Decentralized Coalition Formation with Agent-based Combinatorial Heuristics

Abstract

A steadily growing pervasion of the energy distribution grid with communication technology is widely seen as an enabler for new computational coordination techniques for renewable, distributed generation as well as for bundling with controllable consumers. Smart markets will foster a decentralized grid management. One important task as prerequisite to decentralized management is the ability to group together in order to jointly gain enough suitable flexibility and capacity to assume responsibility for a specific control task in the grid. In self-organized smart grid scenarios, grouping or coalition formation has to be achieved in a decentralized and situation aware way based on individual capabilities. We present a fully decentralized coalition formation approach based on an established agent-based heuristics for predictive scheduling with the additional advantage of keeping all information about local decision base and local operational constraints private. Two closely interlocked optimization processes orchestrate an overall procedure that adapts a coalition structure to best suit a given set of energy products. The approach is evaluated in several simulation scenarios with different type of established models for integrating distributed energy resources and is also extended to the induced use case of surplus distribution using basically the same algorithm.