Ultra-short-term and Short-term Power Load Single-step and Multi-step Prediction Considering Spatial Association

doi:10.19678/j.issn.1000-3428.0069490

Abstract

Abstract:

Accurate ultra-short-term and short-term multi-region power load forecasting is the key to achieving rapid response and real-time scheduling in power systems. Therefore, based on the spatiotemporal correlation of loads in different regions of a power grid, this study proposes a single-step and multi-step prediction model for ultrashort-term and short-term power load forecasting in multiple regions. This model integrates a Gate-controlled Multi-head Temporal Convolutional Network (GMTCN), a Bi-directional Long Short-Term Memory (BiLSTM), and an attention mechanism, denoted as GMTCN-BiLSTM-Attention. First, the Spearman correlation coefficient is used to analyze the spatial correlation of power loads in different regions, and the load sequences of 15 regions are combined into a multivariate time series to be used as input. Then, GMTCN and BiLSTM are employed to obtain the temporal features and spatiotemporal dependencies of different load sequences, and the attention mechanism is applied to assign higher weights to important features, ignoring unimportant information to improve the robustness of the model. Experiments on two datasets reveal a spatiotemporal correlation between the power loads in different regions. The proposed model can effectively obtain the temporal characteristics of load sequences and the spatiotemporal dependencies among load sequences, and it can simultaneously achieve single- and multi-step predictions of ultra-short-term and short-term power loads in multiple regions. Compared with other deep learning models, it has better predictive performance, stronger robustness, and improved generalization.

Key words: multi-region power load forecasting, deep learning, Gate-controlled Multi-head Temporal Convolutional Network (GMTCN), Bi-directional Long Short Term Memory (BiLSTM) network, attention mechanism

摘要：

准确的超短期、短期多区域电力负荷预测是实现电力系统快速响应和实时调度的关键。基于电网不同区域负荷的时空相关性，提出考虑空间关联的多区域电力负荷超短期、短期的单步和多步预测模型。该模型集成门控多头时间卷积网络(GMTCN)、双向长短期记忆(BiLSTM)网络和注意力机制(Attention)，记为GMTCN-BiLSTM-Attention。首先，采用Spearman相关系数分析不同区域电力负荷空间关联，将15个区域的负荷序列组成多元时间序列作为输入。然后，采用GMTCN和BiLSTM获取不同负荷序列的时序特征和时空依赖，并通过Attention机制赋予重要特征更高的权重，忽略不重要的信息，以提高模型的鲁棒性。在2个数据集上的实验结果表明，不同区域变压器的负荷之间存在时空相关性，提出的模型能够有效获取负荷序列以及负荷序列之间的时空依赖，同时实现对多区域的超短期和短期负荷进行单步和多步预测。与其他深度学习模型相比，具有更优的预测性能、更强的鲁棒性和泛化性。

关键词: 多区域电力负荷预测, 深度学习, 门控多头时间卷积网络, 双向长短期记忆网络, 注意力机制

GENG Zhenwei, LI Shenzhang, YU Fengrong. Ultra-short-term and Short-term Power Load Single-step and Multi-step Prediction Considering Spatial Association[J]. Computer Engineering, 2025, 51(10): 392-400.

耿贞伟, 李申章, 于凤荣. 考虑空间关联的超短期、短期多区域电力负荷单步和多步预测[J]. 计算机工程, 2025, 51(10): 392-400.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0069490

https://www.ecice06.com/EN/Y2025/V51/I10/392

Figures/Tables 12

Fig.1 Multi-region load forecasting framework

Fig.2 Causal expansion convolution structure of TCN

Fig.3 GMTCN model structure

Fig.4 BiLSTM model structure

Fig.5 Topological structure of 15 transformers

Fig.6 Spearman correlation coefficient

Fig.7 Prediction curves of different prediction time steps

Fig.8 The predicted values and true values of different models on the test set for one week

