USVs Cooperative Collision Avoidance Based on Man–Machine Interaction and the Artificial Potential Field Method

Abstract

Unmanned surface vehicles (USVs) are sometimes still unable to cope with complex environments and difficult tasks despite already having a high-level autonomous capabilities. Thus, human cognitive abilities are used to promote the safety of USVs. Human–computer interactions can help users solve problems through man–machine collaboration. This paper primarily studies how to improve the capability of USVs to avoid collisions with the help and assistance of users and proposes a method that is based on knowledge accumulation for improving USV autonomous collision avoidance ability. Using this method, parameters are planned according to the tasks given by the user. The knowledge base of the navigation task is constructed through a combination of uncertain environment information and is used to improve the autonomous collision avoidance capability of a USV. The USVs formation consensus motion is controlled by the consensus path planning method. Simulation results show the effectiveness of the proposed method.