Research On Nonlinear Adaptive Control For Dynamic Positioning Ships

This dissertation addresses the control problem of a dynamic positioning ship. In the presence of unkown disturbances and model parameter uncertainties, a series of state feedback/output feedback studies is carried out by the vectorial backstepping method, dynamic surface control technology, adaptive control approaches and the theory of neural networks. Morever, the ships manoeuvre and propulsion simulation system is developed by the author and other participators. The following research works have been completed in this dissertation:Firstly, we consider the problem of trajectory tracking control for a dynamic positioning ship with unknown time-variant environmental disturbances. The adopted mathematicalmodel of the surface ship movement includes the Coriolis and centripetal matrix and the nonlinear damping terms. An adaptive observer is constructed to provide an estimation of unknown disturbances and is applied to design a novel trajectory tracking robust control law through the vectorial backstepping technique. It is proved that the designed tracking control law can force the ship to track the arbitrary reference trajectory and guarantee that all signals of the closed-loop trajectory tracking control system of ships are uniformly ultimately bounded by Lyapunov function and simulation studties.Secondly, adaptive robust control laws are proposed for a dynamic position ship combining dynamic surface control into vectorial backstepping in the presence of unknown bounded time-variant environmental disturbances and konwon ship parameters. The proposed nonlinear control laws utilize the differentiation of the first-order low-pass filter to replace the differentiation of virtual control vector due to introduction of dynamic surface control technique. As a result, the differentiation operations in the control law design are replaced by simple algebraic operations. Hence, the control laws are easily implemented in engineering practice. Leakage terms based on a variation of σ-modification and the high gain observer are incorporated into the state feedback and output feedback control laws respectively. The stability analysis of the closed-loop system is given by means of Lyapunov function and simulation results on a supply ship are presented to validate the effectiveness of the proposed strategies.Thirdly, an adaptive state feedback control law is presented using vectorial backstepping and the neural network (NN) approximator for a dynamic positioing ship with unkown external disturbances and ship parameters. Morever, a robust NN-based adaptive output feedback control scheme is proposed for a dynamic positioning ship by employing the theory of neural networks, observer and dynamic surface control tehnology. It is proved that the proposed control strategy guarante the designed closed-loop dynamic positioning system is uniformly ultimately bounded by means of Lyapunov function. Simulation results on a dynamic positioning ship illustrate the effectiveness of the neural netwok approximator and the control law can force the ship to the desired position.Fourthly, the ship manoeuvre and propulsion simulation system is developed by the author and other participators by CAN-bus, Ethernet technology and VC++programming technology combing the research work of ship motion control. Experimental results show that the system is stable and reliable which can be adopted not only as the teaching system for demonstration but also as the platform for studying advanced control algorithm for ships and integrated bridge system of marine surface vessels. What’s more, it is available as experiment environment for reliability test for ship control systems and has several advantages such as reducing costs, shortening study period, improveing quality of the system.