Abstract:With the rapid development of infrastructure including power grids, managers in power industry these days are faced with an increasingly severe problem of theft of electricity. Theft of electricity has negative effects on many socioeconomic aspects, including impacting stable growth of economic for power enterprises and social development. The traditional anti-theft means require officers checking the integrity of kilowatt-hour meter and the correctness of wiring house by house, which requires enormous manpower and material resources. With the advancement of information collecting technology, power enterprises now possess relatively complete database of power consumption. As a result, performing data mining on existing database and identifying abnormal users has become a hot topic in the field of information technology. The purpose of this paper is to identify abnormal users with machine learning algorithms, providing power enterprises with means to detect theft of electricity at lower cost. The contributions of this paper are listed as follows: 1) Propose various features, providing means to describe users’ behaviors. First, monistic features (mean, difference, coefficient of variation, range, standard deviation), binary features (cosine similarity, Pearson product-moment correlation coefficient) and multivariate features are calculated based on user power consumption (monthly, seasonally, yearly, per holiday, per workday). Then principal component analysis is used to perform dimensionality reduction on the dataset. The dataset is finally scaled and oversampled to form the feature dataset. 2) Study various classification algorithms, optimize hyperparameters, providing models to detect theft of electricity. In this paper, the mechanism behind support vector machine, back-propagation neural network, random forest and XGBoost is studied and the hyperparameters are optimized. 3) Compare and evaluate different classifiers, and provide suggestions for real-world application. In this paper, different classifiers are evaluated with different metrics and the best classifier is recommended based on real-world dataset and different application scenarios.

Abnormal detection and early warning mechanism of power market transaction data: a practice and exploration

Anomaly Detection Algorithm Based on Electric Equipment

Scenario Analysis and Anomaly Detection of Locational Marginal Price in Electricity Market

Anomaly Detection for Power Consumption Patterns Based on Unsupervised Learning

Method for Identifying Abnormal Power Metering Data in Shared Power Information Collection System

A Data-Driven Pattern Extraction Method for Analyzing Bidding Behaviors in Power Markets

Theft of Electricity Detection Based on Supervised Learning

Enhancing power utilization analysis: detecting aberrant patterns of electricity consumption

Anomaly detection of power grid dispatching platform based on Isolation Forest and K-means fusion algorithm

Anomaly electricity detection method based on entropy weight method and isolated forest algorithm

Identification of market power abuse in Chinese electricity market based on an improved cost-sensitive support vector machine

Research on Simulation and identification method of abnormal bidding behavior of power market members based on cloud technology

An Method for Anomaly Detection of State Information of Power Equipment Based on Big Data Analysis

KNN-ADMM Based Online Missing Data Recovery in Electricity Markets

A Long-and Short-term Hot User Identification Method Based on Local Outlier Density

Autoencoder and Kernel Density Estimation-Based Anomaly Detection Method for Power Consumption among Distribution Network Customers

Abnormal Electricity Customer Clustering Method Based on Electricity Big Data

Realizing the Improvement of the Reliability and Efficiency of Intelligent Electricity Inspection: IAOA-BP Algorithm for Anomaly Detection

Active-Passive-Combined Anomaly Detection in Electricity-Carbon Blockchain Trading System under Limited Bandwidth

Detection and Classification of Anomalies in Power Distribution System Using Outlier Filtered Weighted Least Square

Abnormal Electricity Detection with Hybrid Deep Neural Network Model