AI-BASED SMART BUILDING ENERGY MANAGEMENT SYSTEM USING DEEP LEARNING AND EXPLAINABLE AI

Author

KAVIYA S, DR. JAI RUBY

Abstract

 This paper presents a smart building energy management system for next-day household energy consumption forecasting and occupancy-aware energy recommendations. Accurate forecasting can support the planning of building operations and help identify opportunities to reduce avoidable energy use. The study evaluates seven supervised learning models: Random Forest, XGBoost, Gradient Boosting, Long Short-Term Memory (LSTM), bidirectional LSTM (BiLSTM), CNN-LSTM, and an LSTM+CNN-LSTM-Attention hybrid. The models use historical consumption, calendar, weather, and appliance-activity features. Data preprocessing includes the removal of records with unusable values and the construction of lagged, rolling-window, and cyclical temporal features. The data are partitioned chronologically into training, validation, and test sets to reduce information leakage across time. Model performance is assessed using mean absolute error (MAE), root mean squared error (RMSE), mean absolute percentage error (MAPE), and the coefficient of determination (R²). Based on the results reported in Table 1, the deployed LSTM achieves an RMSE of 8.017021 kWh (8.02 kWh rounded), the lowest among the evaluated models, and is retained as the project’s final forecasting model. The system integrates this forecast with YOLOv8-based person detection to incorporate room occupancy into energy-saving recommendations. SHAP and permutation-importance methods are included to support interpretation of the prediction factors. The proposed application combines forecasting, occupancy detection, and recommendation generation within a unified workflow. Further evaluation using additional household data across different buildings and seasons is needed to assess the system’s generalisability.

Keywords

Smart Building Energy Management, Deep Learning, Energy Forecasting, LSTM, Explainable AI, SHAP, YOLOv8, Occupancy Detection, Comparative Analysis, Regression.

Full Text:

Download Paper PDF

References

[1] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997. DOI: 10.1162/neco.1997.9.8.1735.

[2] L. Breiman, “Random Forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001. DOI: 10.1023/A:1010933404324.

[3] T. Chen and C. Guestrin, “XGBoost: A Scalable Tree Boosting System,” in Proc. 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, 2016, pp. 785–794. DOI: 10.1145/2939672.2939785.

[4] S. M. Lundberg and S.-I. Lee, “A Unified Approach to Interpreting Model Predictions,” in Advances in Neural Information Processing Systems (NeurIPS), 2017, pp. 4765–4774.

[5] G. Jocher, A. Chaurasia, and J. Qiu, “Ultralytics YOLOv8,” Ultralytics, 2023. [Online]. Available: https://github.com/ultralytics/ultralytics.

[6] R. Mathumitha, P. Rathika, and K. Manimala, “Intelligent deep learning techniques for energy consumption forecasting in smart buildings: a review,” Artificial Intelligence Review, vol. 57, article 35, 2024. DOI: 10.1007/s10462-023-10660-8.

[7] J. Runge and R. Zmeureanu, “A Review of Deep Learning Techniques for Forecasting Energy Use in Buildings,” Energies, vol. 14, no. 3, 608, 2021. DOI: 10.3390/en14030608.

[8] P. Henkel, T. Kasperski, P. Stoffel, and D. Müller, “Interpretable data-driven model predictive control of building energy systems using SHAP,” Proceedings of the 6th Annual Learning for Dynamics & Control Conference, PMLR 242, pp. 222–234, 2024.

[9] A. Forootani, M. Rastegar, and A. Sami, “Short-term individual residential load forecasting using an enhanced machine learning-based approach based on a feature engineering framework: A comparative study with deep learning methods,” Electric Power Systems Research, vol. 210, article 108119, 2022. DOI: 10.1016/j.epsr.2022.108119.

[10] C. Deb, F. Zhang, J. Yang, S. E. Lee, and K. W. Shah, “A review on time series forecasting techniques for building energy consumption,” Renewable and Sustainable Energy Reviews, 2017. DOI: 10.1016/j.rser.2017.02.085

Share your valuable work from Social Media Buttons