基于改进生成对抗网络的虚假数据注入攻击检测方法

doi:10.12204/j.issn.1000-7229.2022.03.007

电力建设 ›› 2022, Vol. 43 ›› Issue (3): 58-65.doi: 10.12204/j.issn.1000-7229.2022.03.007

• 人工智能在电网故障诊断和定位中的应用·栏目主持王小君教授、罗国敏副教授、石访副教授· • 上一篇下一篇

基于改进生成对抗网络的虚假数据注入攻击检测方法

夏云舒¹(), 王勇¹(), 周林¹(), 樊汝森²()

1.上海电力大学计算机科学与技术学院, 上海市 200120
2.国网上海电力公司青浦供电公司,上海市 201799

收稿日期:2021-09-12 出版日期:2022-03-01 发布日期:2022-03-24
通讯作者: 夏云舒 E-mail:carriexia@163.com;wy616@126.com;zhou@shiep.edu.cn;fanrusen107@163.com
作者简介:王勇(1974),男,博士,教授,从事网络信息安全方面的研究工作,E-mail: wy616@126.com;
周林(1968),男,硕士,副教授,研究方向为计算机网络与应用,E-mail: lin. zhou@shiep.edu.cn;
樊汝森(1989),男,硕士,研究方向为输变电工程建设管理,Email: fanrusen107@163.com。
基金资助:
国家自然科学基金资助项目(61772327);上海市自然科学基金资助项目(20ZR1455900);大数据协同安全国家工程实验室项目(QAX-201803);上海市科委科技创新行动计划(18511105700);上海市科委电力人工智能工程技术研究中心项目(19DZ2252800)

False Data Injection Attack Detection Method Based onImproved Generative Adversarial Network

XIA Yunshu¹(), WANG Yong¹(), ZHOU Lin¹(), FAN Rusen²()

1. School of Computer Science and Technology, Shanghai University of Electric Power, Shanghai 200120, China
2. State Grid Qingpu Electric Power Supply Company, Shanghai 201799, China

Received:2021-09-12 Online:2022-03-01 Published:2022-03-24
Contact: XIA Yunshu E-mail:carriexia@163.com;wy616@126.com;zhou@shiep.edu.cn;fanrusen107@163.com
Supported by:
the National Natural Science Foundation of China(61772327);Natural Science Foundation of Shanghai(20ZR1455900);National Engineering Laboratory for Big Data Collaborative Security Technology(QAX-201803);Science and Technology Commission of Shanghai Municipality(18511105700);Science and Technology Commission of Shanghai Municipality(19DZ2252800)

摘要/Abstract

摘要：

随着新型能源互联网的发展,大规模的传感量测系统为基于数据驱动的虚假数据注入攻击检测方法提供了数据支持,然而攻击样本数据不平衡问题会影响此类方法的性能。提出了基于改进生成对抗网络(generative adversarial network, GAN)和极端随机树的数据重平衡攻击检测模型。首先,为了生成高质量数据,设计GAN的结构使其训练稳定;其次,使用Copula函数构建电力系统状态量之间的空间关联性以适应分布式能源的接入;然后,对改进的GAN进行对抗训练得到重平衡的数据集,采用极端随机树分类器实现攻击检测。此外,设计基于多种分类器的数据有效性指标评估生成数据的质量。通过对比实验对所提方法进行验证,结果表明该方法能生成高质量的量测数据,可以有效解决数据不平衡问题,攻击检测率达98.95%。

关键词: 虚假数据注入攻击, 生成对抗网络, 极端随机树, 不平衡数据, 机器学习, 攻击检测

Abstract:

With the development of new-type energy internet, large-scale sensing measurement systems provide data support for data-driven detection of false data injection attack. However, the problem of unbalanced attack data will affect the performance of such methods. Therefore, a data rebalance attack detection model based on improved generative adversarial network (GAN) and extremely randomized tree is proposed. Firstly, the GAN structure is designed to make the training procedure stable enough to generate high-quality data. Secondly, the Copula function is used to construct the spatial correlation between the power system states to adapt to the integration of the distributed energy resources. Then, a rebalanced dataset is obtained through the adversarial training of the improved GAN, and the extremely randomized tree classifier is used to detect the attack. In addition, the data validity index based on multiple classifiers is designed to evaluate the quality of the generated data. The effect of the proposed method is verified by comparative experiments. Results show that the method can generate high-quality measurement data, solve the problem of data imbalance, and the attack detection rate is 98.95%.

