Roadmap - A : an Algorithm for Minimizing Travel

This paper presents an algorithm for minimizing the travel e ort of a mobile robot navigating in an unknown environment along a graph constructed on-line by the robot. The graph, called a roadmap, represents the free con guration-space of the robot, and the travel e ort is measured by the length of the path traveled by the robot during the search. The algorithm, called Roadmap -A , strives to minimize the travel e ort by combining features of the classical A and local graph-search algorithms. A user-speci ed parameter determines the relative emphasize on the two approaches, where =0 corresponds to A while =1 corresponds to local-search. We study the performance of the algorithm on roadmaps generated by a cylindrical mobile robot navigating in an unknown environment, using a potentialeld roadmap construction method. However, Roadmap -A is general and can be used by any on-line roadmap construction method. We study the performance of Roadmap -A for various values of on simulated environments, and indicate how to best choose for a given class of environments.