基于堆叠自动编码器的电网运行断面相似性匹配研究

doi:10.12204/j.issn.1000-7229.2021.01.013

电力建设 ›› 2021, Vol. 42 ›› Issue (1): 117-124.doi: 10.12204/j.issn.1000-7229.2021.01.013

基于堆叠自动编码器的电网运行断面相似性匹配研究

王铁强¹, 鲁鹏¹, 曹欣¹, 杨晓东¹, 王维¹, 吕昊¹, 冯春贤¹, 田潮², 石皓岩³, 梁海平³

1.国网河北省电力有限公司,石家庄市 050023
2.国网北京市电力公司亦庄供电公司,北京市 100176
3.华北电力大学电力工程系,河北省保定市 071003

收稿日期:2020-06-01 出版日期:2021-01-01 发布日期:2021-01-07
通讯作者: 石皓岩
作者简介:王铁强(1970),男,博士,正高级工程师,主要研究方向为电力系统智能调度;|鲁鹏(1981),男,硕士,高级工程师,主要研究方向为智能技术在电力系统中的应用;|曹欣(1980),男,硕士,高级工程师,主要研究方向为电力系统调度自动化;|杨晓东(1981),男,博士,高级工程师,主要研究方向为电力系统智能调度;|王维(1982),男,硕士,高级工程师,主要研究方向为智能技术在电力系统中的应用;|吕昊(1985),男,学士,高级工程师,主要研究方向为电力系统智能调度;|冯春贤(1984),男,硕士,高级工程师,主要研究方向为智能技术在电力系统中的应用;|田潮(1993),男,硕士,助理工程师,主要研究方向为智能技术在电力系统中的应用;|梁海平(1979),男,博士,讲师,主要研究方向为智能技术在电网中的应用、现代电网评估、电力系统安全防御与恢复控制。
基金资助:
国家电网公司科技项目“基于多源异构时空信息的电网运行方式智能互动决策关键技术研究”(SGTYHT/17-JS-199)

Research on Similarity Matching of Power Network Operation Section Applying Stacked Automatic Encoder

WANG Tieqiang¹, LU Peng¹, CAO Xin¹, YANG Xiaodong¹, WANG Wei¹, Lü Hao¹, FENG Chunxian¹, TIAN Chao², SHI Haoyan³, LIANG Haiping³

1. State Grid Hebei Electric Power Co., Ltd., Shijiazhuang 050023, China
2. State Grid Beijing Electric Power Company, Yizhuang Power Supply Company, Beijing 100176, China
3. Department of Electrical Engineering,North China Electric Power University, Baoding 071003, Hebei Province, China

Received:2020-06-01 Online:2021-01-01 Published:2021-01-07
Contact: SHI Haoyan
Supported by:
Science and Technology Project of State Grid Corporation of China(SGTYHT/17-JS-199)

摘要/Abstract

摘要：

面向电网历史数据挖掘需求,提出了一种基于堆叠自动编码器(stacked automatic encoder, SAE)的电网运行断面相似性匹配方法。基于历史断面信息,结合电网运行特点,筛选有效样本数据。通过无标签的电网运行断面有效样本数据和逐层自动编码器(automatic encoder, AE)得到预训练阶段的权重和偏差参数。进一步地,在参数微调阶段利用带标签的样本数据、初始化后的权重以及偏差对整个网络进行参数优化,得到能够挖掘运行断面深层特征的堆叠自编码网络。所提方法通过SAE算法的深层架构建立历史运行断面数据与相似性度量的非线性映射关系,进而可得到有价值的历史信息。采用IEEE 39节点算例对所提方法进行验证。结果表明,所提方法相较K-means算法匹配准确率较高,错误率随迭代次数的增加下降速度较快。

关键词: 历史信息, 运行断面, 相似性匹配, 堆叠自动编码器(SAE), 特征提取

Abstract:

Aiming at the demand of power grid historical data mining, a method for similarity matching of power grid operation section applying stacked automatic encoder (SAE) is proposed. According to the historical section information, combined with the characteristics of power grid operation, effective sample data is selected. The weights and bias parameters of the pre-training stage are obtained by using unlabeled valid sample data of power grid operation section and layer by layer automatic encoder (AE). Furthermore, in the parameter tuning stage, parameters of the whole network are fine-tuned by using labeled sample data, initialized weights and deviations to obtain a stacked self-coding network capable of mining the deep features of the running section. The proposed method establishes a nonlinear mapping relationship between historical operation section data and similarity measurement through the deep structure of SAE algorithm, and then obtains valuable historical information. IEEE 39-node system is used to verify the proposed method. The results show that the proposed method has higher matching accuracy than K-means algorithm, and the error rate decreases faster with the number of iterations.

