Development of DR energy management optimization at building and district level using GA and NN modeling power predictionsDevelopment of DR energy management optimization at building and district level using GA and NN modeling power predictions
Μεταπτυχιακή Διατριβή
Master Thesis
2018-06-202018enIn broad terms, Demand Response refers to the operational, regulatory and technical framework for inducing changes in the power demand of buildings or settlements during the day. Time of Use (ToU) pricing can be vital to leverage advancements in building or district energy management systems to shift loads, exploit storage capabilities, increase renewable energy penetration and ultimately relief stress from the grid. This is an important feature of the smart grid and a step closer to the necessary open and transparent market framework according to which energy consumption costs reflect actual costs of production, transmission, distribution, infrastructure maintenance and upgrade etc. In this paper Neural Network power predictions are performed and a genetic algorithm based framework for energy management in a group of buildings is developed and tested on real data.
According to the results ToU pricing could be exploited by the industry using ANN based day ahead prediction to perform load shifting and minimize associated costs.http://creativecommons.org/licenses/by-nc/4.0/Πολυτεχνείο Κρήτης::Σχολή Μηχανικών ΠεριβάλλοντοςTsekeri_Elisavet_MSc_2018.pdfChania [Greece]Library of TUC2018-06-20application/pdf5.3 MBfree
Tsekeri Elisavet
Τσεκερη Ελισαβετ
Kolokotsa Dionysia
Κολοκοτσα Διονυσια
Kalaitzakis Konstantinos
Καλαϊτζακης Κωνσταντινος
Karatzas Giorgos
Καρατζας Γιωργος
Πολυτεχνείο Κρήτης
Technical University of Crete
Genetic algorithms
Artificial neural network
Demand response