Variable Admittance Control for Compliant Collaboration in Physical Human-Robot-Environment Interaction

Abstract

This paper explores the development of intelligent and adaptive interaction control strategies for physical human-robot-environment interaction. Collaborative robots with variable admittance control have shown promising results in providing compliant behavior of the robot's end-effector in response to human guidance. However, instability can arise in the transition from free motion to contact motion due to fluctuations in force and imbalance. To address this issue, a decoupled interaction interface is proposed, which separates human and environmental forces for stable interaction. The proposed variable admittance control scheme uses an adaptive controller to update the unknown impact or poor regulation between the robot and its environment. A passivity framework is necessary to maintain the passivity of the overall system, which is derived using the Lyapunov approach. The proposed approach offers better stability and performance in physical human-robot-environment interaction.