Electricity Theft Detection to Reduce Non-Technical Loss using Support Vector Machine in Smart Grid

Toma, Rafia Nishat; Hasan, Nazmul; Nahid, Abdullah-Al; Li, Bo

doi:10.1109/icasert.2019.8934601

Cited by 23 publications

(11 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…International Conference on Critical Information Infrastructures Security (CRITIS) [74] International Conference on Power Systems (ICPS) [75] North American Power Symposium, NAPS 2015 [76] IEEE Symposium on Computational Intelligence and Applications in Smart Grid. [77] International Conference on Advances in Science, Engineering and Robotics Technology 2019 [78] IEEE Power and Energy Society Innovative Smart Grid Technologies Conference. [79] 2017 IEEE Power and Energy Conference at Illinois [80] Energies 2020, 13, 4727 5 of 25…”

Section: Prisma Flowchartmentioning

confidence: 99%

See 1 more Smart Citation

Detection of Non-Technical Losses in Power Utilities—A Comprehensive Systematic Review

et al. 2020

View full text Add to dashboard Cite

Electricity theft and fraud in energy consumption are two of the major issues for power distribution companies (PDCs) for many years. PDCs around the world are trying different methodologies for detecting electricity theft. The traditional methods for non-technical losses (NTLs) detection such as onsite inspection and reward and penalty policy have lost their place in the modern era because of their ineffective and time-consuming mechanism. With the advancement in the field of Artificial Intelligence (AI), newer and efficient NTL detection methods have been proposed by different researchers working in the field of data mining and AI. The AI-based NTL detection methods are superior to the conventional methods in terms of accuracy, efficiency, time-consumption, precision, and labor required. The importance of such AI-based NTL detection methods can be judged by looking at the growing trend toward the increasing number of research articles on this important development. However, the authors felt the lack of a comprehensive study that can provide a one-stop source of information on these AI-based NTL methods and hence became the motivation for carrying out this comprehensive review on this significant field of science. This article systematically reviews and classifies the methods explored for NTL detection in recent literature, along with their benefits and limitations. For accomplishing the mentioned objective, the opted research articles for the review are classified based on algorithms used, features extracted, and metrics used for evaluation. Furthermore, a summary of different types of algorithms used for NTL detection is provided along with their applications in the studied field of research. Lastly, a comparison among the major NTL categories, i.e., data-based, network-based, and hybrid methods, is provided on the basis of their performance, expenses, and response time. It is expected that this comprehensive study will provide a one-stop source of information for all the new researchers and the experts working in the mentioned area of research.

show abstract

Section: Prisma Flowchartmentioning

confidence: 99%

“…SVM has been used plenty of times as the binary classifiers for NTL detection problem because they are immune to the class imbalance issue [1,5,15,21,28,31,40,43,59,72,78]. Many different methodologies have been used, including the cost-sensitive SVM (CS-SVM) and one-class SVM (OC-SVM).…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

Detection of Non-Technical Losses in Power Utilities—A Comprehensive Systematic Review

et al. 2020

View full text Add to dashboard Cite

show abstract

“…According to the mutual information (MI) and the correlation coefficient between features, the optimal feature subset was selected based on the common feature set. Aydin and Gungor [21] and Toma and Li [22] presented two feature construction techniques for NTL characterization. Compared with algorithm improvement, they were more concerned about finding the set of features that best discriminate legal and illegal profiles.…”

Section: Feature Engineering To Obtain the Optimal Feature Setmentioning

confidence: 99%

“…After feature acquisition, the methods of detection can be divided into supervised and unsupervised methods according to whether label information is available in the dataset. Supervised methods include various classifiers and neural network models, such as support vector machines [17], [22], extreme learning machines [23], random forests [24], and deep learning algorithms. Unsupervised methods primarily include a variety of clustering algorithms, e.g., k-means clustering [25], fuzzy clustering [26], and other improved clustering methods [2], [27], [28].…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Detection of Abnormal Electricity Consumption Behavior Based on Feature Engineering

Zhang¹,

Dong²,

Li³

et al. 2020

IEEE Access

View full text Add to dashboard Cite

The detection of abnormal electricity consumption behavior has been of great importance in recent years. However, existing research often focuses on algorithm improvement and ignores the process of obtaining features. The optimal feature set, which reflects customers' electricity consumption behavior, has a significant influence on the final detection results. Moreover, it is not straightforward to obtain datasets with label information. In this paper, a method based on feature engineering for unsupervised detection of abnormal electricity consumption behavior is proposed. First, the original feature set is constructed by brainstorming in the feature engineering step. Then, the optimal feature set, which reflects the customers' electricity consumption behavior, is obtained by features selected based on the variance and similarity between them. After that, in the abnormal detection step, a density-based clustering algorithm, in which the best clustering parameters are selected through iteration and evaluation, combined with unsupervised clustering evaluation indexes, is used to detect abnormal electricity consumption behaviors. Finally, using the load dataset of an industrial park, several typical feature strategies are applied for comparison with the feature engineering proposed in this paper. To perform the evaluation, the label information of abnormal behaviors is obtained by combining the original electricity consumption behavior detection results with abnormal data injections. The abnormal detection method proposed has given good results and outperformed typical feature strategies in an effective and generalizable way.

show abstract

“…This kind of method does not need a labeled data set to train the model, but it is difficult to set the threshold and the detection accuracy is low [4,5]. Supervised classification methods mainly include traditional data mining models such as support vector machine (SVM), multi-layer perceptron (MLP), Bayesian network, extreme gradient boosting tree (XGBoost) [6][7][8][9][10], and new deep learning technologies such as the deep belief network and convolutional neural network (CNN) [11][12][13]. Specifically, SVM is very suitable for binary classification.…”

Section: Introductionmentioning

confidence: 99%

Data Augmentation for Electricity Theft Detection Using Conditional Variational Auto-Encoder

Gong

Zhu

et al. 2020

Energies

View full text Add to dashboard Cite

Due to the strong concealment of electricity theft and the limitation of inspection resources, the number of power theft samples mastered by the power department is insufficient, which limits the accuracy of power theft detection. Therefore, a data augmentation method for electricity theft detection based on the conditional variational auto-encoder (CVAE) is proposed. Firstly, the stealing power curves are mapped into low dimensional latent variables by using the encoder composed of convolutional layers, and the new stealing power curves are reconstructed by the decoder composed of deconvolutional layers. Then, five typical attack models are proposed, and the convolutional neural network is constructed as a classifier according to the data characteristics of stealing power curves. Finally, the effectiveness and adaptability of the proposed method is verified by a smart meters’ data set from London. The simulation results show that the CVAE can take into account the shapes and distribution characteristics of samples at the same time, and the generated stealing power curves have the best effect on the performance improvement of the classifier than the traditional augmentation methods such as the random oversampling method, synthetic minority over-sampling technique, and conditional generative adversarial network. Moreover, it is suitable for different classifiers.

show abstract

Electricity Theft Detection to Reduce Non-Technical Loss using Support Vector Machine in Smart Grid

Cited by 23 publications

References 27 publications

Detection of Non-Technical Losses in Power Utilities—A Comprehensive Systematic Review

Detection of Non-Technical Losses in Power Utilities—A Comprehensive Systematic Review

Unsupervised Detection of Abnormal Electricity Consumption Behavior Based on Feature Engineering

Data Augmentation for Electricity Theft Detection Using Conditional Variational Auto-Encoder

Contact Info

Product

Resources

About