Economic Impact of Demand Response in the Scheduling of Distributed Energy Resources

Demand Response (DR) allows consumers to participate in energy markets, thus assuming an active role. However, the need of an aggregator capable of managing these resources and making decisions accordingly with the objectives of such resources has not been fully addressed. The aggregator activities are complex, and therefore, in the need of intelligent support to accomplish reasonable solutions. This paper proposes a methodology to evaluate the advantages of using DR programs in the resource rescheduling while classification and regression trees are introduced to support the aggregator in terms of scheduling and tariffs definition. Often these techniques are used to help the aggregator decide, as they also learn through training. Focus is given to the use of trees to predict and decide, the consumers' prices and reduction levels to apply, respectively. The case study has 548 distributed generators, 10 external suppliers and 20310 consumers.