An Abnormal Data Identification Method Based on Improved Generative Adversarial Network

doi:10.12204/j.issn.1000-7229.2021.05.002

Abstract

Abstract:

The identification method based on data driving for identifying abnormal data in power grid has become the focus of research in the field of power grid security. Due to the small number of abnormal data in the actual statistical data of power generation, it is extremely difficult to identify abnormal data through data mining. This paper proposes an improved generative adversarial network and isolated forests based abnormal data identification method. Firstly, by using WGAN alternating training generator and discriminator, the distribution characteristics of power generation statistical data are learned and samples are generated, which will generate abnormal samples to enhance the original abnormal samples. According to the accuracy of outlier data identification, the expansion ratio of outlier samples is determined. Then the isolated forest algorithm is used to identify the abnormal data on the expanded balanced dataset. Finally, an evaluation index system consisting of accuracy, recall and precision is selected to evaluate and compare the identification effects of models before and after category equalization. The results show that the proposed method can effectively improve the classification preference of the identification model for most classes and improve the overall identification accuracy.

Key words: generative adversarial networks, Wasserstein distance, sample generating, unbanced data, abnormal data identification

CLC Number:

TM73

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.

Figures/Tables 10

Table 1 State code of power plant's statistic data

Fig.1 Abnormal data identification model based on WGAN

Table 2 Specific distribution of power plant's sample statistic data

Table 3 Optimal parameters of abnormal data identification model

Fig.2 Distribution of data with different tampering degree after dimensionality reduction

Table 4 Comparison of cosine similarity results

Table 5 Comparison of classification results before and after data enhancement

Fig.3 ROC of different data synthesis methods

Table 6 Abnormal data identification results of different data synthesis methods

Table 7 Identification results of the combination of data synthesis and anomaly detection