Key words: false data injection attack, generative adversarial network, extremely randomized tree, imbalanced dataset, machine learning, attack detection

中图分类号:

TM76

夏云舒, 王勇, 周林, 樊汝森. 基于改进生成对抗网络的虚假数据注入攻击检测方法[J]. 电力建设, 2022, 43(3): 58-65.

XIA Yunshu, WANG Yong, ZHOU Lin, FAN Rusen. False Data Injection Attack Detection Method Based onImproved Generative Adversarial Network[J]. ELECTRIC POWER CONSTRUCTION, 2022, 43(3): 58-65.

图/表 9

图1 基于CopulaGAN的FDIA检测框架

Fig.1 Structure of FDIA detection based onCopulaGAN method

表1 二分类混淆矩阵

Table 1 Confusion matrix for binary classification

表2 训练集中各类场景的具体描述

Table 2 Description of the training set

表3 模型超参数优化结果

Table 3 Hyperparameters of the model

表4 生成数据的相似性分数

Table 4 Similarity score of synthetic data

图2 生成数据的有效性评估

Fig.2 Effectiveness evaluation of synthetic data

表5 各种算法在原始数据集上的准确率

Table 5 Accuracy of algorithms on original dataset

图3 FDIA检测混淆矩阵

Fig.3 Confusion matrix of FDIA detection

表6 各数据过采样方法下ET算法的性能

Table 6 Performance of ET algorithm withdiffrent oversampling methods

参考文献 29

[1]	杨杉, 谭博, 郭静波. 基于双马尔科夫链的新型能源互联网虚假数据注入攻击检测[J]. 电力自动化设备, 2021, 41(2):131-137.
	YANG Shan, TAN Bo, GUO Jingbo. Detection of false data injection attack for new-type energy Internet based on double Markov chains[J]. Electric Power Automation Equipment, 2021, 41(2):131-137.
[2]	汤奕, 陈倩, 李梦雅, 等. 电力信息物理融合系统环境中的网络攻击研究综述[J]. 电力系统自动化, 2016, 40(17):59-69.
	TANG Yi, CHEN Qian, LI Mengya, et al. Overview on cyber-attacks against cyber physical power system[J]. Automation of Electric Power Systems, 2016, 40(17):59-69.
[3]	王琦, 邰伟, 汤奕, 等. 面向电力信息物理系统的虚假数据注入攻击研究综述[J]. 自动化学报, 2019, 45(1):72-83.
	WANG Qi, TAI Wei, TANG Yi, et al. A review on false data injection attack toward cyber-physical power system[J]. Acta Automatica Sinica, 2019, 45(1):72-83.
[4]	罗小元, 潘雪扬, 王新宇, 等基于自适应Kalman滤波的智能电网假数据注入攻击检测[J/OL]. 自动化学报, 2020 (2020-07-13)[2021-08-12]. https://kns.cnki.net/kcms/detail/detail/11.2109.TP.20200710.1441.001.html.
	LUO Xiaoyuan, PAN Xueyang, WANG XinYu, et al. Detection of false data injection attack in smart grid via adaptive kalman filtering[J/OL]. Acta Automatica Sinica, 2020(2020-07-13)[2021-08-12]. https://kns.cnki.net/kcms/detail/detail/11.2109.TP.20200710.1441.001.html.
[5]	赵丽莉, 刘忠喜, 孙国强, 等. 基于非线性状态估计的虚假数据注入攻击代价分析[J]. 电力系统保护与控制, 2019, 47(19):38-45.
	ZHAO Lili, LIU Zhongxi, SUN Guoqiang, et al. Cost analysis of the false data injection attack based on nonlinear state estimation[J]. Power System Protection and Control, 2019, 47(19):38-45.
[6]	ZHAO J B, ZHANG G X, LA SCALA M, et al. Short-term state forecasting-aided method for detection of smart grid general false data injection attacks[J]. IEEE Transactions on Smart Grid, 2017, 8(4):1580-1590. doi: 10.1109/TSG.2015.2492827 URL
[7]	KHARE G, MOHAPATRA A, SINGH S N. A real-time approach for detection and correction of false data in PMU measurements[J]. Electric Power Systems Research, 2021, 191:106866. doi: 10.1016/j.epsr.2020.106866 URL
[8]	李元诚, 曾婧. 基于改进卷积神经网络的电网假数据注入攻击检测方法[J]. 电力系统自动化, 2019, 43(20):97-104.
	LI Yuancheng, ZENG Jing. Detection method of false data injection attack on power grid based on improved convolutional neural network[J]. Automation of Electric Power Systems, 2019, 43(20):97-104.
[9]	陈刘东, 刘念. 面向互动需求响应的虚假数据注入攻击及其检测方法[J]. 电力系统自动化, 2021, 45(3):15-23.
	CHEN Liudong, LIU Nian. False data injection attack and its detection method for interactive demand response[J]. Automation of Electric Power Systems, 2021, 45(3):15-23.
[10]	WANG J Y, SUN Z W, BAO B, et al. Malicious synchrophasor detection based on highly imbalanced historical operational data[J]. CSEE Journal of Power and Energy Systems, 2019, 5(1):11-20.
[11]	李艳霞, 柴毅, 胡友强, 等. 不平衡数据分类方法综述[J]. 控制与决策, 2019, 34(4):673-688.
	LI Yanxia, CHAI Yi, HU Youqiang, et al. Review of imbalanced data classification methods[J]. Control and Decision, 2019, 34(4):673-688.
