Output Feedback Finite-Time Control Of State-Constrained Nonlinear System

Scientific and technological progress and industrial development have promoted the wide application of nonlinear systems in industrial production,which in turn has promoted the increasing demand for nonlinear systems control algorithms,and it is becoming more and more important to make breakthroughs in control technology on the existing basis.Considering that the state of an industrial system is constrained in the actual working environment,and due to the complex industrial environment,the nonlinear system state cannot be directly observed,it is unrealistic to use state feedback.Therefore,this paper deeply studies the output feedback finite time control algorithm for nonlinear systems with constrained state,and generalizes the control strategy to the application of robotic arm systems and nonlinear multi-agent systems.The main contents of the thesis research are summarized as:1.For high-order nonlinear systems with state constraints and input saturation,this paper studies the output feedback finite-time adaptive control of nonlinear systems on the basis of command filter backstepping method and barrier Lyapunov function,introduces a command filter to solve the computational explosion problem caused by the backstepping process,and constructs an error compensation mechanism to eliminate the filtering error.Using the adaptive technology based on radial basis neural network to approximate the nonlinear dynamics of the system,the neural network state observer is designed to reconstruct the state of the high-order nonlinear system by collecting output signals.And the fractional power function is introduced so that the output of the nonlinear system tracks the desired signal with good accuracy in a finite-time,and the signals in the closed-loop system are bounded,and all states will not exceed the expected constraints.Finally,the control effect of the scheme is verified by simulation.2.For a class of multi-link manipulator systems with unmeasurable states and constrained states,an adaptive finite time control scheme based on neural network state observer and barrier Lyapunov function is proposed.The finite time command filter is introduced to avoid the direct derivation of the virtual control signal,and an error compensation mechanism is constructed to eliminate the filtering error and further improve the control effect.Considering that the state of the robotic arm system is unobservable in some actual working environments,this paper uses the output feedback control scheme to reconstruct the state of the robotic arm system by using the output signal as feedback.The tracking error of the robotic arm system converges to a small neighborhood near the origin for a finite-time.The effectiveness of the control scheme is verified by simulation.3.For nonlinear multi-agent systems with state-constrained and input saturation,a distributed finite-time adaptive collaborative control scheme based on neural network state observer is designed.The finite time command filter is used to solve the problem of differential explosion,and the error compensation mechanism is further constructed to weaken the adverse effects of the filter error.Neural network techniques can be used to approximate unknown nonlinear dynamics in a system.This control approach ensures that the consistent tracking error of the multi-agent system converges to desired region of the origin in finite time,and ensuring that all states of the system are within the desired constraints.Finally,the feasibility of the proposed algorithm is confirmed in the form of Matlab/Simulink simulation.