Research results on decentralized energy management


Decentral energy management uses locally available information to optimize energy use. Efficient and cost-effective decentral agents convert local information into optimal decisions. Inspired by this challenge, a master student at the University of Chemnitz (co-supervision by Easy Smart Grid) analyzed the best approach to local information use: He identified and compared machine learning algorithms with respect to their ability to predict energy availability and usage patterns. These results were correlated with the resources required to implement such algorithms on typical (embedded) microcontrollers. He also developed a scheduling scheme based on Markov Decision Models to control flexible loads for minimum energy cost. Download Master Thesis or Download presentation

« Back to News                             » Next Article