[12]	陈杰, 张浩天, 汤奕. 基于改进生成式对抗网络的电网异常数据辨识方法[J]. 电力建设, 2021, 42(5):9-15.
	CHEN Jie, ZHANG Haotian, TANG Yi. An abnormal data identification method based on improved generative adversarial network[J]. Electric Power Construction, 2021, 42(5):9-15.
[13]	张阳, 张涛, 陈锦, 等. 基于SMOTE和机器学习的网络入侵检测[J]. 北京理工大学学报, 2019, 39(12):1258-1262.
	ZHANG Yang, ZHANG Tao, CHEN Jin, et al. Research on network intrusion detection based on SMOTE algorithm and machine learning[J]. Transactions of Beijing Institute of Technology, 2019, 39(12):1258-1262.
[14]	石洪波, 陈雨文, 陈鑫. SMOTE过采样及其改进算法研究综述[J]. 智能系统学报, 2019, 14(6):1073-1083.
	SHI Hongbo, CHEN Yuwen, CHEN Xin. Summary of research on SMOTE oversampling and its improved algorithms[J]. CAAI Transactions on Intelligent Systems, 2019, 14(6):1073-1083.
[15]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11):139-144. doi: 10.1145/3422622 URL
[16]	TANAKA F H K D S, ARANHA C. Data augmentation using GANs[EB/OL]. [2021-08-12]. https://arxiv.org/abs/1904.09135
[17]	ZHENG X T, WANG B, XIE L. Synthetic dynamic PMU data generation: A generative adversarial network approach[C]// 2019 International Conference on Smart Grid Synchronized Measurements and Analytics (SGSMA). College Station, TX, USA: IEEE, 2019: 1-6.
[18]	AHMADIAN S, MALKI H, HAN Z. Cyber attacks on smart energy grids using generative adverserial networks[C]// 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP). Anaheim, CA, USA: IEEE, 2018: 942-946.
[19]	MOHAMMADPOURFARD M, GHANAATPISHE F, MOHAMMADI M, et al. Generation of false data injection attacks using conditional generative adversarial networks[C]// 2020 IEEE PES Innovative Smart Grid Technologies Europe. Hague, Netherlands: IEEE, 2020: 41-45.
[20]	LI Y C, WANG Y Y, HU S Y. Online generative adversary network based measurement recovery in false data injection attacks: A cyber-physical approach[J]. IEEE Transactions on Industrial Informatics, 2020, 16(3):2031-2043. doi: 10.1109/TII.2019.2921106 URL
[21]	XU L, SKOULARIDOU M, CUESTA-INFANTE A, et al. Modeling tabular data using conditional GAN[DB/OL]. 2019(2019-01-01)[2021-08-12]. https://arxiv.org/abs/1907.00503v2.
[22]	PATKI N, WEDGE R, VEERAMACHANENI K. The synthetic data vault[C]// 2016 IEEE International Conference on Data Science and Advanced Analytics. IEEE, 2016: 399-410.
[23]	赵渊, 刘庆尧, 邝俊威, 等. 电网可靠性评估中相关性变量的非参数R藤Copula模型[J]. 中国电机工程学报, 2020, 40(3):803-812.
	ZHAO Yuan, LIU Qingyao, KUANG Junwei, et al. A nonparametric regular vine copula model for multidimensional dependent variables in power system reliability assessment[J]. Proceedings of the CSEE, 2020, 40(3):803-812.
[24]	王奕森, 夏树涛. 集成学习之随机森林算法综述[J]. 信息通信技术, 2018, 12(1):49-55.
	WANG Yisen, XIA Shutao. A survey of random forests algorithms[J]. Information and Communications Technologies, 2018, 12(1):49-55.
[25]	CAMANA ACOSTA M R, AHMED S, GARCIA C E, et al. Extremely randomized trees-based scheme for stealthy cyber-attack detection in smart grid networks[J]. IEEE Access, 2020, 8:19921-19933. doi: 10.1109/ACCESS.2020.2968934 URL
[26]	GULRAJANI I, ALMED F, ARJOVSKY M, et al. Improved training of wasserstein gans[DB/OL]. 2017(-03-31)[2021-0812]. https://arxiv.org/abs/1704.00028.
[27]	LIN Z N, KHETAN A, FANTI G, et al. PacGAN: The power of two samples in generative adversarial networks[J]. IEEE Journal on Selected Areas in Information Theory, 2020, 1(1):324-335. doi: 10.1109/JSAIT.2020.2983071 URL
[28]	SMITH M J, SALA C, KANTER J M, et al. The machine learning bazaar: Harnessing the ML ecosystem for effective system development[C]// SIGMOD’20: Proceedings of the 2020 ACM SIGMOD International Conference on Management of Data. 2020: 785-800.
[29]	BEAVER J M, BORGES-HINK R C, BUCKNER M A. An evaluation of machine learning methods to detect malicious SCADA communications[C]// 2013 12th International Conference on Machine Learning and Applications. IEEE, 2013: 54-59.

