Active Disturbance Rejection Control Based On RBF Neural Network For Unmanned Surface Vessel

With the rapid development of the exploitation and utilization of marine resources and marine environment,many countries in the world have strengthened the development,design and manufacture of marine equipment to safeguard their marine rights.Unmanned Surface Vehicle(USV),as a kind of intelligent marine equipment that can be used for monitoring marine environment and safeguarding marine rights has attracted more and more attention.In the extremely complex marine environment,the design of an autonomous motion controller for USV is the premise to ensure that USV can safely navigate and perform specific tasks,which enables USV to track the desired course and track quickly and accurately.Therefore,this paper takes the autonomous motion control of the 'Lanxin' USV as the research object.Aiming at the problem of the precise tracking control of the course and track of the 'Lanxin'USV,an ADRC based on RBF neural network for USV is designed.The feasibility and validity of the designed controller are verified by simulation and ship experiments.The main contents are as follows:Aiming at the problem of the 'Lanxin' USV motion model and disturbance model,the linear mathematical model and the non-linear response mathematical model of 'Lanxin' USV are established respectively,and the disturbance of external wind,wave and current is modeled mathematically,which provides a model basis for course control and path tracking control.Aiming at the course control problem of the 'Lanxin' USV,an ADRC course controller based on RBF neural network is designed by using the advantages of self-learning and self-adapting ability of RBF neural network,and the adjustment method of parameters is given.The simulation experiments show that the ADRC course control based on RBF neural network can adjust the parameters of the controller on-line under the disturbance of wind,wave and current,and has stronger anti-interference ability and better control effect.Aiming at the path tracking control problem of the 'Lanxin' USV,combining LOS algorithm with ADRC heading control based on RBF neural network,a RBF neural network ADRC path tracking controller based on LOS is designed,and the simulation experiments of straight line and curve path tracking control are carried out respectively.The simulation experiments show that the RBF neural network ADRC path tracking control based on LOS has higher tracking accuracy and better control performance under the disturbance of wind,wave and current.Finally,the designed course controller and track controller in this paper are verified by ship.The designed controller is applied to the motion control of the 'Lanxin' USV,and the ship experiments of course and path tracking control are carried out at sea.Through the analysis of the experimental data,it can be seen that the designed controller has good robustness and adaptive ability under the disturbance of external environment,and can realize the precise tracking control of the desired course and track.It provides an effective precise tracking control method for the course and path tracking control of USV.