As a critical technology for enhancing wind energy utilization efficiency, wind power forecasting requires extensive historical data for high-accuracy models. Addressing data scarcity in new wind farms, this study proposes a transfer learning-based LSTM-GRU hybrid model. An optimal feature window preserved temporal dynamics and suppressed redundant noise within a multidimensional feature matrix. The cross-domain framework employs LSTM in data-rich source domains and lightweight GRU in data-scarce targets. Source-domain LSTM parameters transfer to enhance temporal modeling, with transferred layers frozen and only GRU layers fine-tuned, balancing knowledge transfer and domain adaptation. Experimental results show the proposed transfer method reduced MAE by 18.8% and 34.5%, and RMSE by 19.0% and 32.1%, outperforming conventional single-domain models. Freezing transferred parameters decreased trainable parameters, accelerating convergence speed by 26.9% and 17.9%. This study offers managerial support for efficient new wind farm commissioning, improved grid dispatch, and more reliable investment decisions.
AzimiRGhofraniMGhayekhlooM (2016) A hybrid wind power forecasting model based on data mining and wavelets analysis. Energy Conversion and Management127: 208–225.
2.
DoD-PNLeeYChoiJ (2016) Hourly average wind speed simulation and forecast based on ARMA model in Jeju Island, Korea. Journal of Electrical Engineering and Technology11(6): 1548–1555.
3.
El-FoulyTHMEl-SaadanyEFSalamaMMA, et al. (2006) Grey predictor for wind energy conversion systems output power prediction. IEEE Transactions on Power Systems21(3): 1450–1452.
4.
HuSXiangYZhangH, et al. (2021) Hybrid forecasting method for wind power integrating spatial correlation and corrected numerical weather prediction. Applied Energy293: 116951.
5.
HuYHuXYaoX, et al. (2025) Ultra-short-term prediction for wind power via intelligent reductional reconfiguration of wind conditions and upgraded stepwise modelling with embedded feature engineering. Renewable Energy240: 122155.
6.
JalaliSMJAhmadianSKhodayarM, et al. (2022) An advanced short-term wind power forecasting framework based on the optimized deep neural network models. International Journal of Electrical Power & Energy Systems141: 108143.
7.
KhalidMSavkinAV (2012) A method for short-term wind power prediction with multiple observation points. IEEE Transactions on Power Systems27(2): 579–586.
8.
LahouarABen Hadj SlamaJ (2017) Hour-ahead wind power forecast based on random forests. Renewable Energy109: 529–541.
9.
LiZLuoXLiuM, et al. (2022) Wind power prediction based on EEMD-Tent-SSA-LS-SVM. Energy Reports8: 3234–3243.
10.
LiaoQCaoDChenZ, et al. (2023) Probabilistic wind power forecasting for newly-built wind farms based on multi-task gaussian process method. Renewable Energy217: 119054.
11.
LiuXZhouJQianH (2021) Short-term wind power forecasting by stacked recurrent neural networks with parametric sine activation function. Electric Power Systems Research192: 107011.
12.
LiuXLiuJZhaoY, et al. (2024) A bayesian deep learning-based adaptive wind farm power prediction method within the entire life cycle. IEEE Transactions on Sustainable Energy15(4): 2663–2674.
13.
LuJBehboodVHaoP, et al. (2015) Transfer learning using computational intelligence: a survey. Knowledge-Based Systems80: 14–23.
14.
LydiaMSuresh KumarSImmanuel SelvakumarA, et al. (2016) Linear and non-linear autoregressive models for short-term wind speed forecasting. Energy Conversion and Management112: 115–124.
15.
MeiJWangCLuoS, et al. (2024) Short-term wind power prediction based on encoder–decoder network and multi-point focused linear attention mechanism. Sensors24(17): 5501.
16.
MengAChenSOuZ, et al. (2022) A novel few-shot learning approach for wind power prediction applying secondary evolutionary generative adversarial network. Energy261: 125276.
17.
NguyenHKMPhanQ-DWuY-K, et al. (2023) Multi-step wind power forecasting with stacked temporal convolutional network (S-TCN). Energies16(9): 3792.
18.
RiedelPBelkilaniKReichertM, et al. (2024) Enhancing PV feed-in power forecasting through federated learning with differential privacy using LSTM and GRU. Energy and AI18: 100452.
19.
WangJAlShelahiAYouM, et al. (2021) Integrative density forecast and uncertainty quantification of wind power generation. IEEE Transactions on Sustainable Energy12(4): 1864–1875.
20.
WangDYangMZhangW (2023) Wind power group prediction model based on multi-task learning. Electronics12(17): 3683.
21.
WangYZhaoKHaoY, et al. (2024) Short-term wind power prediction using a novel model based on butterfly optimization algorithm-variational mode decomposition-long short-term memory. Applied Energy366: 123313.
22.
XiuCGuoF (2013) Wind speed prediction by chaotic operator network based on kalman filter. Science China Technological Sciences56(5): 1169–1176.
23.
YangMGuoYHuangT, et al. (2025) Power prediction considering NWP wind speed error tolerability: a strategy to improve the accuracy of short-term wind power prediction under wind speed offset scenarios. Applied Energy377: 124720.
24.
YinHLiCChenS, et al. (2025) Few-shot wind power prediction using sample transfer and imbalanced evolved neural network. Energy328: 136375.
25.
ZhaoZYunSJiaL, et al. (2023) Hybrid VMD-CNN-GRU-based model for short-term forecasting of wind power considering spatio-temporal features. Engineering Applications of Artificial Intelligence121: 105982.
26.
ZhaoHXuPGaoT, et al. (2024) CPTCFS: CausalPatchTST incorporated causal feature selection model for short-term wind power forecasting of newly built wind farms. International Journal of Electrical Power & Energy Systems160: 110059.
27.
ZhaoYPanSChenY, et al. (2025) Intraday wind power forecasting by ensemble of overlapping historical numerical weather predictions. IEEE Transactions on Sustainable Energy16(2): 1315–1328.
