Nonlinear Motion Control Methods Of Unmanned Surface Vehicle

The ocean is the second space for human survival. Recently, the exploration andexploitation of marine resources and maritime security have been paid more and moreattention. Unmanned surface vehicles （USV） have become an intense research area becauseof their extensive military and civil applications. As an unmanned intelligent vehicle, USVinvolves the hydromechanics, structural mechanics, ship design, system engineering,computer science, automatic control and many other areas. Therefore, the research on USVhas both important theoretical significance and practical value. Under this background, themotion control of the USV is researched in detailed in this thesis.For the “XL” USV motion control system design, system architecture, speed/directioncontrol （actuated control problem）, filter measurements are mainly discussed. In order tomeet the USV autonomous operation request in the complex ocean environment, the USVintelligent control system architecture is proposed based on the cerebrum basic functioncombination zone theory. Under this framework, the motion control system for “XL” USV isproposed. Considering the USV with nonlinear, time-delay, uncertainty and marineenvironment disturbance, an S surface speed/direction controller independent of precisemathematic models is introduced in the design for “XL” USV controller. To solve the "rudderangle drift" problem of steering pump system, an improved S surface controller based on theactuator compensation is proposed. Many sea tests for “XL” USV demonstrate theeffectiveness and robustness of the above proposed motion control system.For the USV underactuated control, aiming at the stabilization, path-tracking andtrajectory-tracking problems, three nonlinear control strategies are proposed respectively, andnumerical simulation comparison experiments are presented on a USV mathematics model.By combining the Backstepping method with sliding mode control theory, a new methodnamed Backstepping adaptive dynamic sliding mode control （BADSMC） for a class ofsingle input-output uncertain nonlinear systems is constructed for the follow-up controllerdesign. Moreover, an adaptive artificial fish swarm algorithm （AAFSA） is proposed for thecontrol parameter optimization problem. Simulation results demonstrate that the proposedoptimization algorithm could resolve the difficulty of the parameter regulation for underactuated USV.The Stabilization control problem for underactuated USV is addressed. Based on cascadesystem analysis method, the stabilization control problem of resulting cascade system can bereduced to the stabilization problem of cascade subsystem. By drawing lessons from transform method, a smooth time-varying state feedback control law is developed byintroducing an assistant exponentially tapered control item into the feedback control law andtheoretically proved to render the original system globally K exponential asymptoticallystable.The path-following problem with model parameter uncertainties and externaldisturbances for underactuated USV is addressed. By introducing the Serret-Frenet frame, thesingle-input ternary-output system is transformed into an equivalent single-input two-outputsystem under some assumed conditions, which simplifies the controller design. ABackstepping adaptive dynamic sliding mode path-following controller is proposed based onBADSMC method. Theoretical analysis proves that the proposed controller can render theoriginal system global stable.The trajectory-tracking problem with model parameter uncertainties and externaldisturbances for underactuated USV is addressed. This paper assumes that the referencetrajectory is generated by a virtual USV, and then the original system is transformed into anequivalent tracking error system. Therefore, the tracking and regulation problem ofunderactuated USV is equivalent to stabilizing the tracking error system. By borrowing ideasfrom BADSMC method, a Backstepping adaptive dynamic sliding mode trajectory-trackingcontroller is proposed. Theoretical analysis proves that the solutions of original system areultimately bounded using the proposed controller. Accordingly, this can meet the request fortrajectory-tracking control.Numerical simulation comparison experiments on a USV mathematics model fordifferent cases are presented. Simulation results demonstrate the effectiveness, robustness andgood control performance of the above proposed control strategies.