An MT-Transformer Framework for Coordinated Wind–Solar–Load Forecasting with Net-Load-Based Coal-Power Regulation Demand Identification
DOI:
https://doi.org/10.4108/ew.13035Keywords:
MT-Transformer, joint wind–solar–load forecasting, multi-task learning, net-load fluctuation, coal-power regulation demandAbstract
Data-driven forecasting has become increasingly important for describing the temporal interactions among heterogeneous variables in modern power systems. To capture the nonlinear coupling, heterogeneous fluctuations, and multi-scale temporal variations between renewable generation and load demand, this study develops an MT-Transformer framework for coordinated wind–solar–load forecasting. Meteorological variables, historical renewable output, historical load, and temporal labels are integrated as model inputs. A shared Transformer encoder is used to learn common temporal representations, and task-specific forecasting heads are designed to generate synchronized predictions for wind power, PV power, and load. The experimental results show that MT-Transformer achieves an MAE of 0.0848, an RMSE of 0.1254, and an R² of 0.9302. Compared with the Persistence Model, the MAE and RMSE decrease by 36.05% and 30.49%, respectively. The predicted outputs are further converted into a net-load sequence, from which fluctuation indicators are derived. The peak–valley difference reaches 0.462 p.u., and the maximum ramp rate reaches 0.087 p.u./h, indicating evident peak-shaving pressure and short-term regulation demand. These findings confirm that the proposed framework improves coordinated forecasting performance and provides quantitative evidence for coal-power peak regulation, reserve capacity allocation, and ancillary service demand identification.
Downloads
References
[1] Zhang S, Chen W. Assessing the energy transition in China towards carbon neutrality with a probabilistic framework[J]. Nature Communications, 2022, 13: 87.
[2] An K, Zheng X, Shen J, et al. Repositioning coal power to accelerate net-zero transition of China’s power system[J]. Nature Communications, 2025, 16(1): 2400.
[3] Yin G, Li B, Fedorova N, et al. Orderly retire China’s coal-fired power capacity via capacity payments to support renewable energy expansion[J]. iScience, 2021, 24(11): 103287.
[4] Lin B, Liu Z. Assessment of flexible coal power and battery energy storage system in supporting renewable energy[J]. Energy, 2024, 313: 133805.
[5] Hirth L. The market value of variable renewables: The effect of solar wind and solar power variability on their relative price[J]. Energy Economics, 2013, 38: 218-236.
[6] Tziolis G, Livera A, Theocharides S, et al. Net load forecasting: A comprehensive literature review[J]. Sustainable Energy Technologies and Assessments, 2025, 82: 104450.
[7] Kerkau S, Sepasi S, Howlader H O R, et al. Day-ahead net load forecasting for renewable integrated buildings using XGBoost[J]. Energies, 2025, 18(6): 1518.
[8] Telle J S, Upadhaya A, Schönfeldt P, et al. Probabilistic net load forecasting framework for application in distributed integrated renewable energy systems[J]. Energy Reports, 2024, 11: 2535-2553.
[9] Song Z, Gu Y, Liu H, et al. Application of deep learning in wind, solar, and ocean energy: An analysis of prediction, optimization, and operation & maintenance[J]. Renewable and Sustainable Energy Reviews, 2026, 230: 116663.
[10] Hasan M, Mifta Z, Papiya S J, et al. A state-of-the-art comparative review of load forecasting methods: Characteristics, perspectives, and applications[J]. Energy Conversion and Management: X, 2025, 26: 100922.
[11] Malakouti M. Machine learning for renewable energy forecasting: Advances, challenges, and future directions[J]. Sustainable Energy Research, 2026, 13: 14.
[12] Gupta M, Arya A, Varshney U, et al. A review of PV power forecasting using machine learning techniques[J]. Progress in Engineering Science, 2025, 2(1): 100058.
[13] Box G E P, Jenkins G M, Reinsel G C, et al. Time Series Analysis: Forecasting and Control[M]. 5th ed. Hoboken: Wiley, 2015.
[14] Ding J W, Chuang M J, Tseng J S, et al. Reanalysis and ground station data: Advanced data preprocessing in deep learning for wind power prediction[J]. Applied Energy, 2024, 375: 124129.
[15] Cheng J, Luo X, Jin Z. Integrating domain knowledge into Transformer for short-term wind power forecasting[J]. Energy, 2024, 312: 133511.
[16] Piantadosi G, Dutto S, Galli A, et al. PV power forecasting: A Transformer based framework[J]. Energy and AI, 2024, 18: 100444.
[17] Hussan U, Wang H, Peng J, et al. Transformer-based renewable energy forecasting: A comprehensive review[J]. Renewable and Sustainable Energy Reviews, 2026, 226: 116356.
[18] Hu J, Hu W, Cao D, et al. Probabilistic net load forecasting based on transformer network and Gaussian process-enabled residual modeling learning method[J]. Renewable Energy, 2024, 225: 120253.
