Intelligent Dynamic Collision Avoidance Strategy of Hydrogen Fuel Cell Unmanned Ship via Improved Fusion Dynamic Window Method

In this study, a novel fusion dynamic window method is proposed. The safety field obstacle model of obstacle ships is explored to improve the obstacle collection of speed obstacle method, so that the collision avoidance process of Hydrogen Fuel Cell Unmanned Ship (HFCUS) can retain more feasible paths. The speed window is discretized, and the speed vector space is filtered based on the speed obstacle collection in the safety field to realize the fusion of algorithms. The objective evaluation function is improved, and the rule constraint function is presented, so that the collision avoidance behavior of HFCUSs conforms to the constraints of the Convention on the International Regulations for Preventing Collisions (COLREGS), and the safety in the collision avoidance process is improved. Simulation and experiment are verified that the proposed algorithm can effectively avoid dynamic obstacle ships, and the avoidance behavior meets the rules and constraints. HFCUS is realized the safe navigation in the environment of dynamic obstacle ships.