Key words: history information, operation section, similarity matching, Stacked Automatic Encoder(SAE), feature extraction

中图分类号:

TM761

王铁强, 鲁鹏, 曹欣, 杨晓东, 王维, 吕昊, 冯春贤, 田潮, 石皓岩, 梁海平. 基于堆叠自动编码器的电网运行断面相似性匹配研究[J]. 电力建设, 2021, 42(1): 117-124.

WANG Tieqiang, LU Peng, CAO Xin, YANG Xiaodong, WANG Wei, Lü Hao, FENG Chunxian, TIAN Chao, SHI Haoyan, LIANG Haiping. Research on Similarity Matching of Power Network Operation Section Applying Stacked Automatic Encoder[J]. ELECTRIC POWER CONSTRUCTION, 2021, 42(1): 117-124.

图/表 9

图1 自动编码器结构

Fig.1 Structural diagram of automatic encoder

图2 堆叠自动编码器结构

Fig.2 Structural diagram of stacked automatic encoder

表1 原始静态特征量库

Table 1 Original static feature library

图3 3机9节点系统接线图

Fig.3 Connection diagram of 9-node test system

表2 3机9节点样本示例

Table 2 Sample of 9 nodes and 3 generators

图4 IEEE 39节点系统接线图

Fig.4 Connection diagram of IEEE 39-node System

图5 算例中堆叠自动编码器的网络结构

Fig.5 Network structure of stacked self-encoder in the example

表3 不同算法的匹配准确率

Table 3 Matching accuracy of different algorithms

图6 匹配准确率随迭代次数的变化情况

Fig.6 Changing of matching accuracy with iteration number

参考文献 23

[1]	中国电机工程学会电力信息化专业委员会. 中国电力大数据发展白皮书-2013[M]. 北京: 中国电力出版社, 2013: 16.
	Professional Committee of electric power informatization of China Electrical Engineering Society. White paper of China electric power big data development [M]. Beijing: China Electric Power Press, 2013: 16.
[2]	黄天恩, 孙宏斌, 郭庆来 , 等. 基于电网运行仿真大数据的知识管理和超前安全预警[J]. 电网技术, 2015,39(11):3080-3087. doi: 10.13335/j.1000-3673.pst.2015.11.010 URL
	HUANG Tian’en, SUN Hongbin, GUO Qinglai , et al. Knowledge management and security early warning based on big simulation data in power grid operation[J]. Power System Technology, 2015,39(11):3080-3087. doi: 10.13335/j.1000-3673.pst.2015.11.010 URL
[3]	谢先章, 王兆凯, 李亚星 , 等. 基于卷积神经网络的跨领域语义信息检索研究[J]. 计算机应用与软件, 2018,35(8):73-78.
	XIE Xianzhang, WANG Zhaokai, LI Yaxing , et al. Cross-domain semantic information retrieval based on convolutional neural network[J]. Computer Applications and Software, 2018,35(8):73-78.
[4]	唐贤伦, 杜一铭, 刘雨微 , 等. 基于条件深度卷积生成对抗网络的图像识别方法[J]. 自动化学报, 2018,44(5):855-864.
	TANG Xianlun, DU Yiming, LIU Yuwei , et al. Image recognition with conditional deep convolutional generative adversarial networks[J]. Acta Automatica Sinica, 2018,44(5):855-864.
[5]	岳有军, 刘英翰, 赵辉 , 等. 基于CEEMDAN-SE和DBN的短期电力负荷预测[J]. 电测与仪表, 2020,57(17):59-65.
	YUE Youjun, LIU Yinghan, ZHAO Hui , et al. Short-term load forecasting based on CEEMDAN-SE and DBN[J]. Electrical Measurement & Instrumentation, 2020,57(17):59-65.
[6]	马天男, 王超, 彭丽霖 , 等. 计及需求响应和深度结构多任务学习的电力系统短期负荷预测[J]. 电测与仪表, 2019,56(16):50-60.
