Safety-Critical Pursuit-Evasion Differential Game Guidance for Multiple Underactuated Autonomous Surface Vehicles

This paper explores an adversarial pursuitevasion differential game problem of two groups of under-actuated autonomous surface vehicles (ASVs) in an obstacle environment. A pursuit-evasion game method featured with intra-team cooperation and inter-team adversary is proposed. Optimal control theory is utilized to obtain the optimal control strategies of pursuing ASVs and evading ASVs. Speed-heading control barrier functions are utilized to address collision avoidance with obstacles and ASVs within the same team. A quadratic programming (QP) problem is formulated by combining the pursuit-evasion strategies with collision avoidance constraints and input constraints. Real-time optimal control signals are generated by solving QP problem. Simulation results effectively validate the effectiveness of the proposed method.