Computing robot configurations using a genetic algorithm for multimodal optimization

Abstract

This paper describes a general approach for robot inverse kinematic transformation, which computes all possible solutions of this multimodal problem. A function which computes the error between two tool center points (TCPs) is minimized by an advanced genetic algorithm. This multi-population genetic algorithm uses exchange and migration operators for computing all solutions of multimodal optimization problems.