Balancing Accuracy and Interpretability: GradientSHAP for Enhanced Energy Demand Predictions

ABSTRACT

The growing amalgamation of renewable sources of energy in power systems has increased the need for accurate energy demand prediction within smart grids. Recent progress in machine learning has improved predictive capabilities; however, most of these models are complex in structure and lack interpretability. This study proposes a novel GradientSHAP which fuses gradient boosting algorithms with SHAP (SHapley Additive exPlanations) values to enhance predictive performance while improving model interpretability. GradientSHAP is developed to capture complex ad non-linear structure in the time-series and weather data for a robust energy demand predictions. SHAP values are computed together with the boosting algorithm to provide meaningful information into the impact of the individual features on the model predictions. The European energy demand dataset is utilized in this study to evaluate the proposed GradientSHAP, and the model performance is compared with traditional models such as linear regression and support vector regression (SVR). GradientSHAP outweighs these traditional models, obtaining the lowest training and test Mean Squared Error (MSE) and the highest R-squared (R²) score, demonstrating optimal predictive capability. Detailed and concise explanation of feature contributions is presented via SHAP plots to enhance model transparency. The proposed GradientSHAP achieves a significant milestone in energy demand prediction and demonstrates a substantial ability to balance high predictive accuracy and interpretability without a trade-off, which is essential in predicting energy demand in smart grids.

References

• [1] Zambrano, A. F., and Giraldo, L. F., 2020, “Solar Irradiance Forecasting Models Without On-Site Training Measurements,” Renew. Energy, 152, pp. 557–566.
• [2] Sobri, S., Koohi-Kamali, S., and Rahim, N. A., 2018, “Solar Photovoltaic Generation Forecasting Methods: A Review,” Energy Convers. Manage., 156, pp. 459–497.
• [3] Voyant, C., Notton, G., Kalogirou, S., Nivet, M.-L., Paoli, C., Motte, F., and Fouilloy, A., 2017, “Machine Learning Methods for Solar Radiation Forecasting: A Review,” Renew. Energy, 105, pp. 569–582.
• [4] Agüera-Pérez, A., Palomares-Salas, J. C., González de la Rosa, J. J., and Florencias-Oliveros, O., 2018, “Weather Forecasts for Microgrid Energy Management: Review, Discussion and Recommendations,” Appl. Energy, 228, pp. 265–278.
• [5] Qazi, A., Fayaz, H., Wadi, A., Raj, R. G., Rahim, N., and Khan, W. A., 2015, “The Artificial Neural Network for Solar Radiation Prediction and Designing Solar Systems: A Systematic Literature Review,” J. Cleaner Prod., 104, pp. 1–12.
• [6] Antonanzas, J., Osorio, N., Escobar, R., Urraca, R., de Pison, F. M., and Antonanzas-Torres, F., 2016, “Review of Photovoltaic Power Forecasting,” Sol. Energy, 136, pp. 78–111.
• [7] Das, U. K., Tey, K. S., Seyedmahmoudian, M., Mekhilef, S., Idris, M. Y. I., Van Deventer, W., Horan, B., and Stojcevski, A., 2018, “Forecasting of Photovoltaic Power Generation and Model Optimization: A Review,” Renew. Sustain. Energy Rev., 81, pp. 912–928.
• [8] Yadav, A. K., and Chandel, S., 2014, “Solar Radiation Prediction Using Artificial Neural Network Techniques: A Review,” Renew. Sustain. Energy Rev., 33, pp. 772–781.
• [9] Özge, A., and Ümmühan, B. F., 2018, “Estimation Methods of Global Solar Radiation, Cell Temperature and Solar Power Forecasting: A Review and Case Study in Eskişehir,” Renew. Sustain. Energy Rev., 91, pp. 639–653.
• Zendehboudi, A., Baseer, M., and Saidur, R., 2018, “Application of Support Vector Machine Models for Forecasting Solar and Wind Energy Resources: A Review,” J. Cleaner Prod., 199, pp. 272–285.
• Huang, C. M., Huang, Y. C., and Huang, K. Y., 2014, “A Hybrid Method for One-Day Ahead Hourly Forecasting of PV Power Output,” 2014 9th IEEE Conference on Industrial Electronics and Applications, Hangzhou, China, June 9–11, pp. 526–531.
• Lorenz, E., Hurka, J., Heinemann, D., and Beyer, H. G., 2009, “Irradiance Forecasting for the Power Prediction of Grid-Connected Photovoltaic Systems,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 2(1), pp. 2–10.
• Raza, M. Q., Nadarajah, M., and Ekanayake, C., 2016, “On Recent Advances in PV Output Power Forecast,” Sol. Energy, 136, pp. 125–144.\
• Huang, C., Chen, S.-J., Yang, S.-P., and Kuo, C.-J., 2015, “One-Day-Ahead Hourly Forecasting for Photovoltaic Power Generation Using an Intelligent Method With Weather-Based Forecasting Models,” IET Gener. Transm. Distrib., 9, pp. 1874–1882.
• Rozas Larraondo, P., Inza, I., and Lozano, J. A., 2018, “A System for Airport Weather Forecasting Based on Circular Regression Trees,” Environ. Model. Softw., 100, pp. 24–32.
• Qing, X., and Niu, Y., 2018, “Hourly Day-Ahead Solar Irradiance Prediction Using Weather Forecasts by LSTM,” Energy, 148, pp. 461–468.
• Akarslan, E., ad Hocaoglu, F. O., 2017, “A Novel Method Based on Similarity for Hourly Solar Irradiance Forecasting,” Renew. Energy, 112, pp. 337–346.
• Gigoni, L., Betti, A., Crisostomi, E., Franco, A., Tucci, M., Bizzarri, F., and Mucci, D., 2018, “Day-Ahead Hourly Forecasting of Power Generation From Photovoltaic Plants,” IEEE Trans. Sustain. Energy, 9(2), pp. 831–842.
• Semero, Y. K., Zhang, J., and Zheng, D., 2018, “PV Power Forecasting Using an Integrated GA-PSO-ANFIS Approach and Gaussian Process Regression Based Feature Selection Strategy,” CSEE J. Power Energy Syst., 4(2), pp. 210–218.
• Zhang, Y., Beaudin, M., Taheri, R., Zareipour, H., and Wood, D., 2015, “Day-Ahead Power Output Forecasting for Small-Scale Solar Photovoltaic Electricity Generators,” IEEE Trans. Smart Grid, 6(5), pp. 2253–2262.

Download all article in PDF

WSN 199 (2025) 142-155

ADVERTISEMENT