A Novel Unmanned Surface Vehicle Path-Planning Algorithm Based on A* and Artificial Potential Field in Ocean Currents

Abstract

Ocean currents make it difficult for unmanned surface vehicles (USVs) to keep a safe distance from obstacles. Effective path planning should adequately consider the effect of ocean currents on USVs. This paper proposes an improved A* algorithm based on an artificial potential field (APF) for USV path planning in a current environment. There are three main improvements to the A* algorithm. Firstly, the proposed algorithm ignores unnecessary perilous nodes to decrease calculation. Secondly, an adaptive guidance angle is developed to guide the search in the most appropriate direction to reduce the computing time. Thirdly, the potential field force function is introduced into the cost function to ensure that the path designed for the USV always maintains a safe distance from obstacles under the influence of ocean currents. Furthermore, the Bezier curve is adapted to smooth the path. The experimental results show that the USV path-planning algorithm proposed in this paper, which synthesizes the APF and A* algorithms, runs 22.5% faster on average than the traditional A* algorithm. Additionally, the path developed by the proposed A* algorithm effectively keeps appropriate and different distances from obstacles by considering different ocean currents.

Local Path Planning Method for Unmanned Ship Based on Encounter Situation Inference and COLREGS Constraints

“…A method for multi-ship swarm path planning based on local sensor information was proposed by Wang et al [22], achieving collision-free, smooth, and dynamically feasible path generation through dynamic path search and B-spline curve optimization. Yang et al [23] improved the A* algorithm based on the artificial potential field method to consider the influence of ocean currents and static obstacles on path planning, and smoothed the generated paths. Given the focus on constant flow conditions, further research efforts are required for practical applicability.…”