Research On The Control Method Of Nonlinear Switched System With Input Hysteresis

Switched nonlinear systems are an important class of hybrid systems in the current industrial production,which are widely used in important fields such as network control systems,unmanned vehicles,and robotic systems.In designing control schemes for controlled objects,external disturbances,hysteresis nonlinearity,deadband nonlinearity and other factors should be taken into account.If these obstacles are ignored in the design of the controller or in the analysis of stability,the system performance can be seriously degraded and even lead to system instability.In order to overcome these problems,this thesis investigates the control problems of several types of switched nonlinear systems by using Backstepping technique as the control design framework and combining Lyapunov stability theory,switched with mean dwell time,command filtering and Nussbaum function.The specific work studied for this thesis contains as follows:1?We investigate the adaptive neural control problem for a class of uncertain switched nonlinear systems with dwell time.The system contains unknown nonlinear functions,actuator hysteresis,and external disturbances.We design a control scheme using a hysteresis quantizer containing a filter to filter out the high frequency components,while the frequency of the input signal above the boundary causes the hysteresis loop to deform.The adaptive neural network control strategy incorporating in the Backstepping technique framework and it is used to eliminate external disturbances and approximate the unknown nonlinear function.This strategy ensures that the tracking error converges to an adjustable region near zero.2?We investigate switched nonlinear systems with switching hysteresis with output hysteresis.Ulike 1,this section focuses on the output hysteresis,which is more complex than the input hysteresis.The output hysteresis nonlinearity under consideration is captured by the Bouc-Wen model,which leads to an unknown time-varying gain,which can pose a significant challenge to our control design.To overcome this problem,we developed an adaptive neural optimization control scheme.In addition,the Nussbaum-gain technique is applied to deal with the problem of unknown control gains imposed on the optimal controller.Finally,the design scheme ensures the stability of the nonlinear switched system.3?We investigate the problem of adaptive fuzzy optimal control of nonlinear switched systems with dead zone inputs.In this section,the fuzzy logic systems are used to construct a actor-critic architecture in which the actor performs control actions and the critic evaluates the performance of the system.Subsequently,the optimal control design is combined with Backstepping techniques to accomplish the control goals through step-by-step feedback and adjustment of the actor-critic architecture.Given the complexity of the optimal controller design,the introduction of instruction filters can effectively address this problem.Our proposed design scheme can guarantee the stability of the nonlinear switched system,which is well demonstrated in the simulation results.