Iterative Learning-Based Distributed Impedance Control for Robot Teams with Variable Trial Lengths

This article gives a novel distributed iterative learning impedance control scheme for a team of robot manipulators subject to varying trial lengths. The proposed approach enables each manipulator to achieve the desired impedance model by relying solely on neighboring impedance data, without requiring direct access to joint angle profiles. Additionally, it effectively addresses the challenge of differing operation lengths across iterations, and ensures the robust performance for industrial systems with dynamic task requirements. The algorithm is proven to drive the L²-norm of the impedance error for each manipulator to zero as iterations progress. Experimental results demonstrate the method’s effectiveness, adaptability and practicality for real-world applications.