Efficient Path Planning Method of USV for Intelligent Target Search

Abstract

AbstractIn order to accomplish a target search task safely and efficiently and make full use of prior information and real-time information, a path planning method of unmanned surface vehicle (USV) for intelligent target search is proposed. The overall strategy is divided into three parts: global path planning based on prior information, local path planning based on real-time information, and improved A* obstacle avoidance algorithm. Before the start of the task, the global path planning is carried out based on prior information such as the initial position of USV, the predicted position of the target and range of search area. After the start of the task, if USV finds suspicious targets, in order to further approach these suspicious targets, it will enter different local path planning modes according to the characteristics of these targets. During task execution, if obstacles are encountered, an improved A* obstacle avoidance algorithm is adopted. The simulation results show that the proposed method can improve the efficiency of target recognition and reduce the turning cost of USV when encountering obstacles. So, for USV intelligent target search, the proposed path planning method can save resources and improve search efficiency.