Path planning for unmanned automaton based on improved artificial potential field method

Abstract

The artificial potential field method (APF), a highly effective navigation technique, is currently utilized extensively in this area due to the rapid development of unmanned automatons. Traditional APF, however, have several shortcomings, including the issue of unreachable object points and the propensity to sink into local minima that prohibit the automaton from moving on. In this paper, a two-part improved APF model is created to address these issues. First, by including additional constraints, the repulsive field model at the stumbling block is enhanced to address the issue that the object point is impassable when too close a distance between the two stumbling blocks. Secondly, a new potential field is introduced to help the automaton walk out of the local minima. Analogue simulation show that the methods mentioned above can solve these problems better and make the route planning of unmanned automatons come true.