Short-Term Wind and Solar Power Forecasting Method Based on Closed-Loop Empirical Mode Decomposition and Temporal Convolutional Feature Extraction

doi:10.16183/j.cnki.jsjtu.2024.431

Abstract

Abstract:

The strong randomness and volatility inherent in wind and solar power generation results in single forecasting models insufficient for capturing their nonlinear and non-stationary characteristics. To address this issue, this paper proposes a combined short-term forecasting model for wind and solar power based on closed-loop complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and temporal convolutional feature extraction. First, CEEMDAN is applied to decompose the original wind and solar power generation data across multiple time scales, extracting high-energy intrinsic mode functions to effectively reduce the non-stationarity and complexity of the data. Then, a temporal convolutional network (TCN) is employed to extract the temporal features of the wind and solar power data, while a bidirectional gated recurrent unit (BiGRU) further captures the bidirectional dynamic characteristics of the time series. An Attention mechanism is also integrated to enhance the focus on key features, thereby constructing the TCN-BiGRU-Attention forecasting model. Additionally, the salp swarm algorithm (SSA) is introduced to optimize the hyperparameters of the model, and a feedback mechanism is used to dynamically adjust the decomposition parameters of CEEMDAN, further improving the forecasting performance. Finally, experiment results based on wind and solar power generation data from a certain region demonstrate that, compared with other forecasting models, the proposed method improves coefficient of determination R² by 2.35% and 2.92%, respectively, effectively enhancing the prediction accuracy.

Key words: power forecasting, closed-loop empirical mode decomposition, salp swarm algorithm (SSA), temporal convolutional feature extraction, deep learning

CLC Number:

TM732

WANG Danhao, PENG Daogang, HUANG Dongmei, LIU Yu. Short-Term Wind and Solar Power Forecasting Method Based on Closed-Loop Empirical Mode Decomposition and Temporal Convolutional Feature Extraction[J]. Journal of Shanghai Jiao Tong University, 2026, 60(7): 1087-1098.

