Author ORCID Identifier

https://orcid.org/0000-0003-0635-3782

Semester

Summer

Date of Graduation

2026

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Anurag Srivastava

Committee Member

Muhammad Choudhry

Committee Member

Gianfranco Doretto

Committee Member

K. Subramani

Committee Member

David Tucker

Abstract

The increasing integration of distributed energy resources (DERs), advanced sensing, and communication infrastructure is transforming modern power grids into highly interconnected cyber-physical systems. While this enhances observability and controllability, it simultaneously exposes the coupled transmission-distribution grid to cyber threats, dynamic instabilities, and climate-driven extreme events. Addressing these converging challenges requires resilience frameworks that can assess vulnerability and enable intelligent, adaptive control under uncertainty. This dissertation develops a unified methodology for resilient cyber-physical power system operation through attack impact analysis, graph-based digital twins, and hazard-aware intelligent control.

The first part investigates the impact of Dynamic Load-Altering Attacks (DLAAs) on transmission and distribution systems. A cyber-physical simulation framework models coordinated malicious load perturbations and evaluates their effect on voltage stability, frequency response, electromechanical dynamics, and cascading operational stress across interconnected grid layers. The analysis demonstrates how spatially coordinated, temporally varying load manipulation can induce severe dynamic deviations and degrade situational awareness, underscoring the need for fast detection and resilient mitigation in increasingly digitized grids.

The second part proposes a graph-based digital twin framework for resilient dynamic control of DER-rich distribution networks. A spatiotemporal learning architecture integrates graph-based representation learning with temporal sequence modeling to predict real-time control actions for inverter-based DERs and battery storage systems. By embedding physical network topology and system dynamics into a data-driven twin, the framework delivers scalable, adaptive voltage support and corrective control under evolving grid conditions.

The third part extends resilience to climate-driven hazards by developing a weather-aware operational framework for intelligent proactive and corrective control. Integrating environmental risk indicators, infrastructure exposure, and operational decision-making, the framework enables risk-informed preemptive and post-disturbance strategies that reduce outage propagation while preserving service continuity and grid stability. Collectively, this dissertation advances cyber-physical grid resilience by combining physics-informed simulation, graph-based machine learning, and intelligent operational decision support for next-generation power system management.

Download

Available for download on Wednesday, July 28, 2027

Share

COinS