Self-Configuration Fuzzy System for Inverse Kinematics of Robot Manipulators

Abstract

Kinematics is the study of motion without regard to the forces that create it. Generally, kinematics for robot manipulators includes two problems, forward kinematics problem and inverse kinematics problem. Because of the complexity of inverse kinematics, it is difficult to find the solutions for it. This paper applies a self-configuration fuzzy system to finding the solutions for inverse kinematics of robot manipulators. In this paper, the problem of fuzzy approach for inverse kinematics is described first. Then a self-configuration fuzzy system is introduced. Based on a small group of given input-output pairs selected by covering its workspace, an initially simple fuzzy model for inverse kinematics is built, including some basic rules, the number and the parameters of membership functions. Then, by applying given input-output data pairs to this simple fuzzy model, approximating error can be calculated. Consequently, the optimum rule conclusion, optimized membership functions, and new structure can be obtained. Furthermore, an overall analysis in the domain of the whole function is carried out instead of concentrating on the subspace. After optimization problems are solved, a fuzzy system is well defined to solve the inverse kinematics. Finally, the simulation verifies this self-configuration fuzzy system for inverse kinematics of robot manipulators.