Research On Event-Triggered Adaptive Dynamic Programming Algorithm For Networked Systems

In the past ten years,networked control has been one of the hot research directions in academic circles,among which packet loss and low communication efficiency have been the most concerned.Aiming at the problems of packet loss and low communication efficiency in the communication network,this paper proposes to combine the eventtriggered adaptive dynamic programming algorithm,and systematically studies the eventtriggered adaptive dynamic programming control algorithm for nonlinear networked systems.The main content of this paper includes the following three aspects:1.The event triggered double heuristic dynamic programming control algorithm for nonlinear networked systems with packet loss and low communication efficiency is studied.Firstly,in order to save the computing cost and communication resources,the event-triggered condition is designed to determine whether the sensor samples or not;Then,the random variables according to Bernoulli distribution are introduced to describe packet loss;Finally,the algorithm implementation and structure of event triggered double heuristic dynamic programming are given,and the controller that can approximate the optimal feedback control law and the optimal performance index function is obtained.2.The event-triggered globalized dual heuristic dynamic programming control algorithm for nonlinear networked systems is studied,which can improve the control accuracy of the system.Firstly,the event-triggered condition is designed to reduce the number of network communication and the computational burden of the globalized dual heuristic dynamic programming algorithm;Then,the algorithm implementation and design method of event-triggered globalized dual heuristic dynamic programming are given;Finally,the approximate optimal control law is solved at each event-triggered time to reduce the communication and computation of the system.3.The event-triggered adaptive dynamic programming control algorithm for load frequency of power networked systems is studied.Firstly,the control system model of load frequency is established;Then,the action dependent heuristic dynamic programming is used as the controller algorithm to solve the approximate optimal control law.The learning ability of neural network is used to improve the robustness and adaptability of the algorithm on the premise of ensuring the stability of the system;Finally,in order to reduce the computational cost of the system,the event-triggered update mechanism is designed,and it is proved that the controller can ensure the stability of the closed-loop system under the event triggered mechanism.