Backstepping Switching Control Of Nonlinear Systems

Switching scheme in the Control Science and Engineering has attract much attention of control community in the past decades.With the introduction of the switching scheme,the identification control,the backstepping control,the adaptive control,the intelligent control and so on are all combined and embedded with the switching scheme.It promotes the control theory research to enrich the control scheme,expand the control method,increase the control accuracy,improve control performance.From a viewpoint of practical application,with the development of the control technology and the requirement of the higher and higher precision control,the requirement leads to the complex nonlinear structure to described the practice nonlinear process,so the single system model is incapable of covering the whole system operation procedure.In order to make up the gap of the single mode,multiple subsystem models are combined together to describing a practice system and further the switching signal is usually unpredictable.The arbitrary switching systems is a kind of Hybrid system.In order to improve the control performance,several sub-controller connected with switching logic in introduced.The switching logic is designed to embed in the control system to regulate distinct sub-controllers or overcome the unknown parameters in nonlinear system.How to systematic analysis and synthesis the switching mechanism has become an important issue to be solved urgently.Focusing on the effect and impact of switching mechanism in the research of system control science,and combing the backstepping control method,this thesis analysis the switching problem systematically and comprehensively in system control theory,including arbitrarily switching of systems,saturated switching of controllers,adaptive switching of parameters and servo switching of neural network.1.In Chapter 2,this paper begins with the research of the nonlinear switching system.First part of the research is the linearized nonlinear system,the linearized nonlinear system realise the closed-loop nonlinear system control combining identification switching control based on the dynamical switching model subbank.In order to overcome the drawback of nonlinear characteristic missing,the backstepping control algorithm in second part is introduced to control the nonlinearized nonlinear system,among which the control of the switching complex nonlinear systems is an obstacle.The nonlinear system is composed of several arbitrarily switching complex subsystems.A states backstepping controller for switching complex nonlinear system is designed based on an uniform nonlinear decoupling method first.Then based on an uniform gain observer,an output chained backstepping controller is construct to stabilise the observer-backstepping control system.2.In Chapter 3,in the face of uncertain nonlinear system control,due to the uncertainty of the nonlinear system structure,a single controller is difficult to meet all the control requirements and control accuracy.Therefore,it is necessary to design distinguish controllers for a nonlinear system according to different requirements and working environments.Sliding mode variable structure controller is an effective switching controller,mainly used for robust control of nonlinear systems with matched and unmatched uncertain structures.Based on the switching of saturation function,the dynamic controller overcomes the phenomenon of chatting in the design of discontinuous controller and improve the control quality.First a backstepping variable structure switching controller is designed for nonlinear systems with bounded uncertainty and unbounded uncertainty.Then backstepping variable structure adaptive controller is designed to realize dynamic adaptive identification and switching control online.Finally,the command filter is introduced to make up the drawback of explosion of differential terms in backstepping design,and realize the backstepping variable structure switching control.3.In Chpater 4,a global adaptive feedback control of nonlinear systems with unknown parameters entering nonlinearly is proposed.Such unknown parameters are also not required to lie in a known compact set.The adaptive controller is a new double dynamical switching-type controller in which the controller parameter is tuned in a flexible switching manner via a monotonically decreasing switching logic and the controller combines the traditional adaptive theorem with the switching scheme perfectly.Further,an adaptive backstepping flexible switching control scheme is proposed to stabilize a class of mixed switched nonseparated parameterized nonlinear systems.First,the switched subsystems are all parameterized and the nonsepareted parameters are all nonlinearity entering.Next,an uniform backstepping controller under arbitrary switching signal is designed to stabilize the nonlinear systems.Then,a flexible switching mechanism is proposed to inject into the controller to jumping compensate the backstepping controller.4.In Chapter 5,owing to the tracking control problem of the switched uncertain nonlinear systems under arbitrary switchings is hard to realise,the neural network approximator is embedded in the backstepping controller,the neural network is tuning adaptively with the switched subsystem dynamically,realise the states reconstruct of the switched system,form the closed-loop feedback control system.Multilayer neural networks are used as a tool for modeling nonlinear functions up to a small error tolerance.In order to improve the tracking performance and compensate the uncertainty from the arbitrary switching,through the expansion of the traditional neural network controller design,a multilayer neural network adaptive controller with multilayer weight norm adaptive estimation has been designed.Further,the weight norm adaptive laws are not only used to approximate the first layer network but also every layer.The adaptive updated laws of the controller have been derived from the Lyapunov function method,and the adaptive neural network control schemes have been developed to achieve better performance.5.In Chapter 6,the Multilayer-Neural-Network Pre-Observer with the Error Compensator and Command-Filter-Based Adaptive Backstepping Output feedback control is proposed for a class of strict-feedback switched nonlinear system with nonseparated arbitrarily switching parameter.First,a multilayer neural networks observer is developed to estimate the immeasurable states.Next,observer error compensate mechanism is designed to compensate the state estimate error.Then command-filtered backstepping control with compensating signals is designed to realise the reference signal tracking.The proposed method guarantees that all signals in the observe-compensate closed-loop control system are bounded.The main contribution of this paper are the multilayer neural networks observer with compensation.The observer can estimate the real-time states of the nonseparated arbitrarily switching parameterized nonlinear system.The backstepping controller is proposed to combine the command-filtered backstepping controller to overcome the system linearity and explosion of complexity in backstepping design methods.The above content is the whole research contents of doctoral dissertation,the conclusion and perspective are given in the end of the paper.