The design of an obstacle avoiding trajectory in unknown environment using potential fields

Abstract

This paper described the inherent local minimum problem and oscillations in the presence of obstacles or in narrow passages of traditional potential field methods for mobile robots in planar and static environment. Based on the character of a flow of water, this paper proposes a new method of obstacle avoiding trajectory using potential grid method. The main benefits of the method are the quickness of minimum computation and the compensation for the main drawbacks belonging to the potential field method in general. The initial position of the robot is assumed as an artificial headstream, and then the artificial stream flows from the high potential field to the low. When the artificial stream flows to the target position, the track of the artificial stream forms a feasible obstacle avoiding trajectory. Then, the algorithm for constructing obstacle avoiding trajectory is described. And this method has been simulation in long wall, recursive U-shaped, unstructured, cluttered, and maze-like environment. The results show that the proposed method is feasible and effective in path planning.