Multi-Agent Deep Reinforcement Learning-Driven Mitigation of Adverse Effects of Cyber-Attacks on Electric Vehicle Charging Station

An electric vehicle charging station (EVCS) infrastructure is the backbone of transportation electrification. However, the EVCS has myriads of exploitable vulnerabilities in software, hardware, supply chain, and incumbent legacy technologies such as network, communication, and control. These standalone or networked EVCS open up large attack surfaces for the local or state-funded adversaries. The state-of-the-art approaches are not agile and intelligent enough to defend against and mitigate advanced persistent threats (APT). We propose the data-driven model-free distributed intelligence based on multiagent Deep Reinforcement Learning (MADRL)-- Twin Delayed Deep Deterministic Policy Gradient (TD3) -- that efficiently learns the control policy to mitigate the cyberattacks on the controllers of EVCS. Also, we have proposed two additional mitigation methods: the manual/Bruteforce mitigation and the controller clone-based mitigation. The attack model considers the APT designed to malfunction the duty cycles of the EVCS controllers with Type-I low-frequency attack and Type-II constant attack. The proposed model restores the EVCS operation under threat incidence in any/all controllers by correcting the control signals generated by the legacy controllers. Also, the TD3 algorithm provides higher granularity by learning nonlinear control policies as compared to the other two mitigation methods. Index Terms: Cyberattack, Deep Reinforcement Learning(DRL), Electric Vehicle Charging Station, Mitigation.