A New Artificial Potential Field Based Global Path Planning Algorithm for Mobile Robot Navigation

Abstract

This paper proposes a navigation algorithm for indoor mobile robots in unknown and unstructured environments using a modified Artificial Potential Fields (APF) approach. The proposed algorithm takes into consideration the well-known drawbacks of the APF such as local minima, U-Shaped traps, box canyon and the goal non-reachable with obstacles nearby (GNRON) and implies certain modifications to address those problems. The proposed method is tested in a series of simulations performed on actual lidar measurement data and taking into consideration the problems mentioned above. The results of the simulations confirm the performance of the proposed method yielding a robust and fast algorithm for indoor localization and navigation of mobile robots.