基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测

doi:10.11930/j.issn.1004-9649.202306094

中国电力 ›› 2024, Vol. 57 ›› Issue (6): 131-140.DOI: 10.11930/j.issn.1004-9649.202306094

基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测

吴军英¹(), 路欣¹(), 刘宏¹(), 张彬², 柴守亮², 刘蕴春³(), 王佳楠⁴

1. 国网河北省电力有限公司信息通信分公司，河北石家庄　050051
2. 国网河北省电力有限公司邯郸供电分公司，河北邯郸　056035
3. 西安电子科技大学广州研究院，广东广州　510555
4. 北京中电普华信息技术有限公司，北京　102192

收稿日期:2023-06-25 接受日期:2024-01-15 出版日期:2024-06-28 发布日期:2024-06-25
作者简介:吴军英（1982—），男，通信作者，高级工程师（教授级），从事区块链、标识解析、人工智能、物联网、边缘计算等研究，E-mail：wujy@he.sgcc.com.cn
路欣（1986—），女，高级工程师，从事人工智能、物联网、系统应用等研究及数据运营管理工作，E-mail：461604215@qq.com
刘宏（1981—），女，高级工程师，从事区块链、云计算、人工智能、大数据驱动技术在电网业财一体化等领域的应用研究，E-mail：56633194@qq.com
刘蕴春（1999—），女，硕士研究生，从事基于深度学习的流量预测研究，E-mail：21181214072@stu.xidian.edu.cn
基金资助:
河北省重点研发计划资助项目（新一代电子信息技术创新专项：能源工业云网智慧物联关键技术与装备研发及应用示范，22310302D）。

Ultra-short-term Multi-region Power Load Forecasting Based on Spearman-GCN-GRU Model

Junying WU¹(), Xin LU¹(), Hong LIU¹(), Bin ZHANG², Shouliang CHAI², Yunchun LIU³(), Jianan WANG⁴

1. ICT Branch, State Grid Hebei Electric Power Co., Ltd., Shijiazhuang 050051, China
2. Handan Branch, State Grid Hebei Electric Power Co., Ltd., Handan 056035, China
3. Guangzhou Institute of Technology, Xidian University, Guangzhou 510555, China
4. Beijing China Power Puhua Information Technology Co., Ltd., Beijing 102192, China

Received:2023-06-25 Accepted:2024-01-15 Online:2024-06-28 Published:2024-06-25
Supported by:
This work is supported by Key Research & Development Program of Hebei Province (New Generation of Electronic Information Technology Innovation Project: Energy Industry Cloud Network Intelligent IoT Key Technology and Equipment Research and Development and Application Demonstration, No.22310302D).

摘要/Abstract

摘要：

为提升多区域电力负荷的预测精度，聚焦于多区域电力数据的时空相关性分析，提出了一种基于Spearman-GCN-GRU的超短期多区域电力负荷预测模型。该模型通过Spearman相关系数分析不同区域电力负荷的时空相关性，构建Spearman邻接矩阵并输入图卷积神经网络（graph convolutional network，GCN）和门控循环单元（gated recurrent unit，GRU）提取数据中的空间特征和时序特征，最后由多层感知机（multilayer perceptron，MLP）解码输出预测结果。与基于距离邻接矩阵的模型进行对比，验证了Spearman-GCN-GRU模型的可行性。在模型的预测精度上，与传统统计模型和神经网络模型相比，Spearman-GCN-GRU模型在通用的评价指标中均取得最优结果。就均方根误差（root mean square error，RMSE）而言，Spearman-GCN-GRU模型与神经网络模型GRU、GCN和深度神经网络（deep neural network，DNN）相比，RMSE指标分别下降了13.90%、11.66%和8.36%，验证了模型具有更好的预测效果。

关键词: 多区域电力负荷预测, 电力数据时空相关性分析, Spearman相关系数, 图卷积神经网络, 门控循环单元

Abstract:

To improve the prediction accuracy of multi-region power load, an ultra-short-term multi-region power load forecasting model based on Spearman-GCN-GRU is proposed with focus on the spatial-temporal correlation analysis of multi-region power data. Firstly, the Spearman correlation coefficient is used to analyze the spatial-temporal correlation of power load in different regions and construct the Spearman adjacency matrix. And then, the graph convolutional network (GCN) and gated recurrent unit (GRU) are used to respectively extract the spatial and temporal features from the data. Finally, the multilayer perceptron (MLP) is used to decode and output the prediction results. Through comparison with the distance adjacency matrix-based models, the Spearman-GCN-GRU model is proved to be feasible. In terms of prediction accuracy, the Spearman-GCN-GRU model are optimal in common evaluation indexes compared with traditional statistical models and neural network models. Specifically, in terms of the root mean square error (RMSE), the Spearman-GCN-GRU model exhibits a respective decrease of 13.90%, 11.66%, and 8.36% compared to the GRU, GCN and deep neural network (DNN) models, demonstrating its superior predictive performance.

