Enhanced Data-Driven Load Forecasting Framework for High PV Penetration Electrical System: A Case Study of SUT’s Campus
Main Article Content
Abstract
This research proposes an approach to develop a one-day-ahead electrical load forecasting model using deep learning techniques. It compares deep learning models with basic models, namely Moving Average, DNN, LSTM, Bi-LSTM, CNN-LSTM, and Attention-LSTM, to improve accuracy and reliability through Mutual Information (MI) and Shapley Additive Explanation (SHAP) analysis. This approach utilizes data collection and feature engineering to optimize the context of an electrical loads with integrated solar power generation, Using Suranaree University of Technology (SUT) campus electrical system as a case study. Three accuracy indices are examined: Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Square Error (RMSE). Among the NN-based models, Bi-LSTM and LSTM tend to provide the best overall forecasting performance from the input data, and the Bi-LSTM model is the most accurate model with the lowest metric value among all models. The lag load feature, or the value of the electrical load in the past one day, is the feature that is most related to the forecasting target and has the most impact on the model’s decision.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
This journal provides immediate open access to its content on the principle that making research freely available to the public supports a greater global exchange of knowledge.
- Creative Commons Copyright License
The journal allows readers to download and share all published articles as long as they properly cite such articles; however, they cannot change them or use them commercially. This is classified as CC BY-NC-ND for the creative commons license.
- Retention of Copyright and Publishing Rights
The journal allows the authors of the published articles to hold copyrights and publishing rights without restrictions.
References
IEA, World Energy Outlook 2023, Int. Energy Agency, Paris, France, 2023. [Online]. Available: https://www.iea.org/reports/world-energy-outlook-2023
U.S. Energy Information Administration, Electricity explained: Factors affecting electricity consumption, 2023. [Online]. Available: https://www.eia.gov/energyexplained/electricity/
Y. Wang, Q. Chen, T. Hong, and C. Kang, Review of smart meter data analytics: Applications, methodologies, and challenges, IEEE Transactions on Smart Grid, vol. 10, no. 3, pp. 3125–3148, May 2019, doi: 10.1109/TSG.2018.2818167.
T. Panapongpakorn and D. Banjerdpongchai, Short-term load forecast for energy management system using neural networks with mutual information method of input selection, Proc. 2019 SICE International Symposium on Control Systems (SICE ISCS), Tokyo, Japan, Mar. 2019, pp. 13–18, doi: 10.23919/SICEISCS.2019.8758714.
Ministry of Energy Thailand, Thailand Smart Grid Development Plan (2022–2031), Bangkok, Thailand, 2022. [Online]. Available: https://www.energy.go.th
H. S. Hippert, C. E. Pedreira, and R. C. Souza, Neural networks for short-term load forecasting: A review and evaluation, IEEE Transactions on Power Systems, vol. 16, no. 1, pp. 44–55, Feb. 2001, doi: 10.1109/59.910780.
S. Makridakis, E. Spiliotis, and V. Assimakopoulos, Statistical and machine learning forecasting methods: Concerns and ways forward, PLOS ONE, vol. 13, no. 3, Mar. 2018, Art. no. e0194889, doi: 10.1371/journal.pone.0194889.
T. Panapongpakorn and D. Banjerdpongchai, Short-term load forecast for energy management system using neural networks with mutual information method of input selection, Proc. 2019 SICE International Symposium on Control Systems (SICE ISCS), Tokyo, Japan, Mar. 2019, pp. 13–18, doi: 10.23919/SICEISCS.2019.8758714.
S. Atef and A. B. Eltawil, Assessment of stacked unidirectional and bidirectional long short-term memory networks for electricity load forecasting, Electrical Power Systems Research, vol. 189, p. 106489, Jan. 2020, doi: 10.1016/j.epsr.2020.106489.
Y.-G. Lee, J.-Y. Oh, D. Kim, and G. Kim, SHAP value-based feature importance analysis for short-term load forecasting, Applied Science and Convergence Technology, vol. 18, pp. 579–588, Feb. 2023, doi: 10.1007/s42835-022-01161-9.