References 25

1	李练兵, 高国强, 陈伟光, 等. 考虑特征重组和BiGRU-Attention-XGBoost模型的超短期负荷功率预测. 现代电力, 2025, 42(3): 571- 581.
	LI L B, GAO G Q, CHEN W G, et al. Ultra short term load power prediction considering feature recombination and BiGRU-Attention-XGBoost model. Modern Electric Power, 2025, 42(3): 571- 581.
2	陈宋宋, 王阳, 周颖, 等. 基于客户用电数据的多时空维度负荷预测综述. 电网与清洁能源, 2023, 39(12): 28- 40.
	CHEN S S, WANG Y, ZHOU Y, et al. Review of multi-temporal and multi-dimensional load forecasting based on customer electricity consumption data. Power System and Clean Energy, 2023, 39(12): 28- 40.
3	WANG Z Y, ZHOU X J, TIAN J T, et al. Hierarchical parameter optimization based support vector regression for power load forecasting. Sustainable Cities and Society, 2021, 71, 102937. doi: 10.1016/j.scs.2021.102937
4	康重庆, 夏清, 张伯明. 电力系统负荷预测研究综述与发展方向的探讨. 电力系统自动化, 2004(17): 1- 11.
	KANG C Q, XIA Q, ZHANG B M. Review of power system load forecasting research and discussion of development direction. Automation of Electric Power Systems, 2004(17): 1- 11.
5	NIE Y, JIANG P, ZHANG H P. A novel hybrid model based on combined preprocessing method and advanced optimization algorithm for power load forecasting. Applied Soft Computing, 2020, 97, 106809. doi: 10.1016/j.asoc.2020.106809
6	HUANG Q, LI J H, ZHU M S. An improved convolutional neural network with load range discretization for probabilistic load forecasting. Energy, 2020, 203, 117902. doi: 10.1016/j.energy.2020.117902
7	丁国辉, 刘宇琪, 王言开, 等. 基于翻转网络的低相关性序列数据预测研究. 计算机工程, 2024, 50(2): 78- 90. doi: 10.19678/j.issn.1000-3428.0067027
	DING G H, LIU Y Q, WANG T K, et al. Research on low-correlation sequence data prediction based on flip network. Computer Engineering, 2024, 50(2): 78- 90. doi: 10.19678/j.issn.1000-3428.0067027
8	李丹, 张远航, 杨保华, 等. 基于约束并行LSTM分位数回归的短期电力负荷概率预测方法. 电网技术, 2021, 45(4): 1356- 1364.
	LI D, ZHANG Y H, YANG B H, et al. Short time power load probabilistic forecasting based on constrained parallel-LSTM neural network quantile regression mode. Power System Technology, 2021, 45(4): 1356- 1364.
9	赵洋, 王瀚墨, 康丽, 等. 基于时间卷积网络的短期电力负荷预测. 电工技术学报, 2022, 37(5): 1242- 1251.
	ZHAO Y, WANG H M, KANG L, et al. Temporal convolution network-based short-term electrical load forecasting. Transactions of China Electrotechnical Society, 2022, 37(5): 1242- 1251.
10	马越, 温蜜. 基于多尺度LDTW和TCN的空间负荷预测方法. 计算机工程, 2024, 50(3): 106- 113. doi: 10.19678/j.issn.1000-3428.0066944
	MA Y, WEN M. Spatial load forecasting method based on multiscale LDTW and TCN. Computer Engineering, 2024, 50(3): 106- 113. doi: 10.19678/j.issn.1000-3428.0066944
11	邵必林, 纪丹阳. 基于VMD-SE的电力负荷分量的多特征短期预测. 中国电力, 2024, 57(4): 162- 170.
	SHAO B L, JI D Y. Multi-feature short-term forecasting of power load components based on VMD-SE. Electric Power, 2024, 57(4): 162- 170.
12	李晓, 卢先领. 基于双重注意力机制和GRU网络的短期负荷预测模型. 计算机工程, 2022, 48(2): 291-296, 305. doi: 10.19678/j.issn.1000-3428.0060145
	LI X, LU X L. Method for forecasting short-term power load based on dual-stage attention mechanism and gated recurrent unit network. Computer Engineering, 2022, 48(2): 291-296, 305. doi: 10.19678/j.issn.1000-3428.0060145
13	欧阳福莲, 王俊, 周杭霞. 基于改进迁移学习和多尺度CNN-BiLSTM-Attention的短期电力负荷预测方法. 电力系统保护与控制, 2023, 51(2): 132- 140.