Key words: multi-region power load prediction, spatial-temporal correlation analysis of power data, Spearman correlation coefficient, graph convolutional network, gated recurrent unit

吴军英, 路欣, 刘宏, 张彬, 柴守亮, 刘蕴春, 王佳楠. 基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测[J]. 中国电力, 2024, 57(6): 131-140.

Junying WU, Xin LU, Hong LIU, Bin ZHANG, Shouliang CHAI, Yunchun LIU, Jianan WANG. Ultra-short-term Multi-region Power Load Forecasting Based on Spearman-GCN-GRU Model[J]. Electric Power, 2024, 57(6): 131-140.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202306094

https://www.electricpower.com.cn/CN/Y2024/V57/I6/131

图/表 14

图 1 地理距离相关和时空相关

Fig.1 Geographic correlation and spatial-temporal correlation

图 2 距离邻接矩阵和时空邻接矩阵

Fig.2 Geographic adjacency matrix and spatial-temporal adjacency matrix

图 3 基于Spearman相关系数构建区域时空相关图

Fig.3 The regional spatial-temporal correlation graph constructed based on Spearman correlation coefficient

图 4 GCN卷积过程

Fig.4 GCN convolution process

图 5 基于Spearman-GCN-GRU的超短期多区域电力负荷预测模型

Fig.5 Spearman-GCN-GRU-based ultra-short-term multi-region power load forecasting model

图 6 所选10个区域之间的Spearman相关系数

Fig.6 Spearman correlation coefficient among 10 regions

图 7 不同阈值下对应的邻接矩阵

Fig.7 Adjacency matrices with different thresholds

表 1 阈值0.80~0.90对应模型预测精度对比

Table 1 Comparison of prediction accuracy with threshold varying from 0.80 to 0.90

阈值	区域数	RMSE/W	RRMSE/%	MAE/W	MAPE/%	PA/%
0.80	10	8.7759	13.7826	5.4075	8.5570	91.4119
0.85	10	8.2328	12.7745	4.5394	7.2324	92.9375
0.90	10	7.3495	12.5806	4.6102	7.1035	93.0633

表 2 阈值0.90和0.95对应模型预测精度对比

Table 2 Comparison of prediction accuracy between threshold 0.90 and 0.95

阈值	区域数	RMSE/W	RRMSE/%	MAE/W	MAPE/%	PA/%
0.90	6	6.2266	11.7588	4.1887	7.0101	92.8098
0.95	6	6.6723	12.8625	3.8127	6.3530	93.5112

图 8 Spearman邻接矩阵和距离邻接矩阵

Fig.8 Spearman adjacency matrices and geographic adjacency matrices

表 3 不同邻接矩阵对应模型的预测精度对比

Table 3 Comparison of prediction accuracy with different adjacency matrices

方法	区域数	RMSE/W	RRMSE/%	MAE/W	MAPE/%	PA/%
Geographic-GCN-GRU	9	7.4849	12.6664	4.8888	7.9478	92.7178
Spearman-GCN-GRU	9	6.6723	11.8877	4.4881	7.0999	93.0428

图 9 区域1~4在测试集中的预测效果

Fig.9 Prediction results of regions from 1 to 4 in test dataset

表 4 不同模型的预测结果

Table 4 Prediction results of different models

预测模型	RMSE/W	RRMSE/%	MAE/W	MAPE/%	PA/%
SVR （kernel='rbf'）	9.4144	13.6433	6.5627	9.9044	90.1812
GRU	8.5369	13.3252	4.6561	8.3685	92.5577
GCN	8.3195	12.5988	5.0038	7.7763	92.1970
DNN	8.0200	14.0683	5.1853	8.0953	91.0630
Spearman- GCN-GRU	7.3495	11.7929	4.6102	7.1035	93.0633

