Local path planning for USVs based on the fusion algorithm of improved dynamic window approach and velocity obstacle algorithm

The traditional dynamic window approach (DWA) does not consider the motion characteristics of unmanned surface vehicles (USVs), and the planned path bypasses the periphery of the dense area of obstacles, which will cause the path length to be too long. At the same time, the USVs will be too close to the dynamic obstacles, so the dynamic obstacle avoidance effect will be poor. In order to solve these shortcomings, a local path planning algorithm of USVs integrating the improved dynamic window method and the velocity obstacle algorithm was proposed. Firstly, considering the motion characteristics of USVs, the influence of yaw rate and drift angle was considered in the speed selection. Then, the concept of obstacle search angle was proposed to better deal with the impact of obstacles on USV navigation, which ensures USVs to pass through the dense area of obstacles without collision and to greatly shorten the navigation time and path length. Finally, with integration of the velocity obstacle algorithm, a new evaluation function of dynamic window algorithm was obtained according to the relative speed between USVs and dynamic obstacles, so that USVs can avoid dynamic obstacles at a faster speed under the condition of collision avoidance safety. The numerical simulation results show that the path planned by the improved method has less twists and turns and is smoother, which is conducive to the navigation of USVs. The improved method is better than the traditional algorithm in calculation of path length, navigation time and average speed, and has a stronger dynamic obstacle avoidance adjustment ability. © 2023 China Ship Scientific Research Center. All rights reserved.