Intelligent Trajectory Tracking Control Of An Unmanned Surface Vehicle

Unmanned surface vehicle(USV)is a kind of unmanned platforms which can tackle a variety of complex tasks by carrying different military or civilian payloads.Due to the increasing highlights on maritime rights and interests,USV has gradually attracted extensive attention and research from governments and relevant research institutions.Trajectory tracking is a pivotal control scenario of USV motion control,and it has a critical impact on whether it can be fast,effective and accurate in performing the corresponding missions.Therefore,it is of great significance to research the trajectory tracking control problem to promote the follow-up technology development of USV.In this thesis,combined with extreme learning neural networks,finite-time disturbance observer and finite-time extended state observer,the intelligent trajectory tracking control strategy has been proposed to strengthen the robustness and tracking performance of the electric propulsion USV motion control system.The main contributions of this thesis are as follows:Firstly,a hybrid feedforward-feedback trajectory tracking strategy is proposed for an USV which disturbed by external environments such as wind,waves and currents.By adopting a hybrid feedforward-feedback control structure,only reference signals are needed as the approximator inputs,which significantly reduce the inputs dimensions.At the same time,the computational complexity is further simplified by using the extreme learning neural networks as the feedforward approximator.The tracking errors are uniformly ultimately bounded and simulation results verify the effectiveness of the proposed control strategy.Then,considering the trajectory tracking control problem of an USV in the presence of uncertainties and external disturbances.The control scheme is designed based on nonsingular terminal sliding mode technology,extreme learning neural networks and finite-time disturbance observer.Extreme learning neural networks are employed to approximate the uncertainties of the system while the finite-time disturbance observer is applied to observe the minimum approximation errors and external disturbances.The simulation results demonstrate the effectiveness of the proposed control strategy.Finally,a finite-time extended observer(FESO)based trajectory tracking control scheme is proposed for exactly tracking an USV with complex unknows to a reference trajectory.By virtue of nonsmooth analysis,the FESO approach is devised and is incorporated into the nonsingular fast terminal sliding mode control framework,and thereby further enhancing disturbance rejection and tracking accuracy.Moreover,global finite-time stability of the entire FESO-based exact tracking control(FESO-ETC)strategy is derived from rigorously theoretical analysis.Simulation studies and comparisons demonstrate that the proposed FESO-ETC approach can achieve exact trajectory tracking in the presence of complex unknowns.