A COLREGs-compliant guidance strategy for an underactuated unmanned surface vehicle combining potential field with grid map

A guidance strategy conforming to the International Regulations for Preventing Collisions at Sea (COLREGs) is introduced in this paper such that its objective effectively combines unmanned surface vehicle (USV) dynamics with the obstacle avoidance process. The strategy is divided into the global path-guided and local reactive layers based on expanded Theta* (ETheta*) and improved artificial potential field (IAPF). The global layer constructs a nominal mapping relation (NMR) through running the ETheta* reversely. The USV is guided by sub-targets extracted from the NMR to steer clear of all stationary obstacles and reach the destination. The local layer generates the double-attractive potential of adjacent sub-targets and grid-based repulsive potential of irregular shape stationary obstacles. A COLREGs-compliant repulsive potential is additionally created when the USV detects the danger of collision. Taking the gradient of steepest descent in total potential field as the desired heading, PID method is used to control the USV navigation and realize the autonomous obstacle avoidance. The whole guidance strategy is modeled and simulated on unblocked and real maps respectively for some complicated scenarios. The results show that, with this guidance strategy, the vehicle avoids other target ships and obeys COLREGs with appropriate USV dynamics. • The nominal mapping relation is employed for global guidance. • Several potential fields are established to avoid collision for the USV. • COLREGs are considered during collision avoidance maneuvers.