图 10 不同模型对应多区域下的预测结果

Fig.10 Prediction results of different models in multiple regions

参考文献 25

1	杨秀媛, 肖洋, 陈树勇. 风电场风速和发电功率预测研究[J]. 中国电机工程学报, 2005, 25 (11): 1- 5.
	YANG Xiuyuan, XIAO Yang, CHEN Shuyong. Wind speed and generated power forecasting in wind farm[J]. Proceedings of the CSEE, 2005, 25 (11): 1- 5.
2	杨伟, 朱巧明, 李培峰, 等. 基于时间序列的服务器负载预测[J]. 计算机工程, 2006, 32 (19): 143- 145, 148.
	YANG Wei, ZHU Qiaoming, LI Peifeng, et al. Server load prediction based on time series[J]. Computer Engineering, 2006, 32 (19): 143- 145, 148.
3	解振学, 林帆, 王若谷, 等. 基于时序动态回归的超短期光伏发电功率预测方法[J]. 智慧电力, 2022, 50 (7): 45- 51.
	XIE Zhenxue, LIN Fan, WANG Ruogu, et al. Very short-term photovoltaic power forecasting method based on time series dynamic regression[J]. Smart Power, 2022, 50 (7): 45- 51.
4	王勇, 黄国兴, 彭道刚. 带反馈的多元线性回归法在电力负荷预测中的应用[J]. 计算机应用与软件, 2008, 25 (1): 82- 84.
	WANG Yong, HUANG Guoxing, PENG Daogang. Application of multiple linear-feedback regression analysis to electric load forecasting[J]. Computer Applications and Software, 2008, 25 (1): 82- 84.
5	杨海柱, 田馥铭, 张鹏, 等. 基于CEEMD-FE和AOA-LSSVM的短期电力负荷预测[J]. 电力系统保护与控制研究, 2022, 50 (13): 126- 133. DOI
	YANG Haizhu, TIAN Fuming, ZHANG Peng, et al. Short-term load forecasting based on CEEMD-FE-AOA-LSSVM[J]. Power System Protection and Control, 2022, 50 (13): 126- 133. DOI
6	蓝信军, 杨期余, 江辉. 加权模糊回归方法在电力负荷预测中的应用[J]. 长沙电力学院学报(自然科学版), 2003, 18 (4): 34- 37.
	LAN Xinjun, YANG Qiyu, JIANG Hui. The applications of weighted fuzzy regression method to power load forecasting[J]. Journal of Changsha University of Electric Power (Natural Science), 2003, 18 (4): 34- 37.
7	陈晓东, 姚晓林, 晋飞. 电力系统超短期负荷预测方法分析与研究[J]. 城市建设理论研究(电子版), 2014, (34): 602- 603.
8	程志友, 汪德胜. 基于机器学习与疫情关联特征的短期负荷预测[J]. 电力系统保护与控制, 2022, 50 (23): 1- 8. DOI
	CHENG Zhiyou, WANG Desheng. Short-term load forecasting based on machine learning and epidemic association features[J]. Power System Protection and Control, 2022, 50 (23): 1- 8. DOI
9	KONG W C, DONG Z Y, JIA Y W, et al. Short-term residential load forecasting based on LSTM recurrent neural network[J]. IEEE Transactions on Smart Grid, 2019, 10 (1): 841- 851. DOI
10	高明, 郝妍. 基于BiLSTM网络与误差修正的超短期负荷预测[J]. 综合智慧能源, 2023, 45 (1): 31- 40.
	GAO Ming, HAO Yan. Ultra-short-term load forecasting based on BiLSTM network and error correction[J]. Integrated Intelligent Energy, 2023, 45 (1): 31- 40.
11	李闯, 孔祥玉, 朱石剑, 等. 能源互联环境下考虑需求响应的区域电网短期负荷预测[J]. 电力系统自动化, 2021, 45 (1): 71- 78.
	LI Chuang, KONG Xiangyu, ZHU Shijian, et al. Short-term load forecasting of regional power grid considering demand response in energy interconnection environment[J]. Automation of Electric Power Systems, 2021, 45 (1): 71- 78.
12	KUMAR S, HUSSAIN L, BANARJEE S, et al. Energy load forecasting using deep learning approach-LSTM and GRU in spark cluster[C]//2018 Fifth International Conference on Emerging Applications of Information Technology (EAIT). Kolkata, India. IEEE, 2018: 1–4.
13	陈振宇, 刘金波, 李晨, 等. 基于LSTM与XGBoost组合模型的超短期电力负荷预测[J]. 电网技术, 2020, 44 (2): 614- 620.
	CHEN Zhenyu, LIU Jinbo, LI Chen, et al. Ultra short-term power load forecasting based on combined LSTM-XGBoost model[J]. Power System Technology, 2020, 44 (2): 614- 620.
14	SAJJAD M, AHMAD KHAN Z, ULLAH A, et al. A novel CNN-GRU-based hybrid approach for short-term residential load forecasting[J]. IEEE Access, 2020, 8, 143759- 143768. DOI
15	曾囿钧, 肖先勇, 徐方维, 等. 基于CNN-BiGRU-NN模型的短期负荷预测方法[J]. 中国电力, 2021, 54 (9): 17- 23.
	ZENG Youjun, XIAO Xianyong, XU Fangwei, et al. A short-term load forecasting method based on CNN-BiGRU-NN model[J]. Electric Power, 2021, 54 (9): 17- 23.
16	杨胡萍, 余阳, 汪超, 等. 基于VMD-CNN-BIGRU的电力系统短期负荷预测[J]. 中国电力, 2022, 55 (10): 71- 76.
	YANG Huping, YU Yang, WANG Chao, et al. Short-term load forecasting of power system based on VMD-CNN-BIGRU[J]. Electric Power, 2022, 55 (10): 71- 76.
17	刘亚珲, 赵倩. 基于聚类经验模态分解的CNN-LSTM超短期电力负荷预测[J]. 电网技术, 2021, 45 (11): 4444- 4451.
	LIU Yahui, ZHAO Qian. Ultra-short-term power load forecasting based on cluster empirical mode decomposition of CNN-LSTM[J]. Power System Technology, 2021, 45 (11): 4444- 4451.
18	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[J]. 电力系统自动化, 2023, 47 (13): 78- 85.
	ZHANG Pengfei, HU Bo, HE Jinsong, et al. Short-term spatial load forecasting method based on spatio-temporal graph convolutional network[J]. Automation of Electric Power Systems, 2023, 47 (13): 78- 85.
19	董雷, 陈振平, 韩富佳, 等. 基于图卷积神经网络与K-means聚类的居民用户集群短期负荷预测[J]. 电网技术, 2023, 47 (10): 4291- 4301.
	DONG Lei, CHEN Zhenping, HAN Fujia, et al. Short-term load forecasting of residential user groups based on graph convolutional neural network and K-means clustering[J]. Power System Technology, 2023, 47 (10): 4291- 4301.
20	SPEARMAN C. The proof and measurement of association between two things[J]. International Journal of Epidemiology, 2010, 39 (5): 1137- 1150. DOI
21	BARLACCHI G, DE NADAI M, LARCHER R, et al. A multi-source dataset of urban life in the city of Milan and the Province of Trentino[J]. Scientific Data, 2015, 2, 150055. DOI
22	KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[J]. CoRR, 2016. DOI:10.48550/arXiv.1609.02907.
23	CHUNG J, GULCEHRE C, CHO K, et al. Empirical evaluation of gated recurrent neural networks on sequence modeling[J]. ArXiv e-Prints, 2014: arXiv: 1412.3555.
24	任建吉, 位慧慧, 邹卓霖, 等. 基于CNN-BiLSTM-Attention的超短期电力负荷预测[J]. 电力系统保护与控制, 2022, 50 (8): 108- 116.
	REN Jianji, WEI Huihui, ZOU Zhuolin, et al. Ultra-short-term power load forecasting based on CNN-BiLSTM-Attention[J]. Power System Protection and Control, 2022, 50 (8): 108- 116.
25	陆继翔, 张琪培, 杨志宏, 等. 基于CNN-LSTM混合神经网络模型的短期负荷预测方法[J]. 电力系统自动化, 2019, 43 (8): 131- 137.
	LU Jixiang, ZHANG Qipei, YANG Zhihong, et al. Short-term load forecasting method based on CNN-LSTM hybrid neural network model[J]. Automation of Electric Power Systems, 2019, 43 (8): 131- 137.

