Using a multi-agent approach to solve the inverse kinematics problem

Abstract

In robotic application the generation of a trajectory for a robot requires resolution of the inverse kinematics problem by solving a set of equations representing the geometry model of the robot. The resolution of this system is well known, but requires, to define this set of equations, considerable work. A multi-agent approach that solves a reduced form of the inverse kinematics problem is presented. It is shown that this solution results in very simple simple expression. The efficiency of the approach is discussed.