A Dynamic Obstacle Avoidance Method for Unmanned Surface Vehicle under the International Regulations for Preventing Collisions at Sea

A new method is proposed for the dynamic obstacle avoidance problem of unmanned surface vehicles (USVs) under international regulations preventing collisions at sea (COLREGs), which applies particle swarm optimization algorithm (PSO) to window approach (DWA) reduce optimal trajectory finding time and improve timeliness avoidance. Meanwhile, a fuzzy control designed dynamically adjust weight coefficients velocity distance terms in cost function DWA adapt changes environment. The experimentally validated, PSO combined with (PSO-CCDWA) results 42.1%, 11.2% 28.0% reduction navigation USVs three encounter-situations COLREGs than that classical (CCDWA) conforming conventional COLREGs, respectively. (FUZZY-CCDWA) reduces traveled by 15.8%, 0.9% 2.8%, respectively, over CCDWA encounter scenarios. Finally, effectiveness further verified practical experiment USV named “Buffalo”.