An improved algorithm for inverse kinematics solution of redundant modular robots

It is difficult to get solutions that can meet the require-ments of high positioning accuracy and real time control simultaneously among the infinite solutions of a redundant robot. Based on weighted least-norm method(WLNM), an optimized algorithm for inverse kinematics solution of redundant modular robots, with which the joint limits can be avoided, is presented in this paper. According to the characteristic of robot global wrist organ, the position and pose of the robot tag end can be resolved separately, therefore the computing time is reduced without decreasing the solution accuracy. Recently, an Intravehicular Robot Service System’s Ground Demonstration for Space Station has been built in the Space Robot Lab of Bupt, and the operation object of the demonstration is a 9-DOF modular reconfigurable robot. Through the experiments of pulling drawers and pressing buttons by the 9-DOF robot, the accurate and real-time solutions are given to show the applicability and effectiveness of the proposed algorithm. The experiment results demonstrate that the positioning accuracy is up to ±2.0mm and single-step computing time is less than 1 ms.