	OUYANG F L, WANG J, ZHOU H X. A short-term power load forecasting method based on improved transfer learning and multi-scale CNN-BiLSTM-Attention. Power System Protection and Control, 2023, 51(2): 132- 140.
14	唐贤伦, 陈洪旭, 熊德意, 等. 基于极端梯度提升和时间卷积网络的短期电力负荷预测. 高电压技术, 2022, 48(8): 3059- 3067.
	TANG X H, CHEN H X, XIONG D Y, et al. Short-term power load forecasting based on extreme gradient boosting and time convolution network. High Voltage Engineering, 2022, 48(8): 3059- 3067.
15	郑豪丰, 杨国华, 康文军, 等. 基于多负荷特征和TCN-GRU神经网络的负荷预测. 中国电力, 2022, 55(11): 142- 148.
	ZHENG H F, YANG G H, KANG W J, et al. Load forecasting based on multi-load characteristics and TCN-GRU neural network. Electric Power, 2022, 55(11): 142- 148.
16	LIANG J K, TANG W Y. Ultra-short-term spatiotemporal forecasting of renewable resources: an attention temporal convolutional network-based approach. IEEE Transactions on Smart Grid, 2022, 13(5): 3798- 3812. doi: 10.1109/TSG.2022.3175451
17	张鹏飞, 胡博, 何金松, 等. 基于时空图卷积网络的短期空间负荷预测方法. 电力系统自动化, 2023, 47(13): 78- 85.
	ZHANG P F, HU B, HE J S, et al. Short-term spatial load forecasting method based on spatio-temporal graph convolutional network. Automation of Electric Power Systems, 2023, 47(13): 78- 85.
18	吴军英, 路欣, 刘宏, 等. 基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测. 中国电力, 2024, 57(6): 131- 140.
	WU J Y, LU X, LIU H, et al. Ultra-short term multi-region power load forecasting based on Spearman-GCN-GRU model. Electric Power, 2024, 57(6): 131- 140.
19	焦润海, 褚佳杰, 李俊良, 等. 基于数据分解的多区域个性化联邦负荷预测方法. 中国电机工程学报, 2025, 45(5): 1691- 1703.
	JIAO R H, CHU J J, LI J L, et al. Multi-region personalized federal load forecasting method based on data decomposition. Proceedings of the CSEE, 2025, 45(5): 1691- 1703.
20	LIU M P, QIN H, CAO R, et al. Short-term load forecasting based on improved TCN and DenseNet. IEEE Access, 2022, 10, 115945- 115957. doi: 10.1109/ACCESS.2022.3218374
21	杨国华, 郑豪丰, 张鸿皓, 等. 基于Holt-Winters指数平滑和时间卷积网络的短期负荷预测. 电力系统自动化, 2022, 46(6): 73- 82.
	YANG G H, ZHENG H F, ZHANG H H, et al. Short-term load forecasting based on Holt-Winters exponential smoothing and temporal convolutional network. Automation of Electric Power Systems, 2022, 46(6): 73- 82.
22	WANG Y Y, CHEN J, CHEN X Q, et al. Short-term load forecasting for industrial customers based on TCN-LightGBM. IEEE Transactions on Power Systems, 2021, 36(3): 1984- 1997. doi: 10.1109/TPWRS.2020.3028133
23	SONG C R, YANG H D, MENG X B, et al. A novel deep-learning framework for short-term prediction of cooling load in public buildings. Journal of Cleaner Production, 2024, 434, 139796. doi: 10.1016/j.jclepro.2023.139796
24	朱凌建, 荀子涵, 王裕鑫, 等. 基于CNN-BiLSTM的短期电力负荷预测. 电网技术, 2021, 45(11): 4532- 4539.
	ZHU L J, XUN Z H, WANG Y X, et al. Short-term power load forecasting based on CNN-BiLSTM. Power System Technology, 2021, 45(11): 4532- 4539.
25	龚飘怡, 罗云峰, 方哲梅, 等. 基于Attention-BiLSTM-LSTM神经网络的短期电力负荷预测方法. 计算机应用, 2021, 41(S1): 81- 86.
	GONG P Y, LUO Y F, FANG Z M, et al. Short-term power load forecasting method based on Attention-BiLSTM-LSTM neural network. Journal of Computer Applications, 2021, 41(S1): 81- 86.

