Research On Nonlinear Robust Control Of Underactuated Surface Vessels Motion

The underactuated vessel motion system is a very complex nonlinear system, includingmulti-input, multi-output, strong coupling and Nonholonomic. Meanwhile, due totime-varying disturbances from marine environment, the controller is designed byconventional single nonlinear control method can not have the characteristics, which arestrong robustness, adaptability and higher control accuracy. Thereby, it naturally become ahotspot problem that using nonlinear control theory to solve the above problems. With thedevelopment of further research on the underactuated vessel and marine economic strategy,multi-vessel formation control comes into being. Formation control has a wide applicationprospect and space, such as detection of marine resources, fleet coordinated operations,offshore supply, search rescue, fleet patrol, environmental monitoring and other aspects. Tothis end, the paper takes the underactuated vessel as research object, and mainly completes thefollowing aspects:Firstly, to beginning with, three degrees of freedom of ship motion mathematical modelis established, which provide a basic simulation platform to study the characteristics ofclosed-loop system. The control performance of the control system can be assessed though thesimulation research. Then, the basic knowledge, including stability theory, backsteppingmethod, sliding-mode theory and neural network technology is introduced, which lay thefoundation for the subsequent chapters.Secondly, when conventional backstepping method is used to design the trajectorytracking controller of the underactuated vessel, there is a shortcoming of excessive calculation,as it involves analytical and numerical differentiation of the virtual control laws. In order toovercome the problem and realize tracking control for underactuated vessels with parameteruncertainties and external disturbances, a command filter method combined with sliding modemethod is proposed. The tracking error equations are established based on command filtermethod, which puts tracking control into stabilize surge speed and course angle error usingtransformation equation. Then, a nonlinear sliding mode controller is designed based onintegration ideas, without estimating uncertainties of underactuated vessel model and externaldisturbances. The problem of analytical and numerical differentiation of the virtual controllaws is overcome, chattering of control input is circumvented, and the static error andovershoot are decreased. The results of simulation experiments indicate that the proposed controller can effectively implement trajectory tracking of line and curve in the horizontalplane.Thirdly, in order to realize path following control for underactuated vessels underparameter perturbation and external disturbances, a neural sliding mode controller based onvisible distance is presented. The desired position and the current position are used todetermine the reference surge velocity and yawing angle by using visible distance, which areemployed as the virtual control laws in stabilizing the position errors. Then, a new nonlinearfeedback robust controller is designed based on neural network and sliding modes to realizethe tracking control of the virtual control laws without determining upper bound of parameterperturbation and external disturbances. The problem of saturation and chattering of thecontroller are circumvented. The results of simulation experiments indicate that the controlleris robust against parameter perturbation and external disturbances. Moreover, the pathfollowing control with can be achieved by the proposed control method.At last, we further study underactuated vessel formation control problem underparameter perturbation and external disturbances. To beginning with, for leader-followerformation model, considering that the follower can only take advantage of position andheading angle information, the virtual vessel is used, which is designed according to leader’sposition and heading angle and fleet formation information. The motion trajectory of thevirtual vessel converges to position of reference vessel within a limited time, which isregarded as desired trajectory and speed of the follower. Then, the fleet formation controlmodel under leader-follower is established based on the virtual vessel strategy. At last, avessel formation neural sliding mode robust controller is proposed by combiningbackstepping method, neural network technology and sliding-mode control theory underparameter perturbation and external disturbances. Extensive simulations are provided todemonstrate the effectiveness of the presented vessel formation controller.The research results of the dissertation have some academic value and a wide range ofengineering application prospects. Some of the results can also be applied to underwatervehicle, nonholonomic mobile robots and other systems possessing underactuatedcharacteristics, and have good universality.