Adaptive Human–Robot Interaction Control for Robots Driven by Series Elastic Actuators

Abstract

Series elastic actuators (SEAs) are known to offer a range of advantages over stiff actuators for human–robot interaction, such as high force/torque fidelity, low impedance, and tolerance to shocks. While a variety of SEAs have been developed and implemented in initiatives that involve physical interactions with humans, relatively few control schemes were proposed to deal with the dynamic stability and uncertainties of robotic systems driven by SEAs, and the open issue of safety that resolves the conflicts of motion between the human and the robot has not been systematically addressed. In this paper, a novel continuous adaptive control method is proposed for SEA-driven robots used in human–robot interaction. The proposed method provides a unified formulation for both the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">robot-in-charge </i> mode, where the robot plays a dominant role to follow a desired trajectory, and the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">human-in-charge </i> mode, in which the human plays a dominant role to guide the movement of robot. Instead of designing multiple controllers and switching between them, both typical modes are integrated into a single controller, and the transition between two modes is smooth and stable. Therefore, the proposed controller is able to detect the human motion intention and guarantee the safe human–robot interaction. The dynamic stability of the closed-loop system is theoretically proven by using the Lyapunov method, with the consideration of uncertainties in both the robot dynamics and the actuator dynamics. Both simulation and experimental results are presented to illustrate the performance of the proposed controller.