[1]	HUANG Jingui, LIU Peng, TANG Wensheng. MMD-YOLOv7: Vehicle Detection Method Under Dark Conditions [J]. Computer Engineering, 2025, 51(9): 340-349.
[2]	FU Jiacheng, TIAN Jin, ZHANG Yujin, FANG Zhijun. Knowledge Graph Recommendation Based on Previous Triple Set [J]. Computer Engineering, 2025, 51(9): 101-109.
[3]	ZHOU Chenyang, LIU Xueyu, LIANG Shaohua, WU Yongfei. Segmentation and Quantitative Analysis of Renal Artery Vessel Detection Based on Swin Transformer [J]. Computer Engineering, 2025, 51(9): 252-267.
[4]	ZHAI Zhipeng, CAO Yang, SHEN Qinqin, SHI Quan. Traffic Flow Prediction Based on Multiple Spatio-Temporal Graph Fusion and Dynamic Attention [J]. Computer Engineering, 2025, 51(9): 139-148.
[5]	XU Ying, FU Ziwei, ZHANG Wei, CHEN Yunfang. Smart Contract Vulnerability Detection Technology Based on Abstract Syntax Tree Embedding [J]. Computer Engineering, 2025, 51(9): 149-157.
[6]	MA Gan, GU Yu, PENG Dongliang. Combining Improved YOLOv5s and Dynamic Data Augmentation for Sea Surface Ship Detection [J]. Computer Engineering, 2025, 51(9): 294-305.
[7]	CHEN Yanru, LIU Keliang, RAN Maoliang. Real-time Optimization of Instant Meal Delivery Based on Deep Reinforcement Learning [J]. Computer Engineering, 2025, 51(9): 328-339.
[8]	NI Yuansong, HAN Jun, ZOU Xiaoyan, HU Guangyi, WANG Wenshuai. Two-Stage Adaptive Block Transmission Line Bolt Defect Detection Method [J]. Computer Engineering, 2025, 51(8): 281-291.
[9]	HAO Hongda, LUO Jianxu. Multi-Organ Semantic Segmentation Model Based on Multi-Scale Region Feature Fusion [J]. Computer Engineering, 2025, 51(8): 270-280.
[10]	WU Donghui, WANG Jinfeng, QIU Sen, LIU Guozhi. Sign Language Recognition Using Data Gloves Based on EWBiLSTM-ATT [J]. Computer Engineering, 2025, 51(8): 107-119.
[11]	WU Donghui, WANG Jinfeng, QIU Sen, LIU Guozhi. Sign Language Recognition Using Data Gloves Based on EWBiLSTM-ATT [J]. Computer Engineering, 2025, 51(8): 107-119.
[12]	LIN Fan, LI Jianhua. Optical Chemical Structure Recognition Based on Multi-order Gated Aggregation Network [J]. Computer Engineering, 2025, 51(8): 364-372.
[13]	ZHANG Zhaoli, LI Jiahao, LIU Hai, SHI Fobo, HE Jiawen. Personalized Forgetting Modeling for Knowledge Tracing via Transformers [J]. Computer Engineering, 2025, 51(8): 120-130.
[14]	YAN Jianhong, LIU Zhiyan, WANG Zhen. Multi-Scale Convolutional Vehicle Trajectory Prediction Integrating Spatiotemporal Attention Mechanism [J]. Computer Engineering, 2025, 51(8): 406-414.
[15]	LIU Chunxia, MENG Jixing, PAN Lihu, GONG Dali. Remote Sensing Small-Target Detection Method with Fusion of RGB and IR Images [J]. Computer Engineering, 2025, 51(7): 326-338.

Please choose a citation manager

Content to export