A Study on Evaluating the Performance of Robot Motion Using Gradient Generalized Artificial Potential Fields with Obstacles

Abstract

Motion planning (MP) is actually a specialized version with more general Artificial Intelligence (AI) planning problem. The goal of a general purpose planning problem is to come up with a sequence of actions that accomplish a given goal. There are two approaches that utilize random samples. One the Probabilistic Road Map algorithm that sought the construct of a road map. Second, the rapidly exploring random tree procedure, which constructs every evolving trees to explore the free space and forge paths between the start and the goal. Both algorithms have the pleasing property that they work quite well in practice, even on high dimensional configuration spaces. Finally, the behavior of the generalized potential fields with obstacles are discussed. It helps to steer the robot through configuration space by considering the gradient of this artificial potential field. A strength of these potential field methods is that they are relatively simple to implement, and they can often be carried out directly based on sensory input. The finding of the present research is to describe the challenges being faced in the area of motion planning of robot and controlling its speed using gradient in generalized artificial potential fields with obstacles.