F-1226 Robot Motion Planning by Detecting Local Minimum Points of Oval Potential Field
Yukihiro EGUCHI, Hirohisa Kojima · The proceedings of the JSME annual meeting · 2001
This paper studies two-dimensional robot motion planning based on the oval potential field. The drawback of the traditional method is that local minimum pints are generated in the potential field in the case that obstacles locate nearby at each other or are concave. To overcome the above drawback, in this paper, local minimum points are detected by calculating the gaussian and mean curvatures of the potential field, and are combined to the nearest obstacle. Then, the resulting obstacles will be convex. The effectiveness of the present method is shown through numerical simulations.