Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran
1
Department of Architecture, Bo.C., Islamic Azad University, Borujerd, Iran
2
Department of Computer Engineering, Bo.C., Islamic Azad University, Borujerd, Iran
10.61186/jgeri.2025.2069035.1069
Abstract
This study presents a hybrid CNN-LSTM model for short-term energy consumption prediction in Iranian residential buildings, focusing on Tehran. By integrating multimodal data—meteorological, temporal, occupancy proxies, and building metadata—and employing deep feature engineering via a stacked denoising autoencoder, the model achieves high accuracy (R² = 0.89) and robustness against data imperfections. The framework demonstrates the critical role of cultural and contextual features, such as Iranian holidays, in enhancing prediction validity. SHAP analysis provides interpretability, aligning model logic with local realities. The results offer a scalable, context-aware solution for intelligent energy management in Iran’s urban environment.
Niroumand,M. , Jalili,M. and Yarahmadi,H. (2025). Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran. (e729022). Journal of Green Energy Research and Innovation, (), e729022 doi: 10.61186/jgeri.2025.2069035.1069
MLA
Niroumand,M. , , Jalili,M. , and Yarahmadi,H. . "Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran" .e729022 , Journal of Green Energy Research and Innovation, , , 2025, e729022. doi: 10.61186/jgeri.2025.2069035.1069
HARVARD
Niroumand M., Jalili M., Yarahmadi H. (2025). 'Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran', Journal of Green Energy Research and Innovation, (), e729022. doi: 10.61186/jgeri.2025.2069035.1069
CHICAGO
M. Niroumand, M. Jalili and H. Yarahmadi, "Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran," Journal of Green Energy Research and Innovation, (2025): e729022, doi: 10.61186/jgeri.2025.2069035.1069
VANCOUVER
Niroumand M., Jalili M., Yarahmadi H. Short-Term Energy Consumption Prediction in Iranian Buildings Using a Hybrid CNN-LSTM Model with Multimodal Data Fusion: A Case Study on Residential Buildings in Tehran. JGERI, 2025; (): e729022. doi: 10.61186/jgeri.2025.2069035.1069
