In the quest for more efficient and sustainable energy management, a novel approach to load forecasting in microgrids has emerged, offering promising implications for the energy sector. Researchers, led by Georgios Vontzos from the Department of Electrical and Computer Engineering at the University of Thessaly in Greece, have developed a new method that leverages visibility graphs to enhance the accuracy of energy demand predictions. This innovative technique, published in the journal *Electricity* (translated to English), could significantly impact how microgrids, particularly those serving regional airports, optimize their energy use and integrate renewable sources.
Microgrids, which are localized energy systems that can operate independently or in conjunction with the main power grid, are becoming increasingly important in the shift towards renewable energy. However, their effectiveness hinges on accurate load forecasting. “The ability to predict energy demand with high precision is crucial for optimizing energy management, integrating renewable sources, and reducing operational costs,” Vontzos explains. “Our research addresses this need by introducing a methodology that significantly improves forecasting accuracy.”
The proposed method transforms time series data into visibility graphs, which are then analyzed using a superposed random walk method and temporal decay adjustments. This approach weights more recent observations more heavily, providing a more accurate prediction of the next time step in the dataset. The results are impressive, with the method outperforming traditional models like Exponential Smoothing, ARIMA, Light Gradient Boosting Machine, and even advanced techniques like CNN-LSTM.
The study’s findings indicate that the proposed method achieves Symmetric Mean Absolute Percentage Error (SMAPE) and Normalized Mean Root Square Error (NMRSE) values typically ranging from 4–10% and 5–20%, respectively, with an R-squared value reaching as high as 0.96. These metrics demonstrate the method’s effectiveness in forecasting energy consumption for both stationary and highly variable time series.
The implications for the energy sector are substantial. Accurate load forecasting can lead to more efficient energy management, reduced operational costs, and better integration of renewable energy sources. For regional airports, which have diverse energy demands from various types of buildings and wind power generation, this method offers a versatile tool for optimizing energy use.
“This research lays the groundwork for further advancements in time series forecasting, including load forecasting,” Vontzos notes. “It enhances both the theoretical and practical aspects of the field, paving the way for real-world applications that can contribute to more sustainable and efficient energy systems.”
As the energy sector continues to evolve, the ability to accurately forecast energy demand will be paramount. The method developed by Vontzos and his team represents a significant step forward in this endeavor, offering a powerful tool for optimizing energy management in microgrids and beyond. The research not only advances the theoretical understanding of time series forecasting but also provides practical solutions that can be implemented in real-world scenarios, ultimately contributing to a more sustainable energy future.