Partitioning of Large-Scale Systems using Game-Theoretic Coalitional Methods

Abstract

In this paper, tools from cooperative game theory are combined with predictive control to perform the partitioning of large-scale systems (LSS). More specifically, a partitioning algorithm based on the Shapley value to rank the links by using a cooperative cost game is proposed. To this end, coalitional model predictive control, which offers a trade-off between control performance and communication burden, is considered to assess the value of the coalitions in the game. Also, combinatorial explosion issues are relieved by means of an attribution of value to the links based on the nodes they connect. The proposed method is implemented in the Barcelona drinking water network as a real LSS case study, showing the effectiveness of the proposed approach.