Machine Learning-based Fraud Detection in Electric Vehicle Charging Networks
Year: 2026
Type: conference-paper
Abstract: Electric Vehicles (EVs) are becoming increasingly popular, leading to rapid development of EV charging infrastructure. Along with this growth, issues such as unauthorized charging, energy misuse, and abnormal charging sessions are also increasing. Existing monitoring systems generally depend on fixed threshold values and manual observation, which are not always capable of identifying complicated fraud patterns. In this paper, a machine learning-based approach is proposed for detecting unusual charging behavior in EV charging networks. Since labeled fraud data is difficult to obtain, unsupervised learning techniques are used for anomaly detection. The charging data is first cleaned properly and important parameters are selected for analysis. Then, Isolation Forest and K-Means clustering are used to find charging sessions that behave differently from normal patterns. Different visualizations are also used to study charging behavior and anomaly distribution. The obtained results indicate that the proposed system is capable of detecting abnormal charging sessions with good efficiency. This work shows that machine learning can play an important role in improving the security and reliability of modern EV charging infrastructure.
Institution: Netaji Subhash Engineering College
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7203479306