Fuzzy inverse kinematic mapping for a redundant robot

Abstract Inverse kinematics is computational expensive, and for some cases requires numerical solution. Especially for a redundant robot, the inverse kinematic mapping has to be done using a task and/or environmental model and an optimization scheme. This results in inefficient computation and control delay in real-time implementation. In this paper, we propose a fuzzy logic approach for the inverse kinematic mapping which is independent of the number of links, and thus is feasible for both non-redundant and redundant robots. The approach produces zero steady-state error and can be carried out efficiently. The feasibility of this approach has been demonstrated by implementing it in the real-time teleoperation control of a seven-joint space robot.