Application of Soft Computing to Serial Manipulator Kinematic Problems

Abstract

Inspired by the techniques in artificial intelligence, this research applies the constituent of artificial intelligence to perform manipulator positioning tasks. This work combines three methods in artificial intelligence: fuzzy rules, neural networks, and genetic algorithm to form the control block specifically designed to solve kinematic problems of a six degree-of-freedom serial manipulator. The direct kinematic of a manipulator is taken as the feed forward control and the artificial intelligence control block replaces the inverse kinematics when the reverse solving is done. Fine manipulator positioning is achieved from the learning stages executed by the artificial intelligence control block. It is shown experimentally that the technique proposed was capable of producing results with very low errors.