Towards Concise Models of Grid Stability

Decentral Smart Grid Control (DSGC) is a new system implementing demand response without significant changes of the infrastructure. It does so by binding the electricity price to the grid frequency. While models of DSGC exist, they rely on various simplifying assumptions. For example, researchers have assumed that the behavior of all participants in the grid is identical. In this paper we study how data-mining techniques can help to remove some of these simplifications, while keeping the representation of the insights concise. We systematically collect the various assumptions and identify questions regarding the system that are still open. Next, we run many simulations, with diverse input values. Finally, we apply decision trees to the resulting data and show that this indeed provides new insights. For example, we discover that the system can be stable even if some participants adapt their energy consumption with a high delay, or that fast adaptation is preferable for system stability.