Abstract—Managing energy consumption use at airports is becoming more critical, but there’s not much research on airfield lighting systems. This research proposes a hybrid machine learning framework designed to estimate the nominal output current demand by considering weather conditions and time of day. The hybrid model combines two techniques, Gaussian Process Regression (GPR) and K-Nearest Neighbors. This allows it to recognize overall trends and specific changes in energy use. This study also uses data segmentations (25%, 50%, 75% and 100%) and various acquisition functions such as Expected Improvement (EI), Expected Improvement Plus (EIP), Expected Improvement Per Second (EIPS), Expected Improvement Per Second Plus (EIPSP), Lower Confidence Bound (LCB), and Probability of Improvement (POI) to improve predictions. To identify the most effective model, metrics including accuracy, precision, recall, and F1-Score were employed for evaluation. Findings from the study indicate that both data segmentation and the selected acquisition function play a crucial role in influencing model performance. Notably, the model is able to work well even with limited data, meaning it can still be effective even if only a portion of the data is available. This makes it efficient, reducing both costs and time needed to train the model without giving up on accuracy. The solution is lightweight model, making it a practical choice compared to complex models that are hard to interpret. Ultimately, this research presents a new, scalable approach to optimizing energy use in airfield lighting, which can help create more innovative and sustainable airport operations.
 
Keywords—acquisition function, airside energy management, machine learning, nominal output current prediction, sustainable aviation


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).