|
| |
|
|
Predictive Control of DHW and Household Heating Using Machine Learning Methods
Necpál, Dávid ; Zemčík, Pavel (referee) ; Materna, Zdeněk (advisor)
The thesis is devoted to the development of a system for the control of domestic hot water (DHW) and heating. The system uses machine learning methods to predict future data - learning and predicting repetitive routines of people in the home, predicting hot water consumption and temperature trends in the home. The work is divided into two parts, where the first part is the control of domestic hot water heating, and the second part is the control of home heating. Both parts of the system use established hardware devices that implement the data collection and the actual control of the given activities. The devices collect water temperatures from the electric boiler, air temperatures from the rooms of the household and information about the presence of people in the household. The measured data is used to create a model for data prediction, which is used for the final decision on whether DHW heating or home heating should be implemented. The first subsystem for DHW heating achieves energy savings of 20 to 24 %, the second subsystem for domestic heating 18 to 30 %. The benefit of the developed system is the possibility of predictive automatic control of domestic hot water heating and heating, which leads to a reduction of energy consumption required for the above-mentioned activities, as well as a possible increase in comfort in the home.
|
|
| |
guest ::
