Development of path planning algorithm using potential field based on two fixed-angle laser scanners

摘要

Stable and robust environment recognition is a key function required by mobile robots to avoid collision with obstacles. The 2D laser scanner is a compact and stable environment recognition sensor for horizontal scanning of wide areas. However, depending on the obstacle, the complete shape might not be detected by only two-dimensional scanning. To ensure stable detection of obstacles, we propose two fixed-angle laser scanners that can reconstruct the three-dimensional shape of obstacles. The reconstructed shapes are used to produce a potential field map for mobile robot path planning. The validity of the proposed path planning algorithm is confirmed through tests conducted in a real-word outdoor environment.