A Novel Artificial Potential Field Based Ship Path Planning Algorithm Using Model Predictive Control Strategy

Abstract

There exist some limitations and defects when the conventional artificial potential field (APF) based methods are utilized for ship path planning, especially in complex navigation scenarios. In consideration of the path planning requirements for emerging autonomous or smart ships, a novel path planning algorithm that combined the improved artificial potential field method with model predictive control is proposed. A novel potential field function is proposed to achieve collision avoidance and target point arrival in complex waters. Then, ship maneuverability based on the Nomoto model is considered to generate the trackable path. Moreover, the path planning problem is converted to a nonlinear optimization problem, and maneuverability models, position update methods, etc., are considered in the constraints. Finally, the path planning problem is solved using model predictive strategy and online optimization algorithm. We verify the effectiveness of the MPC with different prediction steps in a multi-obstacle scenario. The simulation results show that compared to traditional APF and particle swarm optimization (PSO)-based APF, the proposed algorithm can generate a safer and more feasible path in complex scenarios and meet the real-time requirements.