基于改进生成对抗网络的虚假数据注入攻击检测方法

False Data Injection Attack Detection Method Based onImproved Generative Adversarial Network

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 29

相关文章 11

编辑推荐

Metrics

本文评价

[1]	李彬, 魏吟娬, 祁兵, 孙毅, 陈宋宋. 基于EEMD-LSTM的需求响应终端DDoS攻击检测方法[J]. 电力建设, 2022, 43(4): 81-90.
[2]	张杰, 刘洋, 李文峰, 王磊, 许立雄. 基于两阶段数据增强和双向深度残差TCN的用户负荷曲线分类方法[J]. 电力建设, 2022, 43(2): 89-97.
[3]	王竞才, 李琰, 徐天奇. 虚假数据注入攻击建模及攻击下脆弱线路的快速筛选[J]. 电力建设, 2022, 43(1): 104-112.
[4]	朱庆, 郑红娟, 唐子逸, 韦思雅, 邹子骁, 吴熙. 基于生成对抗网络的综合能源负荷场景生成方法[J]. 电力建设, 2021, 42(12): 1-8.
[5]	梁修锐，刘道伟，杨红英，李卫星，邵广惠，徐兴伟，王克非，李宗翰，赵高尚. 数据驱动的电力系统静态电压稳定态势评估[J]. 电力建设, 2020, 41(1): 126-132.
[6]	蔡星浦，王琦，邰伟，刘科研. 基于多阶段博弈的电力CPS虚假数据注入攻击防御方法[J]. 电力建设, 2019, 40(5): 48-54.
[7]	肖泽青，华昊辰，曹军威. 人工智能在能源互联网中的应用综述[J]. 电力建设, 2019, 40(5): 63-70.
[8]	杨懿男，齐林海，王红，苏林萍. 基于生成对抗网络的小样本数据生成技术研究[J]. 电力建设, 2019, 40(5): 71-77.
[9]	戴彦, 王刘旺, 李媛, 颜拥, 韩嘉佳, 文福拴. 新一代人工智能在智能电网中的应用研究综述[J]. 电力建设, 2018, 39(10): 1-11.
[10]	张春, 田芳, 于之虹, 李岩松, 张爽, 田蓓. 改进Sammon映射算法在分析暂态稳定评估输入特征有效性中的应用[J]. 电力建设, 2016, 37(12): 96-.
[11]	王钦，蒋怀光，文福拴，梅天华. 智能电网中大数据的概念、技术与挑战[J]. 电力建设, 2016, 37(12): 1-.