AI-enhanced smart grid framework for intrusion detection and mitigation in EV charging stations

Arvind R. Singh, R. Seshu Kumar, Rajkumar Singh Rathore*, A. Pandian, Fatma S. Alrayes, Randa Allafi, Nazir Ahmad

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

The rapid growth of electric vehicle (EV) charging stations has brought about substantial security and reliability challenges within smart grid systems. As EV adoption continues to rise, the importance of building a secure and resilient charging infrastructure becomes increasingly evident, especially with the integration of renewable energy sources and decentralized grid operations. This paper introduces an Artificial Intelligence-powered Smart Grid Framework (AI-SGF), developed to tackle intrusion detection and mitigation specifically in EV charging stations. Existing studies have highlighted vulnerabilities in these stations, exposing the smart grid to risks such as unauthorized access, data tampering, and cyber-physical attacks. The AI-SGF addresses these concerns by continuously monitoring real-time charging data, allowing it to detect anomalies and swiftly counter potential cyber threats. The framework has undergone extensive testing, proving its effectiveness with precision, recall, and F1 scores of 96.8 %, 96.0 %, and 96.4 % for reliability, and a cyberattack detection score of 98.9 %. These metrics affirm AI-SGF's capability to exceed current benchmarks in cybersecurity for EV infrastructure, providing robust protection and integration across energy distribution networks. The scalable and adaptable design of AI-SGF further supports seamless deployment within existing EV charging networks, promoting a secure and sustainable EV ecosystem.

Original languageEnglish
JournalAlexandria Engineering Journal
Early online date26 Dec 2024
DOIs
Publication statusE-pub ahead of print - 26 Dec 2024

Keywords

  • Artificial Intelligence (AI)
  • Cyber-physical attacks
  • Electric vehicle
  • Machine Learning (ML)
  • Smart grid

Cite this