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.

Detecting Anomalous Users via Streaming Data Processing in Smart Grid

Research on Clustering Algorithm of Load Decomposition Considering Harmonic Characteristics in Power Safety Monitoring

Regression-based Online Anomaly Detection for Smart Grid Data

Anomaly Detection Algorithm Based on Electric Equipment

Power Consumption-based Anomaly Detection for Relay Protection

Anomaly detection for streaming data based on Grid-clustering and Gaussian distribution

MQTT based Time-Series Anomaly Detection System for Smart Grid

Graph-Based Time Series Edge Anomaly Detection in Smart Grid

Time Series Anomaly Detection for Smart Grids: A Survey

Theft of Electricity Detection Based on Supervised Learning

Anomaly detection of power grid data based on storm cloud platform sliding window

Anomaly Detection in Smart Grids using Machine Learning

Anomaly Detection in Smart Grid Application

Blockchain-based anomaly detection of electricity consumption in smart grids

Dynamic Micro-cluster-Based Streaming Data Clustering Method for Anomaly Detection.

Using Power Side-Channel to Implement Anomaly-Based Intrusion Detection on Smart Grid Terminals

Review of Detection Methods for Abnormal Electricity Consumption Data in Smart Grid

Modified Masking-Based Federated Singular Value Decomposition Method for Fast Anomaly Detection in Smart Grid Systems

Anomaly-based Framework for Detecting Power Overloading Cyberattacks in Smart Grid AMI

A Distributed Anomaly Detection Method of Operation Energy Consumption Using Smart Meter Data

Anomaly Detection for Power Quality Analysis Using Smart Metering Systems