ABSTRACT
Accurate prediction of energy demand is crucial for effective grid planning and strategy management in smart grids. This research proposes a novel BaggingSHAP model, which incorporates an ensemble of bagging algorithm with SHapley Additive exPlanations (SHAP) for improved predictive capability of the model and interpretability. The proposed model entails the collection of weather data and time-series load, data preprocessing steps and the employment of numerous machine learning (ML) models. Aside other model, the proposed model has promising results and outperforms others when evaluated on Mean Squared Error (MSE), R² Score, Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). Proposed BaggingSHAP model achieves the highest R² score of 0.9947 with an inclusion of lowest errors rate for the other metrics, thus indicating superior model predictive performance. Additionally, the BaggingSHAP provides meaningful interpretation of the feature contributions, with its corresponding outcome for the model prediction. Our investigations depict that BaggingSHAP offers an optimal balance between interpretability and accuracy score, thereby proofing it to be an efficient AI-based system for the prediction of energy demand.
References
- [1] Hong and S. Fan, “Probabilistic electric load forecasting: A tutorial review,” Int. J. Forecast., vol. 32, no. 3, pp. 914–938, Jul. 2016.
- [2] Q. Raza and A. Khosravi, “A review on artificial intelligence based load demand forecasting techniques for smart grid and buildings,” Renew. Sustain. Energy Rev., vol. 50, pp. 1352–1372, 2015.
- [3] J. Marín and F. Sandoval, “Short-term peak load forecasting: Statistical methods versus Artificial Neural Networks,” 2005.
- [4] Amral, C. S. Özveren, and D. King, “Short term load forecasting using multiple linear regression,” in Proceedings of the Universities Power Engineering Conference, 2007.
- [5] Chen, W. Wang, C. H.-E. P. S. Research, and undefined 1995, “Analysis of an adaptive time-series autoregressive moving-average (ARMA) model for short-term load forecasting,” Elsevier.
- [6] C.-I. T. on P. A. and and undefined 1971, “Short-term load forecasting using general exponential smooth ing,” ieeexplore.ieee.org.
- [7] Chakhchoukh, P. Panciatici, and L. Mili, “Electric load forecasting based on statistical robust methods,” IEEE Trans. Power Syst., 2011.
- [8] Q. Raza, Z. Baharudin, Badar-Ul-Islam, M. Azman Zakariya, and M. H. M. Khir, “Neural network based STLF model to study the seasonal impact of weather and exogenous variables,” Res. J. Appl. Sci. Eng. Technol., 2013.
- [9] K.-U. M. Thesis, undefined Industrial, and undefined 2008, “Modeling monthly electricity demand in Turkey for 1990-2006,” etd.lib.metu.edu.tr. Valova, D. Szer, N. Gueorguieva, A. B.-Neurocomputing, and undefined 2005, “A parallel growing architecture for self-organizing maps with unsupervised learning,” Elsevier.
- Abd Al-zahra, K. Moosa, and B. H. Jasim, “A comparative Study of Forecasting the Electrical Demand in Basra city using Box-Jenkins and Modern Intelligent Techniques,” Iraqi J. Electr. Electron. Eng., vol. 11, no. 1, pp. 110–123, 2015.
- R. Al-Shakarchi, M. G.-E. M. &Power, and undefined 2000, “Short-term load forecasting for baghdad electricity region,” Taylor Fr.
- Al-Hafid, G. A.-A. R. E. Journal, and undefined 2012, “Short term electrical load forecasting using holt-winters method,” iasj.net. A.Demir, “Elaboration of Electricity Energy for Production-Consumption Relation of Northern-Iraq for the Future Expectations,” Int. J. Acad. Res. Econ. Manag. Sci., vol. 3, no. 5, pp. 101–106, 2014.
- Yildiz, B., Bilbao, J., and Sproul, A., 2017, “A Review and Analysis of Regression and Machine Learning Models on Commercial Building Electricity Load Forecasting,” Renew. Sustain. Energy Rev., 73, pp. 1104–1122.
- Zhao, Z., Houchati, M., and Beitelmal, A., 2017, “An Energy Efficiency Assessment of the Thermal Comfort in an Office Building,” Energy Procedia, 134, pp. 885–893 [Sustainability in Energy and Buildings 2017: Proceedings of the Ninth KES International Conference, Chania, Greece, Jul. 5–7].
- Gu, B., Sheng, V. S., Tay, K. Y., Romano, W., and Li, S., 2015, “Incremental Support Vector Learning for Ordinal Regression,” IEEE Trans. Neural Netw. Learn. Syst., 26(7), pp. 1403–1416.
- Zhang, H., Zhao, F., and Sutherland, J. W., 2015, “Energy-Efficient Scheduling of Multiple Manufacturing Factories Under Real-Time Electricity Pricing,” CIRP Ann., 64(1), pp. 41–44.
- Freitag, M., and Hildebrandt, T., 2016, “Automatic Design of Scheduling Rules for Complex Manufacturing Systems by Multi-Objective Simulation-Based optimization,” CIRP Ann., 65(1), pp. 433–436.
- Zhai, Y., Biel, K., Zhao, F., and Sutherland, J. W., 2017, “Dynamic Scheduling of a Flow Shop With On-Site Wind Generation for Energy Cost Reduction Under Real Time Electricity Pricing,” CIRP Ann., 66(1), pp. 41–44.
Download all article in PDF
