A novel visual servoing feedback control of nonholonomic wheeled mobile robots

Abstract

Currently, some dynamic tracking controllers have been proposed for nonholonomic mobile robots with the unknown visual parameters and has been proved to stabilize the closed loop systems in theory. However, there is lack of the relevant experiment verification except for several simulations. In this paper, we proposed a novel visual servoing feedback control strategy to make the nonholonomic mobile robot track the given reference trajectory. Firstly, we paste two red rectangles with different areas on the top of the robot as marks. The big area is used to describe the location of the robot, and the other is used to auxiliary calculate the orientation angle. The error between the desired marks' contour feature and the current one on the image plane is regarded as the objective of the tracking control of the robot. Secondly, the dynamic feedback tracking controller is used to calculate the robot velocities and the saturating process on them is necessary. Finally, the result of experiments have shown that the image error of the robot converged to the desired value.