Е-Публикации
Технически университет - София

Abstract: This study presents a hybrid price time series forecasting system based on four models: XGBoost, LSTM, Prophet, and ARIMA. The system integrates temperature data as an exogenous regressor to improve predictive accuracy. A graphical environment for data loading, model training, and results visualization is implemented. Metrics such as MAE, MSE, MAPE, WAPE, and Pearson coefficient are used to evaluate the effectiveness. Additionally, SHAP analysis interprets the importance of input features. The results show that combining classical and deep models increases the robustness and precision of forecasts.

References

1. Luo, Xing & Zhu, Xu & Lim, Eng. (2018). A hybrid model for short term real-time electricity price forecasting in smart grid. Big Data Analytics. 3. 10.1186/s41044-018-0036-x.
2. Khalid, R., Javaid, N., Al-zahrani, F. A., Aurangzeb, K., Qazi, E.-U.-H., & Ashfaq, T. (2020). Electricity Load and Price Forecasting Using Jaya-Long Short Term Memory (JLSTM) in Smart Grids. Entropy, 22(1), 10. https://doi.org/10.3390/e22010010
3. Wang, Z., Mae, M., Yamane, T., Ajisaka, M., Nakata, T., & Matsuhashi, R. (2024). Enhanced Day-Ahead Electricity Price Forecasting Using a Convolutional Neural Network–Long Short-Term Memory Ensemble Learning Approach with Multimodal Data Integration. Energies, 17(11), 2687. https://doi.org/10.3390/en17112687
4. Aguiar-Pérez, J. M., & Pérez-Juárez, M. Á. (2023). An Insight of Deep Learning Based Demand Forecasting in Smart Grids. Sensors, 23(3), 1467. https://doi.org/10.3390/s23031467
5. Cheng, Zhengzhe & Xu, Hanyi & Wang, Yuetong & Zhou, Hao. (2023). Time Series Forecasting Based on ARIMA and LSTM Models. 359-364. 10.1109/PEEEC60561.2023.00076.
6. J. Contreras, R. Espinola, F. J. Nogales and A. J. Conejo, "ARIMA models to predict next-day electricity prices," in IEEE Transactions on Power Systems, vol. 18, no. 3, pp. 1014-1020, Aug. 2003, doi: 10.1109/TPWRS.2002.804943.
7. Chen, Tianqi & Guestrin, Carlos. (2016). XGBoost: A Scalable Tree Boosting System. 785-794. 10.1145/2939672.2939785.
8. RS, Rajasree & K, Yashwanth & P, Sree & No, Yashas D & D'Souza, Sonia. (2024). Time Series Forecasting For Energy Consumption Using XGBoost and LSTM. 1-8. 10.1109/APCIT62007.2024.10673558.
9. Gassar, A. A. A. (2024). Short-Term Energy Forecasting to Improve the Estimation of Demand Response Baselines in Residential Neighborhoods: Deep Learning vs. Machine Learning. Buildings, 14(7), 2242. https://doi.org/10.3390/buildings14072242
10. Ahmed, Shamim & Kaiser, M. Shamim & Hossain, Mohammad & Andersson, Karl. (2024). A Comparative Analysis of LIME and SHAP Interpreters with Explainable ML-Based Diabetes Predictions. IEEE Access. PP. 1-1. 10.1109/ACCESS.2024.3422319.
11. Weron, Rafał. (2014). Electricity price forecasting: A review of the state-of-the-art with a look into the future. International Journal of Forecasting. 30. 10.1016/j.ijforecast.2014.08.008.
12. Neumann, Oliver & Turowski, Marian & Mikut, Ralf & Hagenmeyer, V. & Ludwig, Nicole. (2023). Using weather data in energy time series forecasting: the benefit of input data transformations. Energy Informatics. 6. 10.1186/s42162-023-00299-8.
13. Taylor, S. J., & Letham, B. (2018). Forecasting at Scale. The American Statistician, 72(1), 37–45. https://doi.org/10.1080/00031305.2017.1380080
14. Lundberg, Scott & Lee, Su-In. (2017). A Unified Approach to Interpreting Model Predictions. 10.48550/arXiv.1705.07874.
15. Cortez, Juan & Terada, Lucas & Bandeira, Bruno & Soares, João & Vale, Zita & Rider, Marcos J.. (2023). Comparative Analysis of ARIMA, LSTM, and XGBoost for Very Short-Term Photovoltaic Forecasting. 1-6. 10.1109/SEPOC58810.2023.10322588.

Issue

Copyright IEEE