J. Jang, W. Jeong, S. Kim, B. Lee, M. Lee, and J. Moon, RAID: Robust and interpretable daily peak load forecasting via multiple deep neural networks and Shapley values, Sustainability, vol. 15, no. 8, p. 6951, Apr. 2023, doi: 10.3390/su15086951.
X. Liu, Z. Yang, Y. Guo, Z. Li, and X. Xu, A novel correlation features self-assigned Kolmogorov-Arnold Networks for multi-energy load forecasting in integrated energy systems, Energy Conversion and Management, vol. 300, p. 119388, Jan. 2024, doi: 10.1016/j.enconman.2024.119388.
Time Series Prediction Based on LSTM-Attention-LSTM Model, IEEE Access, vol. 11, pp. 48322–48331, 2023, doi: 10.1109/ACCESS.2023.3276628.
P. Wu, Z. Huang, Y. Pian, L. Xu, J. Li, and K. Chen, A Combined Deep Learning Method with Attention-Based LSTM Model for Short-Term Traffic Speed Forecasting, Journal of Advanced Transportation, vol. 2020, pp. 1–12, Nov. 2020, doi: 10.1155/2020/8863724.
K. Ullah, et al., Short-Term Load Forecasting: A Comprehensive Review and Simulation Study With CNN-LSTM Hybrids Approach, IEEE Access, vol. 12, pp. 111858–111881, 2024, doi: 10.1109/ACCESS.2024.3440631.
S. Hu, et al., Short-Term Load Forecasting Based on Mutual Information and BI-LSTM Considering Fluctuation in Importance Values of Features, IEEE Access, vol. 12, pp. 23653–23665, 2024, doi: 10.1109/ACCESS.2023.3323403.
R. Azimi, A. Amjady, and A. Daraeepour, Mutual Information-Based Inputs Selection for Electric Load Time Series Forecasting, Entropy, vol. 15, no. 3, pp. 926–942, Mar. 2013, doi: 10.3390/e15030926.
D. A. Bolstad, U. Cali, M. Kuzlu, and U. Halden, Day-ahead Load Forecasting using Explainable Artificial Intelligence, Proc. IEEE Power & Energy Society Innovative Smart Grid Technologies Conf. (ISGT), New Orleans, LA, USA, Apr. 2022, pp. 1–5, doi: 10.1109/ISGT50606.2022.9817538
M. Ramya and T. Devaraju, Dynamic scheduling and islanding energy management in microgrids using deep learning and adaptive reinforcement learning techniques, Proc. 2025 IEEE International Conference on Intelligent Systems and Sustainable Innovation (ICISS), 12–14 Mar. 2025, doi: 10.1109/ICISS63372.2025.11076495.
P. Shanmugapriya, D. R. P. R., S. V. S., and M. SureshKumar, Effective approach for energy management system using machine learning algorithm, Proc. 2024 IEEE International Conference on Power, Energy, Control and Transportation Systems (ICPECTS), 08–09 Oct. 2024, doi: 10.1109/ICPECTS62210.2024.10780045.
J. Zhang, S. Shi, and N. Zhang, Mitigation strategies for greenhouse gas emissions and energy consumption based on deep learning, Proc. 2024 IEEE International Conference on New Perspectives in Smart Industrial Future (NPSIF), 08–10 Dec. 2024, doi: 10.1109/NPSIF64134.2024.10883169.
J. Das Joy, M. R. Hossain, and A. Chowdhury, A hybrid deep learning model for day ahead short-term load forecasting, IEEE Xplore, Dec. 2024, doi: 10.1109/PEEIACON63629.2024.10800718.
M. A. Hassan, A. A. Eladl, B. E. Sedhom, and M. A. Saeed, A new optimal centralized demand side management for a campus smart microgrid, Proc. 2023 IEEE International Symposium on Industrial Electronics (ISIE), 19–21 Jun. 2023, doi: 10.1109/ISIE51358.2023.10228127.
S. Akhtar, M. Z. bin Sujod, and S. S. H. Rizvi, A novel deep learning architecture for data-driven energy efficiency management (D2EEM) – systematic survey, Proc. 2021 IEEE International Conference on Engineering and Emerging Technologies (ICEET), 27–28 Oct. 2021, doi: 10.1109/ICEET53442.2021.9659737.
