2-D Automatic Micrograsping Tasks Performed by Visual Servo Control

摘要

In this paper, we propose a visual servo control approach and develop the corresponding software to perform micrograsping tasks in a two-dimension plane (2-D). Specifically, using a 6 degree-of-freedom (DOF) robotic manipulator to manipulate a micro-gripper and automatically grasp a micro-part. In the proposed control scheme, two types of position feedback signals are utilized. One is the relative positions of the micro-part with respect to the micro-gripper measured by the developed computer vision system, which is used for visual servo control. The other is the absolute displacements of the micro-part measured by the fixed linear encoders, which is used to check the corresponding motions. The whole control software is developed by LabVIEW, to improve operation speed. Furthermore, a two-stage grasping strategy is employed: in the first stage, pattern-matching is performed once and the bonded micro-gripper is controlled, to directly reach a specific position adjacent to the mating edge of a designated micro-part with the same y coordinate; in the second stage, pattern-matching is performed at each control update interval while the micro-gripper is displaced towards the micro-part with steps on the order of one micron magnitude, until the micro-part is completely grasped. Experiments conducted on a 6-DOF experimental manipulator demonstrate the efficiency and validity of the proposed control approach and grasping strategy.