Efficient path planning with limit cycle avoidance for mobile robot navigation

Abstract

Autonomous navigation in an unknown environment may encounter the limit cycle problem causing lower efficiency when a mobile robot attempts to reach the goal while avoiding obstacles on the way. To cover a wider range of navigation tasks, this paper presents a novel waypoint navigation method in polar coordinate. Navigation limit cycle avoidance is handled by creating and memorizing a specific traversable but less preferred areas during waypoint navigation in which the robot changes its orientation to follow from goal-directed to opportunistic goal-repulsive behavior. This allows the robot to be able to predict, thus avoiding upcoming limit cycle situations. The saving of time and memory resources of path computation using the novel polar-coordinate based waypoint navigation has been verified by simulations, and an implementation of the navigation method on a real mobile robot works well in an indoor experiment with demonstrated planning and route selection capabilities for waypoint navigation.