A Fuzzy Approach to Visual Servoing with A Bagging Method for Wheeled Mobile Robot

Abstract

Visual servoing is a vision-based control method with a mechanism of closed-loop, which uses feedback information extracted from the camera to control the motion of a robot. In this paper, we propose a fuzzy-based visual servoing integrated with a bagging method for the wheeled mobile robot (WMR). Previous studies have shown that the value of the mixture parameter for the image Jacobian matrix affects the performance of image-based visual servoing(IBVS). However, the mixture parameter value is constant in most visual servoing methods. To address this problem, we propose a fuzzy-based method to adjust the mixture parameter during the process of visual servoing. Meanwhile, in order to reduce the effect of image noise and the computational complexity of the pseudoinverse matrix, we propose a bagging method to calculate the inverse kinematics, instead of using the Moore-Penrose pseudoinverse method. The results of simulation and real experiments demonstrate the effectiveness of the proposed IBVS method.