PhD Oral Exam - Mohammad Mahdi Soleymani, Information and Systems Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

The growing adoption of electric vehicles (EVs) provides new opportunities to support power system operation through coordinated charging and discharging. EVs can contribute to services such as frequency regulation and distribution system restoration. However, providing these services requires reliable communication among EV charging stations, EV aggregators (EVAs), and power system operators. This reliance on communication also introduces cybersecurity risks, as attackers may compromise charging infrastructure or manipulate the information and control commands exchanged among these entities. This thesis investigates these risks and develops methods to improve the cybersecurity and resilience of EV-integrated power systems.

First, this thesis investigates the impact of EV-based dynamic load-altering attacks (EV-DLAAs) on power system stability. A data-enabled predictive attack model is developed to determine coordinated attack trajectories using system measurements without requiring perfect knowledge of the power system topology or dynamic parameters. Simulation results demonstrate that coordinated EV-DLAAs can cause frequency instability by manipulating EV charging demand across multiple load buses. To address the localization of such attacks, a phasor measurement unit (PMU)-based framework is developed by combining sparse identification of nonlinear dynamics (SINDy) with a modified basis pursuit de-noising approach. Simulation results demonstrate that the proposed framework accurately identifies compromised load buses and estimates attack magnitudes in real time, without requiring offline training on labelled attack data.

The thesis then addresses resilient frequency regulation with EV participation under covert cyberattacks. A resilient model predictive control framework is developed to protect both feedback measurements and forward control commands. The framework combines primary and auxiliary predictive controllers and a communication restoration mechanism to address both short and prolonged attacks. Real-time simulation results demonstrate that the proposed framework maintains acceptable frequency and tie-line power responses under the considered attacks while respecting generator operating limits and available EV regulation capacity.

Finally, a cyber-resilient methodology is developed for EV-assisted distribution system restoration under false data injection attacks targeting communication between the distribution system operator and the EVA. The methodology combines rolling-horizon stochastic-robust optimization, a trusted EVA safety governor, and corrective support from mobile energy storage systems (MESSs). It coordinates EV charging and discharging, renewable generation, MESSs, and repair activities while prioritizing critical loads and respecting network and EV operating constraints. Simulation results demonstrate that EV participation accelerates critical-load restoration under normal conditions. Under cyberattacks, the proposed methodology enhances critical-load recovery compared with restoration without the proposed methodology.

Overall, this thesis highlights the cybersecurity risks associated with the growing integration of EVs into power systems and develops methods to characterize and identify cyberattacks and maintain essential grid services under compromised communication and control conditions.