The Study On Nonlinear Intelligent Control For Nonlinear Systems With Stochastic Disturbances And Input-Output Constraints

In the traditional linear constant-invariant control system,the controlled object is considered to be linear when the system problem is analyzed and the control method is designed.But in reality,the ideal linear controlled object does not exist.For example,the controlled system has unknown uncertainty,random interference and so on.At the same time,the final control performance of the system is affected not only by the nonlinear effects of the controlled object,but also by the characteristics of various physical devices such as controllers and actuators in the forward path and sensors in the feedback path,as well as the network resources between the components of the system.First of all,due to the existence of external working conditions,engineering imprecise mathematical modeling,the actual industrial system has different degrees of nonlinear and unpredictable irresistible characteristics.Secondly,in the forward path and feedback path of the system,due to the actual component characteristics,there are often non-smooth uncertain constraint characteristics.When the control signal passes through these unfavorable links,unexpected unknown changes will inevitably occur.It may lead to the degradation of the control performance of the system and even lead to instability.Finally,communication between components is limited by network bandwidth,and due to resource constraints,actuator failures and constraints,direct control of failure compensation becomes difficult and challenging.Therefore,this paper,taking backstepping technology as the framework,designs intelligent control strategy for the nonlinear systems with random perturbations.The purpose is to improve the control performance of the nonlinear system,reduce the network bandwidth share of the control method,realize the effective compensation of actuator failure and reduce the influence of constraints on the controlled system.This article is divided into five chapters.The first chapter sumrnarizes the research status of nonlinear control methods,the research on input and output constraint control strategies such as actuator failure,and the research status of intelligent network control.From the second chapter to the fifth chapter,it is the four main research contents of this paper.In chapter 2,we discuss that hysteresis is a common phenomenon in actuators,which can cause hysteresis in the control input of the system.In addition,actuators will inevitably fail in actual operation.Hysteresis effect and actuator failure will weaken the transient tracking performance of the system and even make the system unstable.Through the study of stochastic nonlinear uncertain systems,it is found that how to guarantee the given transient tracking performance remains to be solved when actuator failure and input hysteresis are considered simultaneously.In this paper,fuzzy control method was used to solve the compensation problem of actuator failure and input hysteresis.Combining backstepping technology and fuzzy adaptive control method,the control method was designed for the above problems.Furthermore,all signals can be proved to be bounded and the tracking error can be kept within the set range with input hysteresis and actuator failure.In chapter 3,an adaptive control method for event-triggered is proposed for nonlinear systems with actuator failure and output dead zone.It is a difficult task and challenge to design a compensation controller for uncertain stochastic nonlinear systems.In order to avoid the influence of the nonlinear characteristics of the system on the output,this paper proposes a method which combines the neural network with Nussbaum function to solve the problem.Based on the technique of backtracking Lyapunov function,this method is established to ensure that tracking error constraints can be provided.In addition,the system also has the transmission resource constraints and actuator failure problems,which is a great challenge to the design of the system control method.In case of actuator failure,the system needs more transmission resources.Howevers the transmission resources of the system are limited,and this requirement cannot be satisfied.In addition,how to ensure the tracking performance of the system is also a difficulty and challenge.Using the event-triggered controller and Lyapunov method,a new optimization algorithm is proposed to ensure the stability of the closed-loop system and the convergence of tracking errors.The fourth chapter,aiming at uncertain systems,took backstepping as the technical framework,designed the intelligent control method with event-triggered strategy.Considering the existence of the uncertain part of the system,the neural network is used to approximate it,and the time-varying approximation error is integrated into the design of the approximation system.In addition,in order to better deal with the problem of large dimension of weight vector in neural network,this chapter proposes an adaptive control method.At the same time,aiming at the problem of limited network resources,this paper put forward an effective event-triggered control method combining backstepping technology,which could effectively balance the tracing performance and network resources of the system.According to the above analysis,the intelligent control method in this chapter can ensure the tracking effect of the system.In chapter 5,an extended dimension fuzzy adaptive control scheme is proposed to solve the actuator failure problem and the approximation error problem of intelligent control.In the real physical system,random perturbation is common,and the problem will destroy the transient tracking performance of the system,and even the controlled system.In this section,the influence of time-varying approximation error is considered in the control scheme of stochastic nonlinear system,and the influence of time-varying approximation error on actuator failure is also considered.By further processing the vector norm of the fuzzy logic system,the computation of the control method can be greatly reduced.However,the symbolic function presents a challenging chattering problem.In order to stabilize stochastic nonlinear uncertain systems with random perturbations and actuator failures,backstepping technique is combined with fuzzy logic system technique.The new control method proposed in this paper can guarantee the asymptotic convergence and the predetermined transient tracking error of the system in the case of actuator failure.