Path planning of USV based on improved PRM under the influence of ocean current

Abstract

Multi-extensibility and flexibility of unmanned surface vehicles (USVs) allow them perform many different tasks, further path planning technology is crucial to the safety, autonomy, and intelligent navigation of USVs. Firstly, this paper analyzes the impact of ocean currents and risk constraints on USV based on the electronic chart. Then take the optimal sailing time as the objective function and design a path planning algorithm based on an improved probabilistic roadmap (PRM) algorithm, in which a Gaussian space sampling algorithm based on edge detection is introduced. After building the network topology environment through improved PRM, then a Dijkstra algorithm based on great circle distance is used to solve the optimal path. Finally, the simulation experiment is designed through the MATLAB platform. By comparing the average and the three quartile lengths of the planned paths under three environments, the values of the designed Edge-Gaussion (E-G) PRM algorithm are smaller than Lazy PRM and Gaussian PRM algorithm, which shows that the improved PRM algorithm has better performance. When planning the USV path under the influence of current, compared with traditional length optimal path planning, although the navigation length planned by the designed algorithm is shorter by 972 m, sailing time is improved by 110 s, which efficiency shows the better application on the sea.