Vision-based control of SMA-actuated polymer microgripper with force sensing

Abstract

A vision-based control of SMA-actuated polymer gripper with force sensing is implemented and tested. The gripper system consists of PU (polyurethane) mechanism of gripper with force sensing arm, SMA actuator, and visual-based position and force sensing system. The information of gripper motion is tracked by utilizing computer-vision method through extended Regional Edge Statistics (ERES) algorithm. The position information of gripper tips together with the deflection of force sensing arm is utilized for sensing force. Closed-loop control including PI, PI-FF (PI with feedforward), and FE-PI (fuzzy expert PI) controller are implemented and tested. FE-PI controller is employed for testing the performance of sensing gripping force in grasping of 38μm metal rod.