Figures/Tables 13

Fig.1

Fig.2

Fig.3

Fig.4

Tab.1

Tab.2

Tab.3

Fig.5

Fig.6

Fig.7

Tab.4

Tab.5

Tab.6

References 25

[1]	黄强, 郭怿, 江建华, 等. “双碳”目标下中国清洁电力发展路径[J]. 上海交通大学学报, 2021, 55(12): 1499-1509. doi: 10.16183/j.cnki.jsjtu.2021.272
	HUANG Qiang, GUO Yi, JIANG Jianhua, et al. Development pathway of China’s clean electricity under carbon peaking and carbon neutrality goals[J]. Journal of Shanghai Jiao Tong University, 2021, 55(12): 1499-1509.
[2]	郑珂, 王丽婕, 郝颖, 等. 基于数据集蒸馏的光伏发电功率超短期预测[J]. 中国电机工程学报, 2024, 44(13): 5196-5207.
	ZHENG Ke, WANG Lijie, HAO Ying, et al. Ultra-short-term prediction of photovoltaic power based on dataset distillation[J]. Proceedings of the CSEE, 2024, 44(13): 5196-5207.
[3]	朱琼锋, 李家腾, 乔骥, 等. 人工智能技术在新能源功率预测的应用及展望[J]. 中国电机工程学报, 2023, 43(8): 3027-3048.
	ZHU Qiongfeng, LI Jiateng, QIAO Ji, et al. Application and prospect of artificial intelligence technology in renewable energy forecasting[J]. Proceedings of the CSEE, 2023, 43(8): 3027-3048.
[4]	曹俊波, 周任军, 邓学华, 等. 考虑优化ARIMA模型差分次数的风功率预测[J]. 电力系统及其自动化学报, 2019, 31(1): 105-111.
	CAO Junbo, ZHOU Renjun, DENG Xuehua, et al. Wind power forecast considering differential times of optimal ARIMA model[J]. Proceedings of the CSU-EPSA, 2019, 31(1): 105-111.
[5]	董存, 王铮, 白捷予, 等. 光伏发电功率超短期预测方法综述[J]. 高电压技术, 2023, 49(7): 2938-2951.
	DONG Cun, WANG Zheng, BAI Jieyu, et al. Review of ultra-short-term forecasting methods for photovoltaic power generation[J]. High Voltage Engineering, 2023, 49(7): 2938-2951.
[6]	KHAN S, MUHAMMAD Y, JADOON I, et al. Leveraging LSTM-SMI and ARIMA architecture for robust wind power plant forecasting[J]. Applied Soft Computing, 2025, 170: 112765. doi: 10.1016/j.asoc.2025.112765 URL
[7]	吴硕. 光伏发电系统功率预测方法研究综述[J]. 热能动力工程, 2021, 36(8): 1-7.
	WU Shuo. Review of power forecasting methods of photovoltaic power generation system[J]. Journal of Engineering for Thermal Energy & Power, 2021, 36(8): 1-7.
[8]	彭嘉宁, 徐鹤勇. 基于随机森林和支持向量回归的风力发电预测算法[J]. 热能动力工程, 2024, 39(5): 143-149.
	PENG Jianing, XU Heyong. Wind power prediction algorithm based on random forest and support vector regression[J]. Journal of Engineering for Thermal Energy & Power, 2024, 39(5): 143-149.
[9]	杨雷, 罗雪红, 韩鹍, 等. 基于人工智能算法的电力系统负荷预测研究综述[J]. 电工技术, 2024(12): 57-60.
	YANG Lei, LUO Xuehong, HAN Kun, et al. A review of power system load forecasting based on artificial intelligence algorithms[J]. Electric Engineering, 2024(12): 57-60.
[10]	邓斌, 张楠, 王江, 等. 基于LTC-RNN模型的中长期电力负荷预测方法[J]. 天津大学学报, 2022, 55(10): 1026-1033.
	DENG Bin, ZHANG Nan, WANG Jiang, et al. Medium-and long-term power load forecasting method based on LTC-RNN model[J]. Journal of Tianjin University, 2022, 55(10): 1026-1033.
[11]	李芬, 孙凌, 王亚维, 等. 基于CEEMDAN-GSA-LSTM和SVR的光伏功率短期区间预测[J]. 上海交通大学学报, 2024, 58(6): 806-818. doi: 10.16183/j.cnki.jsjtu.2022.511
	LI Fen, SUN Ling, WANG Yawei, et al. Short-term interval forecasting of photovoltaic power based on CEEMDAN-GSA-LSTM and SVR[J]. Journal of Shanghai Jiao Tong University, 2024, 58(6): 806-818.
[12]	ZHANG C, PENG T, NAZIR M S. A novel integrated photovoltaic power forecasting model based on variational mode decomposition and CNN-BiGRU considering meteorological variables[J]. Electric Power Systems Research, 2022, 213: 108796. doi: 10.1016/j.epsr.2022.108796 URL
[13]	梁宏涛, 刘红菊, 李静, 等. 基于机器学习的短期负荷预测算法综述[J]. 计算机系统应用, 2022, 31(10): 25-35.
	LIANG Hongtao, LIU Hongju, LI Jing, et al. Survey on short-term load forecasting algorithm based on machine learning[J]. Computer Systems & Applications, 2022, 31(10): 25-35.
[14]	WU Q Y, GUAN F, LÜ C, et al. Ultra-short-term multi-step wind power forecasting based on CNN-LSTM[J]. IET Renewable Power Generation, 2021, 15(5): 1019-1029. doi: 10.1049/rpg2.v15.5 URL
[15]	PENG D G, LIU Y, WANG D H, et al. Multi-energy load forecasting for integrated energy system based on sequence decomposition fusion and factors correlation analysis[J]. Energy, 2024, 308: 132796. doi: 10.1016/j.energy.2024.132796 URL
[16]	WANG H, ZHANG N, DU E S, et al. A comprehensive review for wind, solar, and electrical load forecasting methods[J]. Global Energy Interconnection, 2022, 5(1): 9-30. doi: 10.1016/j.gloei.2022.04.002 URL
[17]	崔京港, 王芳, 叶泽甫, 等. 基于QD和因果注意力TCN的光伏功率区间预测[J]. 太阳能学报, 2024, 45(3): 488-495.
	CUI Jinggang, WANG Fang, YE Zefu, et al. Photovoltaic power interval prediction based on QD and causal attention TCN[J]. Acta Energiae Solaris Sinica, 2024, 45(3): 488-495.
[18]	雷柯松, 吐松江·卡日, 伊力哈木·亚尔买买提, 等. 基于WGAN-GP和CNN-LSTM-Attention的短期光伏功率预测[J]. 电力系统保护与控制, 2023, 51(9): 108-118.
	LEI Kesong, KARI Tusongjiang, YAERMAIMAITI Yilihamu, et al. Prediction of short-term photovoltaic power based on WGAN-GP and CNN-LSTM-Attention[J]. Power System Protection & Control, 2023, 51(9): 108-118.
[19]	LIU W H, BAI Y L, YUE X X, et al. A wind speed forcasting model based on rime optimization based VMD and multi-headed self-attention-LSTM[J]. Energy, 2024, 294: 130726. doi: 10.1016/j.energy.2024.130726 URL
[20]	SHI J, TEH J. Load forecasting for regional integrated energy system based on complementary ensemble empirical mode decomposition and multi-model fusion[J]. Applied Energy, 2024, 353: 122146. doi: 10.1016/j.apenergy.2023.122146 URL
[21]	梁恩豪, 孙军伟, 王延峰. 基于自适应樽海鞘算法优化BP的风光互补并网发电功率预测[J]. 电力系统保护与控制, 2021, 49(24): 114-120.
	LIANG Enhao, SUN Junwei, WANG Yanfeng. Wind and solar complementary grid-connected power generation prediction based on BP optimized by a swarm intelligence algorithm[J]. Power System Protection & Control, 2021, 49(24): 114-120.
[22]	HU C J, ZHAO Y, JIANG H, et al. Prediction of ultra-short-term wind power based on CEEMDAN-LSTM-TCN[J]. Energy Reports, 2022, 8: 483-492.
[23]	KARIJADI I, CHOU S Y, DEWABHARATA A. Wind power forecasting based on hybrid CEEMDAN-EWT deep learning method[J]. Renewable Energy, 2023, 218: 119357. doi: 10.1016/j.renene.2023.119357 URL
[24]	DENG Y, YUE Z P, WU Z Y, et al. TCN-Attention-BIGRU: Building energy modelling based on attention mechanisms and temporal convolutional networks[J]. Electronic Research Archive, 2024, 32(3): 2160-2179. doi: 10.3934/era.2024098 URL
[25]	GONG J Q, QU Z G, ZHU Z L, et al. Ensemble models of TCN-LSTM-LightGBM based on ensemble learning methods for short-term electrical load forecasting[J]. Energy, 2025, 318: 134757. doi: 10.1016/j.energy.2025.134757 URL

