoretical research and practical application.