The future of energy markets and the challenge of decentralized self-management
AP2PC'08 Proceedings of the 7th international conference on Agents and Peer-to-Peer Computing
A Reinforcement Learning Approach to Setting Multi-Objective Goals for Energy Demand Management
International Journal of Agent Technologies and Systems
Hi-index | 0.00 |
Managing demand for electrical energy allows generation facilities to be run more efficiently. Current systems allow for management between large industrial consumers. There is, however, an increasing trend to decentralize energy resource management and push it to the level of individual households, or even appliances. In this work we investigate the suitability of using adaptive clustering to improve the scalability of decentralized energy resource management systems by appropriately partitioning resources. We review the area of distributed energy resource management and propose a simple yet realistic model to study the problem. Simulations using this model show that straightforward clustering and distributed planning methods allow systems to scale, but may be limited to only a few hundredthousand appliances. Results indicate that there is an opportunity to apply adaptive clustering techniques in order to discover more advanced grouping criteria that would enable groups to change as appliances' behavior changes. The simulations further suggest that even an extremely limited exchange of information between clusters can greatly improve management solutions.