An improved potential field method for path planning

Abstract

In this paper, an improved potential field method for mobile robots path planning is proposed with modified attractive and repulsive functions.The attractive function decreases the attractive potential around a goal evidently to ensure that the goal position is the global minimum of the total potential field. The repulsive function builds repulsive potential for an arbitrarily shaped obstacle by discretizing outline of the obstacle with points. Therefore, the workspace of the robot is described more exactly. The convergence of this method is discussed to guarantee its reliability for most cases. The simulation has proved that the new potential method can solve the problem of goals nonreachable with obstacles nearby (GNRON), U-shaped trap, and oscillations. Moreover, this method has a great performance in dynamic path planning where a goal and obstacles are moving.