Method and system for formation control for unmanned surface vessel swarm via collaborative exploration deep reinforcement learning (CEDRL)

A system for formation control for aUSV swarm via a CEDRL, comprising:

• a USV desired location library, designing a desired formation pattern based on a formation hierarchical virtual leader strategy, being established, assigning a location index to a desired location of each USV in a formation, and establishing a matching relationship between the USV and the corresponding desired location;
• a location index updating unit, configured to update the desired location of each USV and the corresponding location index via an USV formation local consensus strategy in a case where there is a risk of collision between USVs;
• and
• a USV control unit, configured to acquire an actual geolocation of each USV in real time, and adopt a surface vessel control decision-making network to direct the USV toward a corresponding latest desired location;
• wherein the surface vessel control decision-making network is obtained by training based on the CEDRL.