Distributed Iterative Learning Control for High Performance Consensus Tracking Problem with Switching Topologies

High performance consensus tracking problem operating repetitively has attracted significant research interest in different fields. Recent research apply iterative learning control (ILC) for such problems, since ILC does not require a highly accurate model to achieve the high accuracy requirement (which is in contrast to most of the conventional control methodologies). However, existing ILC designs for high performance consensus tracking problem either focus on the tracking under fixed topology (while the switching topologies structure that is common used in reality has not been taken into account), or can only guarantee the convergence performance when the controller satisfies certain conditions. To address these limitations, this paper proposes a novel ILC algorithm for the high performance consensus tracking problem with switching topologies. The design of the novel performance index guarantees monotonic convergence of the tracking error norm to zero without any restriction on the controller. Furthermore, the proposed algorithm is suitable for homogeneous and heterogeneous networked systems, which is appealing in practice. A distributed implementation using the idea of the alternating direction method of multiplies for the proposed algorithm is provided, allowing the algorithm to be applied to large scale networked dynamical systems. Convergence properties of the algorithm are analysed rigorously and numerical examples are presented to show the algorithm’s effectiveness.