An Adaptive Control Based Approach for Gripping Novel Objects with Minimal Grasping Force

Abstract

This paper presents the design, analysis, and experimental implementation of an adaptive control to facilitate 1-click grasping of novel objects by a robotic gripper. Motivated by a desire to obtain a reduced-order controller, a previously developed grasp model is reparameterized to design an adaptive backstepping controller. A Lyapunov-based analysis is utilized to show asymptotic convergence of the object slip velocity to the origin. Furthermore, the analysis shows that the closed-loop controller is able to estimate the minimal steady-state force required to grasp the object. Simulation and experiment results both show that the object is immobilized within the gripper without any significant deformation.