Balancing electric power in a microgrid via programmable agents auctions

The paper presents an application of an auction algorithm in a multi-agent computer system for managing the unbalanced energy in a microgrid. The main goal of the system is to control and minimize the deviations of the current energy dem and from the actual energy production, using an auction algorithm. Distributed generation is assumed in the microgrid, with renewable power sources. The energy storages and the controllable power sources improve the system operation. The differences between the actual demand and produced energy are caused by unpredictable level of electric power generation by uncontrolled sources, like wind turbines or solar panels, and/or randomness of power utilization. The system will tend to balance out these differences on-line in short time intervals (less than one minute) to follow-up varying levels of local power generation and loads.