Predicting energy consumption of an educational building using group method of data handling based on weather data

• This study uses A1B and B1 scenarios to predict energy consumption and stands out by combining these scenarios with energy prediction models, improving prediction accuracy. • The A1B scenario predicts a temperature increase from 25°C to 30°C , while the B1 scenario forecasts a smaller increase from 25°C to 27°C , leading to significant differences in energy consumption. • The GMDH-PNN model achieved the most accurate energy consumption prediction with an MSE of 1.01 , outperforming ANN and SVM models, which had MSE values of 1.35 and 1.42 , respectively. • The GMDH-PNN model achieved the most accurate energy consumption prediction with an MSE of 1.01 , outperforming ANN and SVM models, which had MSE values of 1.35 and 1.42 , respectively. • This study shows that predicting temperature changes can help in designing long-term energy management strategies and optimized energy systems to address future climate challenges. This study presents a novel approach to predicting energy consumption in educational buildings by combining advanced energy prediction models, including the Group Method of Data Handling-Probabilistic Neural Network (GMDH-PNN), Artificial Neural Network (ANN), and Support Vector Machine (SVM), with climate change scenarios (A1B,B1). While previous research predominantly focuses on conventional methods using historical data, this study innovatively incorporates future temperature projections to assess the impact of climate change on energy demand. The GMDH-PNN model demonstrated superior accuracy in energy prediction, achieving the lowest Mean Squared Error (MSE) of 1.01 for recent data (2007–2021), compared to MSE values of 1.35 for ANN and 1.42 for SVM. Additionally, this study examined the effects of climate change on energy consumption, predicting an increase in demand as temperatures rise.Under Scenario A1B, with a temperature rise from 25°C to 30°C over the next 25 years, energy consumption is expected to increase by 50%, reaching 150 kWh by 2050.Under Scenario B1, with a moderate temperature rise to 27°C, energy consumption will rise to 120 kWh by 2050. The novelty of this work lies in its integration of climate change projections with energy consumption models, providing a comprehensive framework to predict energy trends under varying temperature conditions. The findings suggest that,under future climate scenarios,energy demand will increase significantly, particularly for cooling systems. However, by incorporating climate projections into energy management strategies, we can design and implement energy-efficient systems that are adaptable to changing environmental conditions, offering a hopeful solution to the challenges posed by climate change.