能源	R²
能源	K_IMF=5	K_IMF=6	K_IMF=7	K_IMF=8	K_IMF=9	K_IMF=10	K_IMF=11	K_IMF=12	K_IMF=13
风	0.9434	0.9632	0.9666	0.9727	0.9798	0.9735	0.9729	0.9718	0.9706
光	0.9329	0.9518	0.9612	0.9691	0.9774	0.9823	0.9784	0.9761	0.9755

能源	P_n
能源	IMF1	IMF2	IMF3	IMF4	IMF5	IMF6	IMF7	IMF8	IMF9	IMF10
风	0.0059	0.0074	0.031	0.0303	0.0751	0.0608	0.0849	0.0639	0.6408
光	0.0124	0.0083	0.0101	0.0151	0.0872	0.3413	0.1285	0.0297	0.0147	0.3527

能源	R²
能源	P_th>0	P_th>1%	P_th>3%	P_th>5%	P_th>10%	P_th>30%
风	0.9618	0.9765	0.9765	0.9798	0.9032	0.9032
光	0.9711	0.9798	0.9823	0.9823	0.9616	0.9514

发电类型	RMSE/kW	MAE /kW	R²
风力发电	1674.7352	1039.5361	0.9798
光伏发电	490.1308	301.0943	0.9823

预测模型	风力发电评估指标			光伏发电评估指标			模型性能指标
预测模型	RMSE/kW	MAE/kW	R²	RMSE/kW	MAE/kW	R²	参数量	FLOPs
LSTM	2752	1771	0.9193	777	457	0.9311	2901	20
Transformer	3558	2982	0.8652	871	542	0.9134	86069	8064020
BiLSTM	2255	1523	0.9459	811	465	0.9250	5801	40
LSTM-Attention	2960	1975	0.9066	1049	689	0.8744	2922	1940
TCN	2369	1669	0.9402	750	355	0.9358	2401	222740
TCN-LSTM	2199	1501	0.9485	764	371	0.9335	14357	1903124
VMD-LSTM	2124	1424	0.9489	618	444	0.9544	2901	20
EMD-LSTM	2534	1646	0.9316	742	362	0.9373	2901	20
VMD-GRU	1943	1262	0.9573	702	496	0.9411	2181	20
EMD-GRU	2517	1825	0.9325	769	383	0.9326	2181	20
VMD-TCN	2383	1751	0.9357	661	466	0.9478	2401	222740
EMD-TCN	2964	2059	0.9064	762	396	0.9337	2401	222740
所提模型	1675	1040	0.9798	490	301	0.9823	22317	2619204