[1]	朱沐雨, 马宏忠, 郭鹏宇, 宣文婧. 典型调峰/调频工况下储能电池组荷电状态估计[J]. 中国电力, 2024, 57(6): 18-26.
[2]	杨国华, 祁鑫, 贾睿, 刘一峰, 蒙飞, 马鑫, 邢潇文. 基于CEEMD-SE的CNN&LSTM-GRU短期风电功率预测[J]. 中国电力, 2024, 57(2): 55-61.
[3]	陈子含, 滕伟, 胥学峰, 丁显, 柳亦兵. 基于图卷积网络和风速差分拟合的中长期风功率预测[J]. 中国电力, 2023, 56(10): 96-105.
[4]	冯裕祺, 李辉, 李利娟, 周彦博, 谭貌, 彭寒梅. 基于CNN-GRU的光伏电站电压轨迹预测[J]. 中国电力, 2022, 55(7): 163-171.
[5]	郑豪丰, 杨国华, 康文军, 刘志远, 刘世涛, 伍弘, 张鸿皓. 基于多负荷特征和TCN-GRU神经网络的负荷预测[J]. 中国电力, 2022, 55(11): 142-148.
[6]	杨胡萍, 余阳, 汪超, 李向军, 胡奕涛, 饶楚楚. 基于VMD-CNN-BIGRU的电力系统短期负荷预测[J]. 中国电力, 2022, 55(10): 71-76.
[7]	曾囿钧, 肖先勇, 徐方维, 郑林. 基于CNN-BiGRU-NN模型的短期负荷预测方法[J]. 中国电力, 2021, 54(9): 17-23.

基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测

Ultra-short-term Multi-region Power Load Forecasting Based on Spearman-GCN-GRU Model

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 25

相关文章 7

编辑推荐

Metrics

模态框（Modal）标题

基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测

Ultra-short-term Multi-region Power Load Forecasting Based on Spearman-GCN-GRU Model

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 25

相关文章 7

编辑推荐

Metrics