Sonia Leva

Presentations:

MultiGood MicroGrid and the City of the Future

Microgrids (multi-energy) and new energy management systems (EMS) represent a very up to date way to increase the penetration of renewable energy sources (RES). Multi-goods microgrids are systems with multiple production systems (PV, wind, etc.), storage capacities (batteries, hydrogen) and goods produced (electricity, heat, mobility). Being complex systems, their optimal operation is not straightforward and requires the development of dedicated control and optimization tools with relevant efforts for the modelling, analysis and design of both devices and systems. In fact, EMS play a crucial role in the control of any energy systems, and they are more and more important in grids with high penetration of RES. Optimization tools require accurate load and PV/Wind production profiles to reduce the fuel consumption or dependence on the main grid. Finally, electric vehicle (EV) integration, in terms of charging stations in distribution networks and power electronics, is totally new and raises the need for new models and methods.

PV Nowcasting and EV Forecasting

Very-short-term photovoltaic power forecast, namely nowcasting, is gaining increasing attention to face grid stability issues. All-Sky images or Satellite imaging can be used to develop the clouds movement model which might interfere with the sun beams towards a specific geographic target. However, estimating the weather conditions from these images – sun intensity, cloud appearance and movement, etc., is a very challenging task that the community has yet to solve with traditional computer vision techniques. With respect to load forecasting, a novel issue has been identified in electric vehicle (EV) forecasting, both in terms of peak power and energy. EV forecasting can be very different from load forecasting: if we have to define a daily operation plan of the charging station, it is important to know not only the power profile but also how many charging sessions will there be and when they are requested. The forecasting has become increasingly challenging and important considering small e-stations and taking into account vehicle-to-grid (V2G) scenarios and programmable charging.

PV Forecasting and Nowcasting Methods

Energy systems around the world are undergoing substantial changes, with an increasing penetration of Renewable Energy Sources (RES). For this reason, the availability of a pool of suitable forecasting models specific for the needed time horizon and task is becoming crucial in the grid operation. Novel comprehensive methodologies aiming at computing the PV power forecast at different time horizons and resolutions are introduced: moving from the 24-hours ahead prediction provided by the Artificial Neural Network and hybrid methods, a technique to refine the power forecast for the following 3 hours with an hourly granularity is analysed, leveraging on newer information available during the operation. Moreover, to provide the power forecast for the following 30 minutes on a minute-basis, an innovative modification of a statistical technique is proposed, the robust persistence. The proposed comprehensive approach allows to greatly reduce the overall error when compared with the benchmark models. Finally, a thorough analysis of new forecasting error metrics is provided. As part of the tutorial, an open-source dataset of the PV-related measurements acquired in the SolarTechLAB facility installed at Politecnico di Milano in 2017 is provided. The dataset also includes meteorological measurements to allow the integration and refinement of the adopted forecasting methods.