Online distributed design for control cost reduction

In this paper we consider the problem of reducing control costs in Multi-Agent Systems in a distributed fashion, while maintaining or seeking a certain level of global performance. The problem is formulated to minimize the controller gains locally, while respecting a global bound on the state deviations due to effects of external disturbances We show that with the chosen performance measure this can be reformulated as a convex optimization problem. However, due to the limited knowledge of the agents and structural changes in the networks, currently existing algorithms for solving distributed optimization problems cannot be used as is. We therefore derive an algorithm that is capable of solving the problem, while also respecting limitations caused by the aforementioned issues. We end the paper with an example where the problem is solved in a Multi-Terminal DC-grid. (c) 2022 Elsevier Ltd. All rights reserved.