HEMStoEC: Home Energy Management Systems to Energy Communities DataSet

The building sector is responsible for about 1/3 of all final energy consumed in the world. It is also responsible for about 30 % of CO² emissions from the end-use sector when accounting for indirect emissions from the use of electricity and heat in buildings. Focusing on EU, the use of electricity to satisfy the loads of lighting and most electrical appliances represents about 14.5 % of the energy consumed in residential sector, excluding heating and cooling systems, and 24.8 % including the latter. Therefore, we are in the presence of a sector that has a significant weight in the final energy consumption figures. Thus, innovative energy initiatives should contribute towards reducing energy consumption, reducing the effects on the climate, and achieving greater energy efficiency. These initiatives begin to emerge to a certain extent from small consumers, as they become more aware of environmental issues, either isolated or grouped in an energy community, where generated or stored energy is shared between stakeholders. In addition, energy markets go through a transition period and begin to give way, recognize, and promote the emerging role of prosumers (producers+consumers).

It is within this context that this dataset is introduced. It allows, for a single prosumer, to:

1. Test and validate different control strategies for home energy management systems;
2. Design forecasting energy consumption models;
3. Design forecasting PV energy generation models;
4. Test and validate different non-invasive load monitoring (NILM) algorithms;
5. Design forecasting thermal comfort models, as well as test and validate control strategies for Heating, Ventilation and Air Conditioning (HVAC) systems.

Additionally, for a community of 4 houses, it allows to:

1. Test and validate different control strategies for the community energy management system;
2. Design forecasting community energy consumption models;
3. Test and validate transfer learning strategies for NILM.

The data, spanning more than three years, is stored in Matlab -v7 format, . This allows to be read by other languages, such as python.