Parameter Optimization of Artificial Potential Field Algorithm for Local Path Planning Problem
Rafał Szczepański · 2025
Local path planning problems are crucial for the safe operation of autonomous mobile robots in an unknown environment. One of the most commonly used algorithms is called an Artificial Potential Field. Its principle can be described as follows: the goal is pulling the mobile robot to itself, whereas the obstacles push it away. Although the principle is simple, the algorithm requires the definition of a few parameters. Improper selection may lead to unbalanced forces generated by the goal and the obstacles. This imbalance may cause a collision, highly unsmooth movement, or failure to reach the goal position. In this paper, the selection procedure for these parameters is considered. For this purpose, the optimization algorithm was used to provide the optimal solution for the dynamics of mobile robots and the environment for the defined objective function. The optimal parameters allow the selection of efficient Artificial Potential Field parameters, which provide superior performance for the considered scenario. The parameters obtained have been examined using a real mobile robot, Husarion ROSbot 2.0 PRO, in the Mobile Robotics Laboratory at Nicolaus Copernicus University in Torun.