Evaluation of distributed multi-agent Energy Management System - cost calculation

An Energy management system (EMS) is a concept that spans various possible solutions, ranging from basic implementations over solutions that use simple intelligent computer methods to systems that employ advanced intelligent methods. We designed and implemented an intelligent multi-agent based approach that uses a market mechanism to manage power in the microgrid of a simulated Research and Education Center. In this article, the performance of our system regarding cost optimality is compared against two other solutions: the first is a simple solution that uses predefined profiles that define the use of the controllable sources this provides an upper limit to the cost, the second is the perfect artificial optimum which provides a lower limit to the cost. The perfect optimum resembles a centrally controlled EMS, with the difference that it does not suffer from the delays of detecting power imbalances. The tests show that distributed approach closely resembles the optimal one.