IBVS with Fuzzy Sliding Mode for Robot Manipulators

Abstract

Closed loop control of robot manipulator's end effector pose with visual feedback is called as visual servoing (VS). As one of the approaches for VS, image-based visual servoing (IBVS) has the advantage of no pose estimation for commonly used eye-in-hand configured manipulators. VS aims to minimize the error derived from k feature points vector s in image feature space and it controls the velocity of the end effector from error signals. This velocity control is based on sliding mode control (SMC) with a fixed gain. Choice of an appropriate gain plays a critical role in the performance of this controller. This study is focused on varying gain for fast convergence with varying sliding slope approach. Computing gain using fuzzy logic that is an approach in fuzzy SMC is proposed.