| The rapid development of global business has promoted the rapid development of the shipbuilding industry,and the vessels are developing towards large scale.Due to the shortcomings of large vessels,such as large inertia,unstable navigation control,and difficult mooring operations,large vessels are prone to accidents during the berthing process in the port,thus the automatic berthing technology of vessels has become a hot research issue in this field.In view of the problems of low speed path following control effect,large energy consumption and stable control outside the berth during automatic ship berthing,this thesis focuses on two key points: track keeping control at low speed and external stabilization control of berth without displacement overshoot.Based on the research of these two control processes,the main work done in this thesis includes:With the help of the rich research foundations of the predecessors,this thesis established a three-degree-of-freedom mathematical motion model and interference model for large vessels based on the MMG separation modeling idea,which provides a model basis for subsequent controller designs.Aiming at the problem of poor rudder efficiency and real ship maneuverability during low-speed path tracking in the harbor,the idea of indirect track control was applied,and the track control was divided into a guidance link,a course control link and a rudder angle control link.In terms of guidance,a line of sight(LOS)guidance algorithm is used.To solve the problem that the guidance algorithm converges slowly to the target track when the lateral error is large and the tracking effect is poor at large corners,the LOS algorithm is implemented.In terms of heading control,a single neuron adaptive PID heading controller is designed to solve the problems of large rudder angle amplitude and high steering frequency at low speed.Finally,the adaptive LOS guidance algorithm and the single neuron adaptive PID heading control are combined into a track controller.The simulation results show that when tracking at low speed,the path following controller can guide the ship to quickly converge to the target path and face different turning angles.Both can smoothly complete the steering task,and the rudder angle amplitude is smaller and the steering frequency is smaller during the following process.In order to solve the problem of final sedation in automatic berthing,artificial neural network(ANN,Artificial Neural Network)and strong non-linear function mapping ability can be used to replicate the automatic berthing operation of experienced captains.A simple docking simulation platform is built by the established mathematical model of the ship’s motion and GUI(Graphical User Interface)controls.Sample data is generated by automatic docking on the simulation platform,and then three-layer BP neural network is used for parameter training.The trained ANN is used to perform docking experiments in different initial states.The experimental results show that the controller achieves a good control effect and meets the requirements of actual docking.All simulation experiments in this thesis are implemented with MATLAB programming.The simulation results show the effectiveness of the controller design.This research has important practical significance for promoting the development of automatic berthing control of vessels,and has laid a theoretical foundation for the construction of safe and energy-efficient marine transportation. |