[19] Ahmad A, Xiao X, Mo H, et al. TFTformer: A novel transformer based model for short-term load forecasting[J]. International Journal of Electrical Power & Energy Systems, 2025, 166: 110549.
[20] Pandžić F, Capuder T. Advances in short-term solar forecasting: A review and benchmark of machine learning methods and relevant data sources[J]. Energies, 2023, 17(1): 97.
[21] Wang Y, Zou R, Liu F, et al. A review of wind speed and wind power forecasting with deep neural networks[J]. Applied Energy, 2021, 304: 117766.
[22] Manandhar P, Rafiq H, Rodriguez-Ubinas E, et al. New forecasting metrics evaluated in Prophet, Random Forest, and Long Short-Term Memory models for load forecasting[J]. Energies, 2024, 17(23): 6131.
[23] Taylor S J, Letham B. Forecasting at scale[J]. The American Statistician, 2018, 72(1): 37-45.
[24] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems. 2017, 30: 5998-6008.
[25] Wu X, Zhan H, Hu J, et al. Non-stationary GNNCrossformer: Transformer with graph information for non-stationary multivariate spatio-temporal wind power data forecasting[J]. Applied Energy, 2025, 377: 124492.
[26] Vandenhende S, Georgoulis S, Van Gansbeke W, et al. Multi-task learning for dense prediction tasks: A survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(7): 3614-3633.
[27] Dou X, Wang Y, Zhang H, et al. An adaptive forecasting method for net load ramping demand in power systems[J]. IET Generation, Transmission & Distribution, 2025, 19(1): e70176.
[28] Verdone A, Galletti A, Mazzoni S, et al. A review of solar and wind energy forecasting: From single-site to spatio-temporal deep learning approaches[J]. Applied Energy, 2025, 392: 125843.
[29] Lim B, Zohren S, Roberts S. Time-series forecasting with deep learning: A survey[J]. Philosophical Transactions of the Royal Society A, 2021, 379(2194): 20200209.
[30] Tao K, Zhang X, Wang Y, et al. Operational day-ahead photovoltaic power forecasting using transformer-based deep learning models[J]. Applied Energy, 2024, 373: 123986.
[31] Liu H, Zhang Z, Wang Y, et al. Probabilistic forecasting of multiple plant day-ahead renewable power generation sequences with data privacy preserving[J]. Energy and AI, 2025, 19: 100463.
[32] Yang T, Yang Z, Li F, et al. A short-term wind power forecasting method based on multivariate signal decomposition and variable selection[J]. Applied Energy, 2024, 360: 122759.
[33] Caruana R. Multi-task learning[J]. Machine Learning, 1997, 28(1): 41-75.
[34] Hyndman R J, Athanasopoulos G. Forecasting: Principles and Practice[M]. 3rd ed. Melbourne: OTexts, 2021.
[35] Zhou H, Zhang S, Peng J, et al. Informer: Beyond efficient Transformer for long sequence time-series forecasting[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2021, 35(12): 11106-11115.
[36] Bashir T, Wang H, Tahir M, et al. Wind and solar power forecasting based on hybrid CNN-ABiLSTM, CNN-transformer-MLP models[J]. Renewable Energy, 2025, 239: 122055.
[37] Lim B, Arık S Ö, Loeff N, et al. Temporal Fusion Transformers for interpretable multi-horizon time series forecasting[J]. International Journal of Forecasting, 2021, 37(4): 1748-1764.
[38] Liu L, Zhang J, Xue S. Photovoltaic power forecasting: Using wavelet threshold denoising combined with VMD[J]. Renewable Energy, 2025, 249: 123152.
[39] Zhai C, Zhang Y, Liu X, et al. Photovoltaic power forecasting based on VMD-SSA-Transformer: Multidimensional analysis of dataset length, weather mutation and forecast accuracy[J]. Energy, 2025, 324: 135971.
[40] Wang B, Chen J, Zhu Y, et al. SP-Transformer: A medium- and long-term photovoltaic power forecasting model integrating multi-source spatiotemporal features[J]. Applied Sciences, 2025, 15(21): 11846.
[41] Wu M, Feng W, Li X, et al. Short-term power load forecasting using an improved model integrating GCN and Transformer[J]. Applied Sciences, 2025, 15(13): 7003.
[42] Lafuente-Cacho M, Martínez S, López G, et al. State of the art for solar and wind energy-forecasting methods for sustainable grid integration[J]. Journal of the Knowledge Economy, 2025, 16: 16432-16467.
[43] Feng W, Deng B, Chen T, et al. Probabilistic net load forecasting based on sparse variational Gaussian process regression[J]. Frontiers in Energy Research, 2024, 12: 1429241.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Meng Huang, Lei Wang, Teng Luo, Chao Ju, Wenchao Guo, Chenghao Ning
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This is an open-access article distributed under the terms of the Creative Commons Attribution CC BY 4.0 license, which permits unlimited use, distribution, and reproduction in any medium so long as the original work is properly cited.
