In the heart of Poland, researchers are pioneering a new approach to optimize energy production and consumption in buildings, with significant implications for the energy sector. Andrzej Marciniak, from the Department of Transportation and Informatics at WSEI University in Lublin, has developed an innovative algorithm that could revolutionize how we understand and manage energy in smart buildings. His work, published in the journal Energies (translated from Polish as ‘Energies’), focuses on determining the energy production and consumption signatures of a university administration building equipped with a 50 kWp photovoltaic (PV) system.
The study begins with a seemingly simple task: assessing whether the chosen PV system adequately meets the building’s energy demands. However, Marciniak’s approach is anything but straightforward. He employs an advanced PV inverter, a device connected to the Internet of Things and a smart metering system, to collect and analyze time-series data on power production and consumption. “The key is to understand the behavior of these systems,” Marciniak explains. “By analyzing the data, we can identify patterns and optimize energy management.”
The research involves traditional statistical analyses, dividing the generated power into self-consumption and grid-fed power, and the consumed power into PV-produced and grid-sourced power. But here’s where it gets interesting: Marciniak uses unsupervised clustering, specifically the k-means algorithm, to divide the power generation and consumption space into distinct states. These states are then categorized based on their nature and usefulness in managing power.
The results are presented as heatmaps, which allow for the localization of specific states at different times of the day. This visualization leads to a better understanding and quantification of the building’s energy dynamics. “The heatmaps provide a clear picture of when and how power is generated and consumed,” Marciniak notes. “This can help in making informed decisions about energy management.”
One of the most significant findings is the confirmation of the potential use of an energy storage system. By identifying surplus power states, the algorithm can help determine the optimal size and operation strategy for such a system. This has substantial commercial implications for the energy sector, as it can lead to more efficient use of renewable energy sources and reduced reliance on the grid.
The algorithm also serves as a basis for further analysis, including the prediction of power generated by renewable energy sources and the energy consumed by various types of buildings. This could pave the way for more intelligent, adaptive energy management systems in the future.
Marciniak’s work is a testament to the power of data-driven decision-making in the energy sector. By leveraging advanced analytics and machine learning techniques, we can gain deeper insights into energy production and consumption patterns. This, in turn, can lead to more efficient, sustainable, and cost-effective energy management strategies. As the energy sector continues to evolve, such innovations will be crucial in meeting the challenges of the future.