A period-varying iterative learning control scheme is proposed for a robotic manipulator to learn a target trajectory that is planned by a human partner but unknown to the robot, which is a typical scenario in many applications. The proposed method updates the robot’s reference trajectory in an iterative manner to minimize the interaction force applied by the human. Although a repetitive human–robot collaboration task is considered, the task period is subject to uncertainty introduced by the human. To address this issue, a novel learning mechanism is proposed to achieve the control objective. Theoretical analysis is performed to prove the performance of the learning algorithm and robot controller. Selective simulations and experiments on a robotic arm are carried out to show the effectiveness of the proposed method in human–robot collaboration.

This document is made available in accordance with publisher policies and may differ from the published version or from the version of record. If you wish to cite this item you are advised to consult the publisher's version. Please see the URL above for details on accessing the published version.

No abstract

This paper presents a novel method for learning and tracking of the desired path of the human partner in physical human-robot interaction. Combining the Adam optimization algorithm with iteration learning control (ILC), a path learning method is designed to generate and update reference waypoints according to the human partner's desired path. This method firstly uses the Adam optimization algorithm to update the robot's reference waypoints in an online manner. Then, an ILC is developed to further modify the waypoints and reduce the difference between the robot's actual path and the human partner's desired path in an iterative manner. Simulations and experiments on a 7-DOF Sawyer robot are carried out to show the effectiveness of our proposed method.

This article proposes a dynamic time warping (DTW)-based iterative learning control (ILC) scheme for discrete-time nonlinear systems to tackle the path learning problem with varying trial lengths and terminus constraint. By incorporating the improved DTW algorithm, the varying trial lengths are aligned as a desired length. Meanwhile, this algorithm can find the most similar waypoints between the output and the desired paths, which can be used to design an updating law and facilitate the convergence of path learning. Different from the existing ILC methods based on the probability distribution function for learning trajectory in the time domain, the method in this article can be applied to learn the spatial path corresponding to the desired trajectory. Furthermore, the learning property in the presence of variable initial states is discussed under the proposed method. Several illustrative examples manifest the validity of the proposed DTW-based ILC algorithm.