Master thesis: Prediction of Indoor Temperature with a Self-learning model
Are you interested in innovations in the energy sector? Would you like to learn more about digitalization of energy systems? RISE Electric Power Systems is looking for a highly motivated student with an interest in energy system and data science.
The project aims to investigate modelling approaches which can be implemented in a large-scale building level optimization of the heat demand.
Background and purpose
Demand response can play an important role for balancing the electricity production and consumption, and therefore facilitate the large-scale integration of intermittent electricity generation. A large amount of demand-side flexibility exists in heating systems. In Sweden, electrical space heating represents about 50% of residential electricity demand. Previous studies have estimated that the residential electric heating system can provide 2000-2400 MW flexibility for balancing the production variation. With enabling control and optimization strategies, such flexibility can be exploited by temporally shifting the heating load while maintaining the indoor temperature at a comfort level. It is important to have a generic modelling approach which can be easily implemented in a large scale to enable the building level optimization of the heating demand. The approach will support utilities, energy service providers or heating device manufactures for further development of enabling control in heating systems.
About the project
The aim of this project is to investigate modelling approaches for predicting the indoor temperature in buildings, which is affected by multiple factors. Such approaches can be utilized in the optimization control of modern heating systems. The thesis will start with the student gaining an understanding about present modelling methods by literature studies. The student is expected to test and compare different machine learning techniques when developing and implementing the model(s). The proposed modelling methods should be validated with public dataset and compared with existing models.
Scope: 30 hp
Location: RISE Electric Power Systems, Drottning Kristinas väg 61, Stockholm
Who are you?
For this degree project, the student should have background knowledge in heating systems, optimization, data analytics and machine learning. Possible education programs include, but are not limited to, electrical engineering, energy engineering and engineering physics.
Welcome with your application!
For more information, please contact Meng Song, Researcher, Electric Power Systems, RISE, Sweden, +46-72 212 08 69. Application deadline: 2021-09-30. The application should include your academic CV and a small description of yourself. Applications will be evaluated continuously, and the start date will be agreed with the successful applicant.
Visstidsanställning 3-6 månader
Student - examensarbete/praktik
+46-72 212 08 69
2021-09-30Skicka in din ansökan