For achieving navigational safety and continuous communication link between swarms of unmanned marine vehicle (UMV) during mission execution, the cooperative path planning of unmanned surface vehicle (USV) and unmanned aerial vehicle (UAV) swarms is studied.

Keep-in and keep-out geo-fences are used to carry out scene modelling, and the problems of threat and obstacle avoidance are transformed into geo-fence constraints. Aiming at collision avoidance and continuous communication link between vehicles, a judgment criterion for the constraints of collision avoidance and communication link via time sequence detection is proposed. The average travel time (ATT) of the swarm is taken as the path optimization function, and the multiple constraints are transformed into penalty functions. A self-adaptive differential evolution algorithm is adopted to solve the optimization problem.

The proposed method can ensure safe navigation and communication link between USV and UAV swarms in hostile and obstacle-filled environment, and achieve the shortest ATT under multiple constraints.

This method has practical value for the off-line path planning of UMV swarms in the hostile and obstacle-filled environment.