Adaptive Impedance Control of Robots With Reference Trajectory Learning
Qing Xu, Xiaoming Sun
Abstract
In this paper, we propose an adaptive impedance control with reference trajectory learning for the robots interacting with unknown environment. A cost function considering its tracking errors and interaction force is introduced and a reference trajectory learning law based on iterative learning is presented to minimize it. Also, an adaptive impedance control is designed to follow the target impedance model with the adaptive reference trajectory to implement the convergence of tracking errors and interaction force. Through simulation and experimental studies, we find that the robot can autonomously adjust its trajectory or interaction mode when the environment types or cost function parameters are tuned, so that a more humanized and intelligent interaction can be realized.