Image space trajectory tracking of 6-DOF robot manipulator in assisting visual servoing

Abstract

As vision is a versatile sensor, vision-based control of robot is becoming more important in industrial applications. The control signal generated using the traditional control algorithms leads to undesirable movement of the end-effector during the positioning task. This movement may sometimes cause task failure due to visibility loss. In this paper, a sliding mode controller (SMC) is designed to track 2D image features in an image-based visual servoing task. The feature trajectory tracking helps to keep the image features always in the camera field of view and thereby ensures the shortest trajectory of the end-effector. SMC is the right choice to handle the depth uncertainties associated with translational motion. Stability of the closed-loop system with the proposed controller is proved by the Lyapunov method. Three feature trajectories are generated to test the efficacy of the proposed method. Simulation tests are conducted and the superiority of the proposed method over a Proportional Derivative – Sliding Mode Controller (PD-SMC) in terms of settling time and distance travelled by the end-effector is established in the presence and absence of depth uncertainties. The proposed controller is also tested in real-time by integrating the visual servoing system with a 6-DOF industrial robot manipulator, ABB IRB 1200.

Keywords

• feature point trajectory
• image space tracking
• sliding mode control
• depth uncertainties