Research On Path Planning Algoritms Of Unmanned Surface Vehicle

Path plannin is the key technology of intelligent robots,it is also a solving process of strategic problem.Unmanned Surface Vehicles(USVs) must be capable of a relatively high degree of autonomous navigation in highly dymamic and unexpected marine environment,so efficient,realiable and real-time obstacle avoidance path planning is very important.This thesis mainly researches on global path planning which is on static environmental informations and local path planing obeying marine rules in pre-defined navigation ways and dynamic obstacle enviroments.The main work accomplished concretely follows:a brief overview of the research status at home and abroad and the algorithms of robot path planning are first given.And then,the motion characteristics of the USV and the International Marine Collision Regulations are also introduced.After that,the real-time dynamic obstacle avoidance of the USV is focused on.In this dissertation,at first the global waypoints is got by A Star search algorithm which is can find an optimal solution in a short amount of time.Then in the process of the USV sailing along the course of the pre-defined trajectory,according to real-time sensor data,using fusion of the marine rules of velocity obstacles method to modify the trajectory,at the same time taking into account the uncertainty of the obstacle movement and the data uncertainty that caused by USV self-positioning systems in process of movement and real-time detective sensor systems'accuracy,error,drift,models the obtained obstacle information uncertainty and combines the process of prediction and velocity obstacles.This dissertation focuses on the discussion and verification for real-time obstacle avoidance strategy,an USV path planning platform is built by simulation,realizes real-time static or dynamic obstacle avoidance simulation in the process of sailing along the intended routes based on Marine rules,it verifies the feasibility and effectiveness of above approaches.