Digitalisation and decentralisation of the electricity grid, as well as self-consumption, have become key issues in the integration of renewable energies. In this respect, energy management systems (EMS) based on demand response (DR) in buildings with photovoltaic (PV) energy production play an important role. In the case of shared or collective self-consumption (CSC) in buildings, optimising and acting on flexible loads allows the integration of PV energy at low cost, with high reliability and improving the energy efficiency of the building.

The general objective of this line of research is the study and design of smart EMS for the CSC of electricity produced by PV panels in buildings. The objectives of EMS are diverse and can be contradictory. In this line of research, the following objectives are planned to be fully or partially addressed, depending on the application:

• Maximisation of the self-consumption rate.
• Minimisation of energy consumption.
• Ensuring the thermal comfort of the building's users.
• Maximisation of the useful life of the components on which action is to be taken - in particular heat pumps and electric vehicle (EV) batteries.
• Obtaining the highest possible economic profitability for the installation.

Depending on these objectives, the EMS should act on the main flexible loads available in buildings:

• Heating, ventilation and air-conditioning (HVAC) systems based on heat pump units.
• EV chargers installed in underground garages or in car parks around buildings.
• Electricity consumption of building users (lighting, charging of laptops, mobile phones, etc.) through awareness campaigns and communication systems.

In certain buildings, especially in residential buildings, domestic hot water generation systems and in particular their storage tanks also represent a considerable flexible inertia load. In these cases, the EMS could also act on them.

Depending on the characteristics of flexible loads, it may be appropriate for the EMS to be structured in two levels, just like EMS in microgrids:

• Day-ahead: the objective of this stage is to predict when and with what intensity the next day would need to be consumed in order to achieve the above-mentioned objectives. To this end, the EMS will generate the setpoints related to the flexible loads for the 24 h of the following day. This step has several intermediate milestones:

1. Prediction of the next day's electricity consumption of the building to be acted upon.
2. Prediction of the next day's PV production.
3. Update of the dynamic HVAC model, which includes the thermal behaviour of the building.
4. Optimisation of the thermal energy consumption and EV load setpoints according to the EMS objectives mentioned above.

• Intra-day: The objective of the intra-day EMS is to apply the setpoints generated the previous day and adjust them to cope with variations from the predictions made.

Depending on the type of flexible loads and, in general, the type of application, this two-level structure can be reduced to a single level. In any case, predictions, modelling of the flexible loads, design of optimisation algorithms and action on these loads are indispensable to achieve the objectives set for the EMS.

Within the general objective of the proposed line of research, the following specific objectives are envisaged:

1. Development of a building energy production and consumption simulator. This simulation model has to include: the PV production; the consumption in the building, including that of the flexible loads considered (HVAC, EV load and eventually water heating); the technical characteristics of the flexible loads and their environment; the connection to the distribution network; and the way in which the EMS acts on these loads, so that the effect of the EMS on the electricity consumption of the building can be simulated.
2. Development of predictive models for energy consumption and PV production in the building based on artificial intelligence (AI) and data. These will be trained with data from the above mentioned simulator and real measurement data (building consumption, PV production, meteorological data, etc.).
3. Identification of control and optimisation models for flexible loads. They will be used for the optimisation of the EMS and the control of certain flexible loads.
4. Optimisation of EMS and design of flexible load controllers. In particular, the aim is to control the thermal consumption (heating and cooling) of buildings using the model predictive control (MPC) technique.
5. Development of strategies to raise awareness and influence the energy consumption habits of building users.
6. Validation of the EMS developed in numerical simulations and in real systems.