A Novel Approach To Local Minima Problem In Robot Path Planning Using Artificial Potential Field

Abstract

The Autonomous mobile robots have vast applications which include industrial robots, service robots, drone-based robots for supplying customer goods, robots for space exploration and much more. Path planning is one of the important steps of using mobile robots in any application. In autonomous mobile robot navigation, local minima describe a situation in which an object gets stuck with no next possible move attainable. This paper introduces a new concept to deal with the local minima problem in Artificial Potential based robot navigation methods. We propose a virtual plane based method which basically gets the robot out of the local minima successfully and inhibits it from getting into the same area again, thereby guaranteeing the convergence of the method. The proposed technique uses the concept of histograms and temporary objectives to achieve its goal. In particular, the histograms are used to find the wide enough gaps for the robot to pass through and the temporary objectives are used to escape the robot from local minima. The above technique is used, in particular, with the C and H type obstacles. The simulation results verify the effectiveness of this method.