The increasing penetration of renewable energy, the use of storage systems, the adoption of cogeneration and trigeneration, and the possibility of establishing energy communities in a context of high price volatility require advanced tools to define the optimal management of energy systems. This is essential to achieve maximum profitability from existing equipment and to better plan future investments.
We have a decade of experience in developing numerical codes that robustly and efficiently address the optimal management of multi-good systems, supported and validated through extensive experimental activity at our Multi-Good MicroGrid Laboratory (MG2Lab). Specifically, our activities include:
- development and testing of advanced optimisation methods that consider every possible uncertain outcome in the microgrid, adopt adaptive optimal dispatch policies instead of fixed control structures, allow the best exploitation of renewable resources and guarantee reliable and efficient operation;
- design and optimal management of integrated energy systems, focusing on synergies between energy sectors, integration of electric vehicles and provision of ancillary services;
- deployment of predictive multi-layer Energy Management Systems (EMS) for the optimal scheduling of dispatchable generators, predictive operation planning of storage systems and correction rules to deal with forecast errors in real time;
- load forecasting by means of computational intelligence techniques (deep learning, such as LSTM and Bi-LSTM), a key point for estimating the expected baseline of the microgrid;
- day-ahead PV forecasting with Physical Hybrid Artificial Neural Networks (PHANN) and nowcasting based on image processing techniques.
Projects
Latest publications
Loading latest publications…






