A view-time based potential field method for moving obstacle avoidance

Abstract

This paper proposes a new motion planning method of a mobile robot avoiding moving obstacles. We use an artificial potential field method and the view-time concept. The view-time is defined as the period between one sampling time to the next. It is assumed that the position and the velocity of obstacles are monitored at each sampling time. From the position and velocity at sampling time, we calculate the probability density of the obstacle trajectory for a view-time, based on the assumption that the obstacle motion is being modeled as a random walk process. The probability that the position can be swept by an obstacle during a view-time is calculated from the probability density function of the obstacle trajectory. Then artificial potential is assigned at every position by considering the probability of obstacle trajectory. The force induced by the artificial potential field repels the robot away from the probable obstacle trajectory. This method is a look ahead scheme and is effective for moving obstacle avoidance.