Review of non-technical loss detection methods

Messinis, George M.; Hatziargyriou, Nikos D.

doi:10.1016/j.epsr.2018.01.005

Cited by 187 publications

(116 citation statements)

References 40 publications

Supporting

Mentioning

108

Contrasting

Unclassified

Order By: Relevance

“…For confidentiality reasons, the variables information are condensed in classes as described in Table 1. The resulting features are very similar to the ones described in [Messinis and Hatziargyriou 2018]. The fraud event variable describes exclusively the fraudulent or non-fraudulent events observed by the company investigators.…”

Section: Datasetsmentioning

confidence: 56%

“…Several applications of supervised and unsupervised machine learning algorithms for prediction of fraud and irregularity in electric utility can be found in literature. [Messinis and Hatziargyriou 2018]. Examples of implementations of usual supervised methods include the application of support vector machines to identify customer's abnormal consumption behavior based on previous energy usage data [Nagi et al 2010, Alfarra et al 2018.…”

Section: Related Workmentioning

confidence: 99%

“…They contain data of a medium-size brazilian city with around 700 thousands customers between February of 2014 and September of 2018. Considering [Messinis and Hatziargyriou 2018] definitions, the dataset contains low-resolution energy data, with a time resolution of one day, at consumer level. After feature engineering and categories aggregation a total of 64 features was employed in statistical modeling studies.…”

Section: Datasetsmentioning

confidence: 99%

See 2 more Smart Citations

Employing Gradient Boosting and Anomaly Detection for Prediction of Frauds in Energy Consumption

Albiero¹,

Santos²,

Uyrá³

et al. 2019

Anais Do XVI Encontro Nacional De Inteligência Artificial E Computacional (ENIAC 2019)

View full text Add to dashboard Cite

show abstract

Section: Datasetsmentioning

confidence: 56%

Section: Related Workmentioning

confidence: 99%

Section: Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

Employing Gradient Boosting and Anomaly Detection for Prediction of Frauds in Energy Consumption

Albiero¹,

Santos²,

Uyrá³

et al. 2019

Anais Do XVI Encontro Nacional De Inteligência Artificial E Computacional (ENIAC 2019)

View full text Add to dashboard Cite

show abstract

“…Concerning residential consumers with power theft scenario 3, it can be said that it is impossible to be detected with the proposed method because no smart meters are connected to the power grid in that scenario. It is worth noting that other equally efficient metrics could be: AUC (area under curve), precision metric, F1 score, etc [18].…”

Section: Resultsmentioning

confidence: 99%

Efficient Power Theft Detection for Residential Consumers Using Mean Shift Data Mining Knowledge Discovery Process

Blazakis¹,

Stavrakakis²

2019

IJAIA

View full text Add to dashboard Cite

Energy theft constitutes an issue of great importance for electricity operators. The attempt to detect and reduce non-technical losses is a challenging task due to insufficient inspection methods. With the evolution of advanced metering infrastructure (AMI) in smart grids, a more complicated status quo in energy theft has emerged and many new technologies are being adopted to solve the problem. In order to identify illegal residential consumers, a computational method of analyzing and identifying electricity consumption patterns of consumers based on data mining techniques has been presented. Combining principal component analysis (PCA) with mean shift algorithm for different power theft scenarios, we can now cope with the power theft detection problem sufficiently. The overall research has shown encouraging results in residential consumers power theft detection that will help utilities to improve the reliability, security and operation of power network.

show abstract

“…With the appearance of the smart grid comes a great deal of smart meter (SM) data and extra opportunities to solve NTL. Hence, a lot of data oriented methods have been proposed recently, due to the development of machine learning and ease of implementation [5]. Researchers adopt methods of different fields of knowledge with machine learning, such as anomaly detection, cybersecurity, etc.…”

Section: Introductionmentioning

confidence: 99%

Knowledge Embedded Semi-Supervised Deep Learning for Detecting Non-Technical Losses in the Smart Grid

Zhou

Wang

et al. 2019

Energies

View full text Add to dashboard Cite

Non-technical losses (NTL) caused by fault or electricity theft is greatly harmful to the power grid. Industrial customers consume most of the power energy, and it is important to reduce this part of NTL. Currently, most work concentrates on analyzing characteristic of electricity consumption to detect NTL among residential customers. However, the related feature models cannot be adapted to industrial customers because they do not have a fixed electricity consumption pattern. Therefore, this paper starts from the principle of electricity measurement, and proposes a deep learning-based method to extract advanced features from massive smart meter data rather than artificial features. Firstly, we organize electricity magnitudes as one-dimensional sample data and embed the knowledge of electricity measurement in channels. Then, this paper proposes a semi-supervised deep learning model which uses a large number of unlabeled data and adversarial module to avoid overfitting. The experiment results show that our approach can achieve satisfactory performance even when trained by very small samples. Compared with the state-of-the-art methods, our method has achieved obvious improvement in all metrics.

show abstract

Review of non-technical loss detection methods

Cited by 187 publications

References 40 publications

Employing Gradient Boosting and Anomaly Detection for Prediction of Frauds in Energy Consumption

Employing Gradient Boosting and Anomaly Detection for Prediction of Frauds in Energy Consumption

Efficient Power Theft Detection for Residential Consumers Using Mean Shift Data Mining Knowledge Discovery Process

Knowledge Embedded Semi-Supervised Deep Learning for Detecting Non-Technical Losses in the Smart Grid

Contact Info

Product

Resources

About