퍼지추론을 利用한 로보트 逆자코비안의 解에 관한 硏究

Abstract

The purpose of this paper is to solve the robot inverse Jacobian by using fuzzy reasoning. Manipulator inverse kinematics is the problem of finding the required joint angles to control the position and orientation of the end-effector of robot, and the solution is usually very complex and nonlinear. Robot inverse Jacobians relate Cartesian velocities to joint velocities, and are very complex to be solved. So, we attempt to represent the relation of joint angles and the position of the end-effctor by a fuzzy model. This paper proposes to use TSK fuzzy reasoning method to solve inverse kinematics and inverse Jacobian. TSK fuzzy reasoning method uses If—then—rules, of which consequents are linear systems very well and can be identified by using input-output data without analytical knowledge of the systems. The inverse Jacobians is easily solved from the fuzzy model of inverse kinematics. A two-link manipulator example shows the velocity of the proposed method.