AI based profit strategies in a smart energy market

Abstract

This master’s thesis proposes an approach to predict prices in a smart grid’s energy trading market using artificial neural networks to maximize the profit of a participating household. Because energy demand and supply must be balanced at all times to guarantee the stability of the energy grid, producers, wholesalers and brokers are equally interested in keeping the equilibrium to prevent blackouts, poor customer satisfaction and the respective costs. In the past couple of years, the ever-expanding energy demand and the addition of volatile renewable resources into the energy system has created new challenges. Recent advances in information technologies, on the other hand, introduce possibilities of how to tackle these challenges ahead. One approach is to decentralize the energy distribution and having smart microgrids balancing the demand and supply within themselves, stabilizing the whole system in the process. To do so, the Smart Stability Network incorporates an auction scheme where each participating household can offer to produce or consume energy for a price. Because market participants usually act selfishly and want to increase their profits, knowing the ideal price to win the auction in advance would enable them to maximize their profit and ensure their active participation in stabilizing the grid....