NAPF-A*: A nonlinear APF fused A* algorithm for smooth path planning of port AGVs

Automated Guided Vehicles (AGVs) serve as key equipment for horizontal cargo handling at container terminals, and their path planning performance is of great significance for achieving smart and efficient port operations. To address the issues of unreachable targets and path oscillations in the traditional Artificial Potential Field (APF) algorithm, as well as the excessive polylines and poor adaptability to dynamic environments in the A* algorithm when applied to port AGVs, this paper proposes a NAPF-A* fusion path planning method for port AGVs. By constructing a nonlinear potential field function, a novel Nonlinear Artificial Potential Field (NAPF) algorithm is developed to overcome the inherent limitations of the conventional APF approach. This NAPF algorithm is then integrated with an A* algorithm enhanced by eight-direction neighborhood search and cubic Bézier curves, forming the NAPF-A* hybrid algorithm that achieves an effective combination of global optimal planning and real-time dynamic obstacle avoidance. Simulation experiments under static and dynamic environments as well as physical tests on the TurtleBot3 platform verify the superiority of the proposed algorithm, which can reduce the peak acceleration and path length of AGVs.