Distributed Iterative Learning Impedance Control for a Team of Robot Manipulators with Varying Trial Lengths

Abstract

This article presents a distributed iterative learning impedance control algorithm for a team of robot manipulators with varying trial lengths. This approach enables each manipulator to achieve the desired impedance model only using the impedance information of its neighbors, eliminating the need for direct access to the desired joint angle profiles. Furthermore, the proposed scheme addresses the challenge of randomly varying operation lengths across iterations. This capability is particularly important for ensuring the robustness of practical industrial systems, where trial durations vary due to dynamic task requirements. It is demonstrated that the impedance errorL² -norm of each manipulator converges to zero as the iteration index approaches infinity even under variable-length operations. Finally, experimental validation using collaborative robot manipulators confirms the effectiveness, adaptability, and practicality of the proposed method.