Teng Zhang, XuSheng Qian, Yu Zhou, GaoJun Xu, Ming Wu
{"title":"Robust Clustering and Anomaly Detection of User Electricity Consumption Behavior Based on Correntropy","authors":"Teng Zhang, XuSheng Qian, Yu Zhou, GaoJun Xu, Ming Wu","doi":"10.1049/gtd2.70027","DOIUrl":null,"url":null,"abstract":"<p>Anomaly detection in power systems is crucial for ensuring the safety and stability of electrical grids. Traditional methods struggle to extract meaningful features from electricity consumption data due to significant differences in usage patterns across various user types, such as residential and industrial users. Applying a single model for all user categories increases feature complexity and computational demands. Additionally, non-Gaussian outliers caused by equipment and measurement noise can significantly deviate from normal data patterns, making them difficult to filter using standard methods. To address these challenges, this paper proposes a robust, user-type-specific anomaly detection method. After data preprocessing, a correntropy-based K-means clustering method is used to separate users with noisy data. A two-stage detection framework combining fuzzy logic and a convolutional neural network (CNN)-long short-term memory (LSTM) model enhances both detection efficiency and accuracy. The experiments were conducted using open-source datasets, and the results demonstrated that our method achieved an accuracy of 95%, which is approximately 4% higher than the traditional Isolation Forest method. This indicates that our approach effectively balances efficiency and accuracy in anomaly detection, with its generalizability further validated on an additional dataset.</p>","PeriodicalId":13261,"journal":{"name":"Iet Generation Transmission & Distribution","volume":"19 1","pages":""},"PeriodicalIF":2.0000,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/gtd2.70027","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Iet Generation Transmission & Distribution","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1049/gtd2.70027","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
Anomaly detection in power systems is crucial for ensuring the safety and stability of electrical grids. Traditional methods struggle to extract meaningful features from electricity consumption data due to significant differences in usage patterns across various user types, such as residential and industrial users. Applying a single model for all user categories increases feature complexity and computational demands. Additionally, non-Gaussian outliers caused by equipment and measurement noise can significantly deviate from normal data patterns, making them difficult to filter using standard methods. To address these challenges, this paper proposes a robust, user-type-specific anomaly detection method. After data preprocessing, a correntropy-based K-means clustering method is used to separate users with noisy data. A two-stage detection framework combining fuzzy logic and a convolutional neural network (CNN)-long short-term memory (LSTM) model enhances both detection efficiency and accuracy. The experiments were conducted using open-source datasets, and the results demonstrated that our method achieved an accuracy of 95%, which is approximately 4% higher than the traditional Isolation Forest method. This indicates that our approach effectively balances efficiency and accuracy in anomaly detection, with its generalizability further validated on an additional dataset.
期刊介绍:
IET Generation, Transmission & Distribution is intended as a forum for the publication and discussion of current practice and future developments in electric power generation, transmission and distribution. Practical papers in which examples of good present practice can be described and disseminated are particularly sought. Papers of high technical merit relying on mathematical arguments and computation will be considered, but authors are asked to relegate, as far as possible, the details of analysis to an appendix.
The scope of IET Generation, Transmission & Distribution includes the following:
Design of transmission and distribution systems
Operation and control of power generation
Power system management, planning and economics
Power system operation, protection and control
Power system measurement and modelling
Computer applications and computational intelligence in power flexible AC or DC transmission systems
Special Issues. Current Call for papers:
Next Generation of Synchrophasor-based Power System Monitoring, Operation and Control - https://digital-library.theiet.org/files/IET_GTD_CFP_NGSPSMOC.pdf