Research On Guaranteed Performance Control Of Nonlinear Systems Based On Adaptive Dynamic Programming

With the development of science and technology,the complexity of engineering control systems continues to increase.In the event of a failure,it may cause the performance of the entire system to drop,and even affect the stability of the system,causing serious losses.It is especially important to improve the safety and reliability of your control system.Due to the complexity of the nonlinear system itself,the development of its control theory is not perfect,and most of the existing results mainly study the stability of the faulty system.On this basis,the system performance optimization problem is rarely considered.Therefore,the research on the performance control problem of nonlinear systems has very important theoretical and practical value.In this thesis,the problem of guaranteed performance control of nonlinear systems is studied.The main contents are as follows:Firstly,the problem of guaranteed cost control for uncertain nonlinear systems is considered.The nonlinear system under consideration is more general,and the goal of guaranteed cost control is to reduce the uncertainty and the impact of actuator failure on system performance.In this thesis,the uncertainty matching and the constraints defined by the upper bound are relaxed,and a modified performance function is introduced as the optimization target to realize the guaranteed performance control of the system.On this basis,the adaptive dynamic programming method is proposed to learn the guaranteed cost control strategy.Finally,numerical simulation examples verify the effectiveness of the proposed method.Secondly,we consider the reliability guaranteed performance control problem of large systems under time-varying actuator failure.In order to weaken the influence of actuator failure and ensure the performance of the system,a kind of guaranteed cost control strategy is designed based on adaptive dynamic programming method.An unknown actuator failure is assumed to have a known upper bound.In the case where the interconnection items are unknown and do not match,the guaranteed cost control problem of a large-scale nonlinear system can be transformed into the optimal control strategy for designing isolated systems,which is necessary to solve the HJB equation.The conclusion is drawn by using an adaptive dynamic programming(ADP)method of the execution-evaluation network.Even with time-varying actuator failures and unknown interconnections,the convergence of neural network weights and the stability of closed-loop systems are rigorously demonstrated.Finally,the effectiveness of the proposed method is verified by simulation examples.Next,we consider a class of performance-based tracking control problems for large systems based on event triggering.In order to save storage resources and avoid constant updates of the controller,we introduce an event trigger mechanism.We extend the system to convert the performance tracking control problem of large systems into the optimal control problem of the extended system.On the basis of the extended system,the event trigger condition is designed,and when the event-based error is greater than the designed threshold,the event is triggered and the controller is updated.The ADP method is used to approximate the control strategy,and the stability analysis is implemented based on the Lyapunov stability theorem.Finally,the work of the full text is summarized and the research direction is further prospected.