	MA Tiannan, WANG Chao, PENG Lilin , et al. Short-term load forecasting of power system considering demand response and multi-task learning based on deep structure[J]. Electrical Measurement & Instrumentation, 2019,56(16):50-60.
[7]	于群, 李浩, 屈玉清 . 基于深度神经网络和内外部因素的大电网安全态势感知研究[J/OL]. 电测与仪表:1-10[2020-04-30]. https://kns.cnki.net/kcms/detail/23.1202.TH.20200417.1650.002.html.
	YU Qun, LI Hao, QU Yuqing . Research on security situation awareness of large power grid based on deep neural network and internal and external factors [J/OL]. Electrical Measurement & Instrumentation: 1-10[ 2020- 04- 30]. Http://kns.cnki.net/kcms/detail/23.1202.th.20200417.1650.002.html.
[8]	张秀钊, 王志敏, 钱纹 , 等. 基于时空神经网络的充电桩时空动态负荷预测[J]. 智慧电力, 2019,47(8):50-54.
	ZHANG Xiuzhao, WANG Zhimin, QIAN Wen , et al. Spatio-temporal dynamic load forecasting of charging pile based on spatio-temporal neural network[J]. Smart Power, 2019,47(8):50-54.
[9]	朱乔木, 党杰, 陈金富 , 等. 基于深度置信网络的电力系统暂态稳定评估方法[J]. 中国电机工程学报, 2018,38(3):735-743.
	ZHU Qiaomu, DANG Jie, CHEN Jinfu , et al. A method for power system transient stability assessment based on deep belief networks[J]. Proceedings of the CSEE, 2018,38(3):735-743.
[10]	黄彦浩, 于之虹, 史东宇 , 等. 基于海量在线历史数据的大电网快速判稳策略[J]. 中国电机工程学报, 2016,36(3):596-603. doi: 10.13334/j.0258-8013.pcsee.2016.03.002 URL
	HUANG Yanhao, YU Zhihong, SHI Dongyu , et al. Strategy of huge electric power system stability quick judgment based on massive historical online data[J]. Proceedings of the CSEE, 2016,36(3):596-603.
[11]	张宇航, 邱才明, 杨帆 , 等. 深度学习在电网图像数据及时空数据中的应用综述[J]. 电网技术, 2019,43(6):1865-1873.
	ZHANG Yuhang, QIU Robert, YANG Fan , et al. Overview of application of deep learning with image data and spatio-temporal data of power grid[J]. Power System Technology, 2019,43(6):1865-1873.
[12]	高昆仑, 杨帅, 刘思言 , 等. 基于一维卷积神经网络的电力系统暂态稳定评估[J]. 电力系统自动化, 2019,43(12):18-26.
	GAO Kunlun, YANG Shuai, LIU Siyan , et al. Transient stability assessment for power system based on one-dimensional convolutional neural network[J]. Automation of Electric Power Systems, 2019,43(12):18-26.
[13]	仉怡超, 闻达, 王晓茹 , 等. 基于深度置信网络的电力系统扰动后频率曲线预测[J]. 中国电机工程学报, 2019,39(17):5095-5104.
	ZHANG Yichao, WEN Da, WANG Xiaoru , et al. A method of frequency curve prediction based on deep belief network of post-disturbance power system[J]. Proceedings of the CSEE, 2019,39(17):5095-5104.
[14]	王怀远, 陈启凡 . 基于堆叠变分自动编码器的电力系统暂态稳定评估方法[J]. 电力自动化设备, 2019,39(12):134-139.
	WANG Huaiyuan, CHEN Qifan . Transient stability assessment method of electric power systems based on stacked variational auto-encoder[J]. Electric Power Automation Equipment, 2019,39(12):134-139.
[15]	梁海平, 田潮, 王铁强 , 等. 基于改进K-means算法的电网运行断面相似性匹配研究[J]. 电力自动化设备, 2019,39(7):119-124.
