Intelligent Adaptive Control Strategy Research For Nonlinear Systems With State Constraints

Most of practical industrial physical systems contain many nonlinearities and uncertainties,which make the control and analysis become complicated.All kinds of constraints are ubiquitous in the practical nonlinear systems because of surroundings,safety specifications and external factors.These constraints would influence the system stability and lead to control performance decrease.A reasonable approach is to consider the state constraints in the problem formulation,and then design an effective controller which ensures that the constraints are satisfied.This thesis investigates the control problem for three classes of nonlinear systems with full state constraints,such as uncertain nonlinear systems,switched nonlinear systems and stochastic nonlinear systems.Based on the adaptive backstepping technique and barrier Lyapunov function,we study the intelligent adaptive control problem for nonlinear systems with state constraints associating with the approximated neural networks and fuzzy logic systems.The main contents are summarized as follows:1.For the nonlinear systems with full state constraints,the adaptive fuzzy control law,the adaptive fuzzy finite-time control law and the adaptive fuzzy asymptotic tracking control law are designed based on the barrier Lyapunov function and the adaptive backstepping technique.Firstly,an event-triggered adaptive fuzzy control law is designed for the strict feedback nonlinear systems with state constraints and time-delay.The barrier Lyapunov functions are utilized to address the state constraints and the time-delay is compensated by choosing the appropriate Lyapunov-Krasovskii functional.To reduce the transmission data,the event-triggered mechanism is integrated into the adaptive backstepping design framework.The designed control law can guarantee semi-global uniform ultimate boundedness of all the signals in the closed-loop system.Secondly,an adaptive fuzzy finite-time control law is designed for nonstrict feedback nonlinear systems with state constraints.The property of fuzzy basis function is used to address the nonstrict feedback structure,which avoids the algebraic loop problem.The designed adaptive fuzzy finite-time controller can make system states maintain the predefined constraints and all error signals of the closed-loop system converge into a small neighborhood of the zero in a finite time.Finally,an adaptive fuzzy asymptotic tracking control law is designed for multi-input and multi-output nonlinear systems with state constraints.By introducing some smooth functions and positive time-varying integral functions into the backstepping design procedure,an adaptive fuzzy asymptotic control law is recursively constructed.The asymptotic tracking control effect is achieved without violating the state constraints.2.For the switched nonlinear systems with full state constraints,the adaptive fuzzy uniformly ultimately bounded control law,the adaptive fuzzy event-triggered control law and the adaptive fuzzy asymptotic tracking control law are designed in the adaptive backstepping design framework.Firstly,an adaptive fuzzy control strategy is designed for the switched nonlinear systems with state constraints and time-delay.By constructing the suitable Lyapunov-Krasovskii functional,the adverse effect of time-delay is effecively compensated.The fuzzy logic systems are explored to approximate the unknown dynamics,and then the adaptive fuzzy tracking control law is constructed based on the adaptive backstepping technique.Combining the Lyapunov theory with average dwell time method,it is shown that all signals of the whole closed-loop systems are bounded under switching signals and the predefined constraints are not violated.Secondly,the event-triggered mechanism is integrated into the controller design process,and an adaptive fuzzy event-triggered control strategy is presented for switched nonlinear time-delay systems.The stability of the closed-loop systems is ensured by applying the common Lyapunov function.Finally,an adaptive fuzzy asymptotic tracking control law is presented for switched nonlinear systems with full state constraints.The state constraints are addressed with the aid of the barrier Lyapunov functions.Moreover,drawing support from some smooth functions and positive time-varying integral functions,the adaptive fuzzy asymptotic tracking control law is recursively constructed.Through Lyapunov analysis and the common Lyapunov function method,it is shown that the system output tracking error asymptotic converges to zero and the closed-loop system is stable.3.For the stochastic nonlinear systems with full state constraints,the adaptive neural network output feedback control law,the adaptive neural network uniformly ultimately bounded control law and the adaptive neural network asymptotic tracking control law are designed based on the quartic barrier Lyapunov functions and adaptive backstepping technique.Firstly,an adaptive neural network output feedback control scheme is designed for the stochastic nonlinear systems with unmeasurable states.The neural network state observer is established to estimate the unmeasurable states.In the adaptive backstepping design framework,an adaptive neural network output feedback control strategy is devised to address the state constraints by using quartic barrier Lyapunov functions.The presented control law can guarantee all states maintain the predefined scopes and that all signals of the closed-loop systems are semi-global uniform ultimate bounded in probability.Secondly,considering the stochastic nonlinear systems with state constraints and time-varying delays,an event-triggered adaptive neural network control law is designed by combining the event-triggered mechanism with adaptive backstepping technique.The Lyapunov-Krasovskii functional is employed to compensate the adverse effect originating from time-varying delays and the barrier Lyapunov functions are used to make all states maintain the prescribed scopes.The neural networks are employed to approximate the nonlinear functions.The constructed event-triggered controller not only can effectively offset the influence coming from time-varying delays and state constraints,but also reduce the transmission events.The system stability is analyzed by applying Lyapunov theory.Thirdly,an adaptive neural network uniformly ultimately bounded control law is designed for stochastic nonlinear systems with time-varying state constraints based on the backstepping technique.The time-varying barrier Lyapunov functions are utilized to address the time-varying state constraints,and the neural networks are utilized to approximate the unknown combinational nonlinear functions.Finally,an adaptive neural network asymptotic tracking scheme is presented for the stochastic nonlinear systems with full state constraints and unknown virtual control coefficients.Some smooth functions and bounded estimation method are used to construct the asymptotic tracking controller.By skillfully constructing the Lyapunov function and several useful inequalities,the stability and asymptotic tracking performance can be ensured.