Research on AMR Rapid Path Planning Algorithm Based on Variable-Scale Canyon

Abstract In large-scale map environments, the increased computational load of the A* algorithm and existing shortest path principles often lead Autonomous Mobile Robot (AMR) to favor one side of corridors, which does not provide sufficient safety for the robots. Inspired by the mountainous terrain and river flow patterns within canyons, this paper proposes the Canyon algorithm. First, a variable-scale grid map is utilized to generate a coarse-precision map through map scale resizing. Meanwhile, the cost principle is improved by gradually decreasing the cost of cells that are further away from obstacles, generating a canyon cost map from which traversable river paths can be planned. Second, the river paths are expanded according to the diagonal principle and then mapped back to the original map to obtain the river area. Finally, within the river area, the A* algorithm is employed to plan a safe path. The results indicate that the AMR path planning algorithm based on the variable-scale canyon algorithm can effectively plan safe paths, keeping the AMR paths away from potential collision areas and enhancing safety.