Researchers from the University of the Basque Country, including B. da Costa Paulo, N. Aginako, J. Ugartemendia, I. Landa del Barrio, M. Quartulli, and H. Camblong, have developed a novel approach to optimize energy consumption in buildings, which could have significant implications for the energy industry.
The team’s work focuses on creating demand response (DR) strategies to reduce energy waste and consumption in buildings, which are among the largest energy consumers. The researchers propose using surrogate models based on active learning to decrease the computational effort required to simulate a building’s energy consumption behavior. This approach allows for a more efficient and cost-effective way to manage energy use.
The researchers first applied their active learning approach to a smaller problem: regressing the curve of voltage versus current of a thermo-resistor with reduced simulations. They then implemented a surrogate model of energy consumption in a building, aiming to learn consumption patterns based on a limited number of simulations. The goal is to use the surrogate model’s results to set the reference temperature, maximizing photovoltaic (PV) self-consumption and reducing energy usage from the grid.
The surrogate model significantly reduces the time spent mapping all possible consumption scenarios, achieving a reduction of around seven times. This efficiency gain could be particularly valuable for energy companies and building managers looking to optimize energy use and reduce costs.
The research was published in the journal Energy and Buildings, providing a valuable contribution to the field of energy efficiency and demand response strategies. The practical applications of this work could help the energy industry and building managers to reduce energy waste, lower costs, and mitigate environmental impact.
This article is based on research available at arXiv.