To enhance the accuracy of ultra-short-term photovoltaic power forecasting under meteorological disturbance scenarios, a hybrid model integrating an improved rime optimization algorithm (IRIME), a temporal convolutional network (TCN), a multi-strategy optimization temporal channel attention (MSOTCA) mechanism, and a bidirectional gated recurrent unit (BiGRU) was proposed. Firsty, key meteorological features were selected using Pearson correlation analysis and fed into the TCN for preliminary temporal feature extraction. Next, an MSOTCA mechanism was designed by integrating multi-scale temporal information, meteorological ramp enhancement, and residual connections to strengthen the model’s dynamic responsiveness to complex disturbance patterns. Subsequently, an IRIME algorithm enhanced with a meteorological-fluctuation-aware dynamic convergence factor was developed to automatically optimize the model hyperparameters. The BiGRU was then employed as the temporal decision layer to output point forecasts. Finally, prediction intervals were generated via Monte Carlo sampling to quantify uncertainty. Experiments were conducted using measured data from photovoltaic power stations in Yanggao, Shanxi and Hami, Xinjiang. The results demonstrate that compared to benchmark models such as Transformer and Informer, the proposed model achieves average reductions in mean absolute error (MAE) and root mean square error (RMSE) ranging from 19.20% to 59.54% and from 9.22% to 49.57%, respectively. These results indicate superior prediction accuracy and stability, validating the effectiveness and robustness of the proposed method under meteorological disturbance scenarios.
TAN Hua-qiang, ZHANG Chao, QIN Tao, CHEN Chang-sheng, GUO Yuan-ping, YANG Jing.
Ultra-short-term Photovoltaic Power Prediction Using TCN-BiGRU with Improved Optimization and Attention Mechanism[J].
Science Technology and Engineering, 2026, 26(21): 9156-9168 DOI:10.12404/j.issn.1671-1815.2506105
预处理后的数据输入时域卷积网络(temporal convolutional network,TCN)。TCN采用因果卷积确保仅依赖历史信息,避免未来数据泄露;并通过膨胀卷积扩展感受野,空洞因子按指数规律增长(如1,2,4,…),使模型能够有效捕获不同时间尺度的长短期依赖关系。每个卷积块内含两层卷积,其间加入修正线性单元(rectified linear unit, ReLU)激活与Dropout层以抑制过拟合;块间采用残差连接,缓解梯度消失,提升深层网络训练稳定性。
WangQi, MaLi. Spatial differentiation and endogenous driving mechanisms of China’s energy transition[J]. Resources Science, 2024, 46(9): 1709-1722.
[3]
GuptaK A, SinghK R. A review of the state of the art in solar photovoltaic output power forecasting using data-driven models[J]. Electrical Engineering, 2024, 107(4): 1-44.
ChenQian, BaiJing, ChenXiaoyue, et al. Review of the impact of distributed PV grid integration on the power grid[J]. Science Technology and Engineering, 2024, 24(27): 11491-11504.
[6]
HuZ, GaoY, JiS, et al. Improved multistep ahead photovoltaic power prediction model based on LSTM and self-attention with weather forecast data[J]. Applied Energy, 2024, 359: 122709.
FengPeiru, JiangGuifen, XuJiayin, et al. Short-term photovoltaic power prediction based on CEEMD-BiLSTM-RFR[J]. Science Technology and Engineering, 2024, 24(5): 1955-1962.
DongCun, WangZheng, BaiJieyu, et al. Review of ultra-short-term forecasting methods for photovoltaic power generation[J]. High Voltage Engineering, 2023, 49(7): 2938-2951.
ZhengXiaoliang, WangHu. Short term photovoltaic power generation prediction based on similar day clustering combination model[J]. Science Technology and Engineering, 2025, 25(27): 11632-11641.
WangDongfeng, LiQingbo, ZhangBoyang, et al. research on photovoltaic power prediction model considering seasonality and trend characteristics[J]. Acta Energiae Solaris Sinica, 2025, 46(3): 348-356.
ChenSiyu, JiangJunxia, LiShuaibing. Review of photovoltaic power prediction methods under extreme climate conditions[J]. Journal of Electrical Engineering, 2025, 20(1): 281-290.
RuanGuangchun, HeYiliu, TanZhenfei, et al. Review of hybrid data-driven and physics-based modeling for the operation of new-type power systems[J]. Proceedings of the CSEE, 2024, 44(13): 5021-5037.
[19]
CasanovaH R, CondeA, PérezS C. Hour-ahead photovoltaic power prediction combining BiLSTM and bayesian optimization algorithm, with bootstrap resampling for interval predictions[J]. Sensors, 2024, 24(3): 882.
[20]
LiuM, WangX, ZhongZ. Ultra-short-term photovoltaic power prediction based on BiLSTM with wavelet decomposition and dual attention mechanism[J]. Electronics, 2025, 14(2): 306.
[21]
ShiY, ZhangL, WangS, et al. A short-term photovoltaic power interval forecasting method based on fuzzy granular computing and CNN-BiGRU[J]. International Journal of Low-Carbon Technologies, 2024, 19: 306-314.
[22]
WangJ, JiaM, LiS, et al. Short-term power-generation prediction of high humidity island photovoltaic power station based on a deep hybrid model[J]. Sustainability, 2024, 16(7): 2853.
CaiYuan, WuHao, TangDan. Probabilistic prediction of photovoltaic power interval based on feature mining and improved TCN-BiGRU[J]. Science Technology and Engineering, 2025, 25(10): 4145-4155.
LuoXiaoyuan, LiuJie, YangBin, et al. Ultra short-term wind power prediction based on improved osprey optimization algorithm and VMD-LSTM[J]. Acta Energiae Solaris Sinica, 2025, 46(3): 652-660.
[27]
ZhouN, ShangB, ZhangJ, et al. Research on prediction method of photovoltaic power generation based on transformer model[J]. Frontiers in Energy Research, 2024, 12: 1452173.
[28]
QuanR, QiuZ, WanH, et al. Dung beetle optimization algorithm-based hybrid deep learning model for ultra-short-term PV power prediction[J]. iScience, 2024, 27(11): 111126.
[29]
DaiY, WangY, ChenY, et al. Combining meteorological and power information of station-measurement and model-prediction with the hybrid CNN-Transformer and CNN-BiLSTM for ultra-short-term photovoltaic power forecasting[J]. International Journal of Electrical Power and Energy Systems, 2025, 171: 111009.
[30]
MahdiS M, MichelaL. High-resolution PV power prediction model based on the deep learning and attention mechanism[J]. Sustainable Energy, Grids and Networks, 2023, 34: 101025.
ZhouWen, GaoQiang, MaoZemin, et al. Multivariable time-sequential short-term photovoltaic power interval prediction based on improved AT-QRCNN-BiLSTM model[J]. Electric Engineering, 2024(22): 72-79.
[33]
ChenL, ZhangH, XiaoJ, et al. SCA-CNN: spatial and channel-wise attention in convolutional networks for image captioning[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI, USA: IEEE, 2017: 6298-6306.
[34]
HangS, DongZ, AsgharA H, et al. RIME: a physics-based optimization[J]. Neurocomputing, 2023, 532: 183-214.