Iterative learning control for human-robot collaborative output tracking

Abstract

This article studies human-robot learning control for collaborative output-tracking tasks. We propose an algorithm to adaptively tune the frequency-dependent iteration gain and apply it to two cases: when the desired output is directly available to the robot and when the robot infers the desired output from human-achieved output. Experiment results are presented to illustrate the application of the proposed approach to a human-robot collaborative output-tracking task. Results show that the error converges to less than the closed-loop robot tracking error, and that the approach can provide varying levels of robot assistance by selecting the desired human-robot collaboration level.