References 21

[1]	国家能源局综合司. 国家能源局综合司关于进一步做好电力监管统计信息报送工作的通知(国能综通综合[2017] 90号)[EB/OL]. (2017-10-18)[2020-05-15]. http://fjb.nea.gov.cn/news_view.aspx?id=28398.
[2]	陈凌, 韩伟, 张经炜. 基于数据融合的光伏组件故障诊断[J]. 电网技术, 2017,41(6):1864-1873.
	CHEN Ling, HAN Wei, ZHANG Jingwei. PV module fault diagnosis based on data fusion[J]. Power System Technology, 2017,41(6):1864-1873.
[3]	颜永龙, 李剑, 李辉, 等. 采用信息熵和组合模型的风电机组异常检测方法[J]. 电网技术, 2015,39(3):737-743. doi: 10.13335/j.1000-3673.pst.2015.03.023 URL
	YAN Yonglong, LI Jian, LI Hui, et al. A wind turbine anomaly detection method based on information entropy and combination model[J]. Power System Technology, 2015,39(3):737-743. doi: 10.13335/j.1000-3673.pst.2015.03.023 URL
[4]	何蓓, 邹波, 周峰, 等. 基于决策树群的多维电能表状态检验技术及其应用[J]. 电测与仪表, 2018,55(7):92-98.
	HE Bei, ZOU Bo, ZHOU Feng, et al. Technology and application of multi-dimensional remote monitoring system for Watt-hour meter based on decision tree group[J]. Electrical Measurement & Instrumentation, 2018,55(7):92-98.
[5]	SHAH A M, BHALJA B R. Fault discrimination scheme for power transformer using random forest technique[J]. IET Generation, Transmission & Distribution, 2016,10(6):1431-1439.
[6]	黄建明, 李晓明, 瞿合祚, 等. 考虑小波奇异信息与不平衡数据集的输电线路故障识别方法[J]. 中国电机工程学报, 2017,37(11):3099-3107.
	HUANG Jianming, LI Xiaoming, QU Hezuo, et al. Method for fault type identification of transmission line considering wavelet singular information and unbalanced dataset[J]. Proceedings of the CSEE, 2017,37(11):3099-3107.
[7]	刘洋, 刘洋, 许立雄, 等. 计及数据类别不平衡的海量用户负荷典型特征高性能提取方法[J]. 中国电机工程学报, 2019,39(14):4093-4104.
	LIU Yang, LIU Yang, XU Lixiong, et al. A high performance extraction method for massive user load typical characteristics considering data class imbalance[J]. Proceedings of the CSEE, 2019,39(14):4093-4104.
[8]	李倩倩, 刘胥影. 多类类别不平衡学习算法: EasyEnsemble.M[J]. 模式识别与人工智能, 2014,27(2):187-192.
	LI Qianqian, LIU Xuying. EasyEnsemble. M for multiclass imbalance problem[J]. Pattern Recognition and Artificial Intelligence, 2014,27(2):187-192.
[9]	刘凯伟, 张冬梅. 基于流形学习的异常检测算法研究[J]. 计算机工程与应用, 2013,49(13):105-109.
	LIU Kaiwei, ZHANG Dongmei. Manifold learning-based anomaly detection algorithm[J]. Computer Engineering and Applications, 2013,49(13):105-109.
[10]	熊冰妍, 王国胤, 邓维斌. 基于样本权重的不平衡数据欠抽样方法[J]. 计算机研究与发展, 2016,53(11):2613-2622. doi: 10.7544/issn1000-1239.2016.20150593 URL
	XIONG Bingyan, WANG Guoyin, DENG Weibin. Under-sampling method based on sample weight for imbalanced data[J]. Journal of Computer Research and Development, 2016,53(11):2613-2622.
[11]	BATISTA G E A P A, PRATI R C, MONARD M C. A study of the behavior of several methods for balancing machine learning training data[J]. ACM SIGKDD Explorations Newsletter, 2004,6(1):20-29. doi: 10.1145/1007730.1007735 URL
[12]	CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique[J]. Journal of Artificial Intelligence Research, 2002,16:321-357. doi: 10.1613/jair.953 URL
[13]	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.
[14]	GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Advances in Neural Information Processing Systems, 2014,3:2672-2680.
[15]	高强, 姜忠昊. 基于GAN等效模型的小样本库扩增研究[J]. 电测与仪表, 2019,56(6):76-81.
	GAO Qiang, JIANG Zhonghao. Amplification of small sample library based on GAN equivalent model[J]. Electrical Measurement & Instrumentation, 2019,56(6):76-81.
[16]	谭本东, 杨军, 赖秋频, 等. 基于改进CGAN的电力系统暂态稳定评估样本增强方法[J]. 电力系统自动化, 2019,43(1):149-157.
	TAN Bendong, YANG Jun, LAI Qiupin, et al. Data augment method for power system transient stability assessment based on improved conditional generative adversarial network[J]. Automation of Electric Power Systems, 2019,43(1):149-157.
[17]	刘云鹏, 许自强, 和家慧, 等. 基于条件式Wasserstein生成对抗网络的电力变压器故障样本增强技术[J]. 电网技术, 2020,44(4):1505-1513.
	LIU Yunpeng, XU Ziqiang, HE Jiahui, et al. Data augmentation method for power transformer fault diagnosis based on conditional Wasserstein generative adversarial network[J]. Power System Technology, 2020,44(4):1505-1513.
[18]	LIU Y F, ZHOU Y, LIU X, et al. Wasserstein GAN-based small-sample augmentation for new-generation artificial intelligence: A case study of cancer-staging data in biology[J]. Engineering, 2019,5(1):156-163. doi: 10.1016/j.eng.2018.11.018 URL
[19]	张煜, 吕锡香, 邹宇聪, 等. 基于生成对抗网络的文本序列数据集脱敏[J]. 网络与信息安全学报, 2020,6(4):109-119.
	ZHANG Yu, LYU Xixiang, ZOU Yucong, et al. Differentially private sequence generative adversarial networks for data privacy masking[J]. Chinese Journal of Network and Information Security, 2020,6(4):109-119.
[20]	田力, 向敏. 基于密度聚类技术的电力系统用电量异常分析算法[J]. 电力系统自动化, 2017,41(5):64-70.
	TIAN Li, XIANG Min. Abnormal power consumption analysis based on density-based spatial clustering of applications with noise in power systems[J]. Automation of Electric Power Systems, 2017,41(5):64-70.
[21]	廖伟涵, 郭创新, 金宇, 等. 基于四阶段预处理与GBDT的油浸式变压器故障诊断方法[J]. 电网技术, 2019,43(6):2195-2203.
	LIAO Weihan, GUO Chuangxin, JIN Yu, et al. Oil-immersed transformer fault diagnosis method based on four-stage preprocessing and GBDT[J]. Power System Technology, 2019,43(6):2195-2203.