We aim to harness the power of data-driven approaches, specifically Bayesian optimization and sequential learning, to optimize energy systems and materials. This innovative approach holds immense potential for improving the performance and efficiency of various energy technologies, including novel photovoltaic (PV) cells, photosynthetic processes for solar fuel generation, and energy storage materials.
Bayesian optimization offers a robust framework for exploring and exploiting the design space of complex energy systems and materials. By leveraging prior knowledge and incorporating real-time data, Bayesian optimization enables the efficient identification of optimal solutions while minimizing the number of costly and time-consuming experiments. This approach is particularly valuable in the field of energy technology, where the design space is vast and traditional trial-and-error methods may be prohibitively expensive. Sequential learning complements Bayesian optimization by actively selecting the most informative data points to improve the optimization process iteratively. It allows for the systematic acquisition of data, adaptation of models, and refinement of strategies, leading to accelerated convergence towards optimal solutions. By iteratively learning from data, sequential learning enables researchers to uncover hidden relationships, identify critical parameters, and guide the design and synthesis of energy materials and systems with enhanced performance.
Referenti:
Prof. Eliodoro Chiavazzo
Prof. Matteo Fasano
Prof. Pietro Asinari
Link: https://www.denerg.polito.it/it/la_ricerca/gruppi_di_ricerca/gruppo_di_ricerca_m3es