基于DDPG制导的无人帆船路径跟踪控制OA

Aiming at the problem that the traditional guidance approach does not consider the tacking and gybing maneuvering characteristics of the unmanned sailboat,and cannot complete the path following task under actual ocean wind conditions,a DDPG-PID path following approach for the unmanned sailboat based on the Deep Deterministic Policy Gradient(DDPG)guidance is proposed.The motion state and position information of the unmanned sailboat are taken as the inputs of the guidance algorithm,and the policy network of the DDPG algorithm is utilized to output the reference heading angle of the unmanned sailboat.The heading controller is designed by using the PID algorithm to realize the effective tracking of the actual heading of the unmanned sailboat to the reference.Considering the maneuvering characteristics of the unmanned sailboat,a reward function is designed and the neural networks are iteratively trained to maximize the reward value of each round.Finally,through simulation comparison experiments,the effectiveness and reliability of the proposed DDPG-PID algorithm in the reference path tracking task of unmanned sailboats are verified.