Artificial Potential Fields in a Complex Robotic System Configuration Space

Artificial Potential Fields for robotic path planning were firstly introduced by Khatib as a solution of the basic planning problem, where a single point shaped robot with no kinematic constraints has to move from a starting point to a goal, in presence of fixed obstacles. In this paper, it is presented a mathematical tool useful for dealing with Artificial Potential Fields in configuration spaces of different kinds of complex robotic system, such as manipulation structures and cooperating mobile robot team, also (in both cases) in narrow environment and with agents of arbitrary shape. It is also presented the resulfs of the application of this mefhod to the problem of path planning for a six joint planar manipulator in presence of obstacles, using the gradient expression as a base for a potential guided variational planning algorithm