	LIANG Haiping, TIAN Chao, WANG Tieqiang , et al. Running section similarity matching based on improved K-means algorithm[J]. Electric Power Automation Equipment, 2019,39(7):119-124.
[16]	KAN M N, SHAN S G, CHANG H , et al. Stacked progressive auto-encoders (SPAE) for face recognition across poses [C]//CVPR ‘14: Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2014: 1883-1890.
[17]	HA M, BYUN Y, KIM J , et al. Selective deep convolutional neural network for low cost distorted image classification[J]. IEEE Access, 2019,7:133030-133042. doi: 10.1109/Access.6287639 URL
[18]	CHEN Y S, LIN Z H, ZHAO X , et al. Deep learning-based classification of hyperspectral data[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014,7(6):2094-2107. doi: 10.1109/JSTARS.2014.2329330 URL
[19]	BENGIO Y . Learning deep architectures for AI[M]. Now Foundations and Treads, 2009.
[20]	邓俊锋 . 基于稀疏自动编码器和边缘降噪自动编码器的深度学习算法研究[D]. 武汉:武汉科技大学, 2016.
	DENG Junfeng . Deep learning algorithm based on sparse auto-encoder and marginalized denoising automatic encoder [D] Wuhan: Wuhan University of Science and Technology, 2016.
[21]	徐行, 李华强, 赵祥云 , 等. 基于运行状态和网络结构的节点综合脆弱性评估[J]. 电网技术, 2014,38(3):731-735. doi: 10.13335/j.1000-3673.pst.2014.03.028 URL
	XU Hang, LI Huaqiang, ZHAO Xiangyun , et al. Assessment on nodal comprehensive vulnerability based on operational state and network structure[J]. Power System Technology, 2014,38(3):731-735. doi: 10.13335/j.1000-3673.pst.2014.03.028 URL
[22]	李宏仲, 吕风磊, 范宏 . 基于分形维数的500 kV/220 kV电网结构量化方法[J]. 电力系统自动化, 2015,39(10):87-92.
	LI Hongzhong, LÜ Fenglei, FAN Hong . Quantitative structure method for 500 kV /220 kV power grid based on fractal dimension[J]. Automation of Electric Power Systems, 2015,39(10):87-92.
[23]	戴远航, 陈磊, 张玮灵 , 等. 基于多支持向量机综合的电力系统暂态稳定评估[J]. 中国电机工程学报, 2016,36(5):1173-1180. doi: 10.13334/j.0258-8013.pcsee.2016.05.001 URL
	DAI Yuanhang, CHEN Lei, ZHANG Weiling , et al. Power system transient stability assessment based on multi-support vector machines[J]. Proceedings of the CSEE, 2016,36(5):1173-1180.

基于堆叠自动编码器的电网运行断面相似性匹配研究

Research on Similarity Matching of Power Network Operation Section Applying Stacked Automatic Encoder

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 23

相关文章 5

编辑推荐

Metrics

本文评价

[1]	黄冬梅, 林孝镶, 胡安铎, 孙锦中. 基于卷积自编码器的日负荷深度嵌入聚类方法[J]. 电力建设, 2021, 42(1): 132-138.
[2]	李浩茹, 丁保迪, 季宇, 王永刚, 陈继开, 张利伟. 考虑光伏出力间歇性的微电网故障辨识方法[J]. 电力建设, 2020, 41(9): 76-85.
[3]	庄家懿, 杨国华, 郑豪丰, 王煜东, 胡瑞琨, 丁旭. 并行多模型融合的混合神经网络超短期负荷预测[J]. 电力建设, 2020, 41(10): 1-8.
[4]	陈倩，齐林海，王红. 基于LSTM网络的谐波多标签分类[J]. 电力建设, 2019, 40(7): 101-106.
[5]	闾海庆，翁向阳，封加会，雷远华,刘佳明. 一种自适应的无人机影像角点特征提取方法[J]. 电力建设, 2013, 34(7): 104-107.