PhD defence by Mojtaba Yousefi on Energy Management System for Smart Homes: Modeling, Control, Performance and Profit Assessment
30.10.2020 kl. 09.30 - 12.30
Mojtaba Yousefi, Department of Energy Technology, will defend the thesis "Energy Management System for Smart Homes: Modeling, Control, Performance and Profit Assessment"
Energy Management System for Smart Homes: Modeling, Control, Performance and Profit Assessment
Associate Professor Amin Hajizadeh
Associate Professor Mohsen Soltani
Associate Professor Matthias Mandø
Professor Tamas Kerekes, Dept. of Energy Technology, Aalborg University (Chairman)
Professor Mohan Lal Kolhe, Faculty of Engineering and Science, University of Agder, Norway
Associate Professor Hamid Rexa Shaker, Center for Energy Informatics, The Maersk Mc-Kinney Moller Institute, University of Southern Denmark
Recent studies of designing home energy management systems (HEMSs) have indicated that considering uncertainties and random parameters of various home resources such as photovoltaic (PV) arrays, plugged-in electric vehicles (PEV), and home load demand can significantly improve the HEMS optimal performance. Therefore, the subject of this thesis is to design a HEMS incorporating uncertainties arise from PV power output, home load demand (thermal and electricity), PEV trip time, and PEV state of charge (SOC) at plugged-in time. the contribution of this thesis is divided into three parts: 1) modeling of building resources 2) control-oriented modeling strategies and 3) performance and profit assessment of implementing the designed HEMS for building with different energy labels (determine the building storage efficiency).
A contribution of this thesis is to provide a comprehensive comparison of the existing modeling techniques such as physics-based modeling (equation-based models), data-driven or combination of them (hybrid modeling) techniques for different resources of a building. Then, these techniques are employed to capture the uncertainties of PV, home load demand (thermal and electricity) and PEV. The accuracy of the obtained models from each technique are validated by the historical data. Then, the pros and cons of each technique are presented. The results demonstrate the conditions, under which the methods can provide a reliable and accurate description of smart home dynamics. Eventually, a holistic model for the entire building is provided with considering building electrical and thermal parts. This holistic model is used by the controller to minimize the main objective of the problem while should not violate the problem constraints. In this section, some famous empirical PV models and current machine learning techniques such as artificial neural networks
The second contribution is to develop a closed-loop online optimization controller to deal with uncertainties, stochastic parameters and nonlinearities of the problem. Therefore, a predictive HEMS is designed through nonlinear model predictive control (MPC) to minimize the building cost of energy and meet the user's preference in terms of the need for electricity and thermal energy. To the best of author’s knowledge, this is the first study in the smart home context that considered the user's thermal and electrical requirements by using the following home energy storages (HESs) technologies; 1) PEV battery and 2) building thermal mass (heating/cooling the building through HP) as home energy storages. Using the following, technologies as the building storages make the system economic. Furthermore, a trade-off is made between the HEMS optimal operation and PEV battery lifetime degradation cost. The last but not least, the simulation results are validated by comparing it with an off-line optimization counterpart in which all the future inputs are known in advance.
Finally, the third contribution is to investigate the profit assessment of the designed HEMS in different buildings with different storage efficiency (different thermal resistance) which indicates by the building energy label ranges from “A” to “G”. Moreover, the impact of having different heating emission systems, which affect the building thermal capacity, is investigated as well. The simulation results prove that not even the HEMS optimal performance in building with proper storage efficiency (Label “A”) is much better than the poor storage efficiency, but also the HEMS performance in meeting the optimization constrains is much close to desire point than the building with poor insulation quality. The last but not least, it is shown that in a building with the same energy label, the floor-radiator heating system can improve the HEMS performance in both energy cost minimization and fulfilling constraints than the radiator-only heating systems, because the floor-radiator heating system increases the building thermal time constant (by improving building thermal capacity). Although, the improvements reduce as the building energy label moves to label “G”.
The defence will be in english - all are welcome
Department of Energy Technology