Redundant manipulators have extra degrees of freedom. It is possible for the redundant manipulator to reach an object by many different manipulator configurations. Because of it, redundant manipulators are used effectively in obstacle avoidance. However, it is very difficult to calculate a single set of joint angles to reach the end-effector to a target. In this paper, we propose a new method to solve the inverse kinematics using topology representing network. © 2005 SICE.