Date of Award

Summer 8-22-2026

Document Type

Dissertation

Degree Name

Ph.D. in Electrical Engineering

Organizational Unit

Daniel Felix Ritchie School of Engineering and Computer Science, Electrical and Computer Engineering

First Advisor

David W. Gao

Second Advisor

Mohammad A. Matin

Third Advisor

Rui Fan

Copyright Statement / License for Reuse

All Rights Reserved
All Rights Reserved.

Keywords

Closed-loop validation, Distributed energy resources (DERs), IEEE Std. 1547-2018, Optimal power flow (OPF), Reinforcement learning (RL), Voltage regulation

Abstract

Distribution networks face increasing challenges in maintaining acceptable voltage profiles and phase balance due to their radial structure, high R/X ratios, and the growing integration of distributed energy resources (DERs). While these factors exacerbate voltage magnitude deviations and unbalance, DERs equipped with smart inverters (SIs) also provide new opportunities for grid support through controllable active and reactive power injection. Effectively leveraging these capabilities requires systematic modeling, coordinated optimization, and scalable control under network constraints, while accounting for practical implementation requirements and operational uncertainties.

This dissertation develops a unified framework for voltage regulation and unbalance mitigation in distribution networks with high DER penetration. First, a multi-phase optimal power flow (OPF) formulation based on a linearized LinDistFlow model is established to enable computationally efficient optimization. Building on this foundation, a comprehensive voltage regulation framework is developed that jointly incorporates legacy grid devices and SI-based DER control functionalities under IEEE 1547-compliant modes, enabling coordinated support for system-level voltage regulation and unbalance mitigation. Furthermore, to bridge the gap between theoretical optimization and practical deployment, a closed-loop validation framework is constructed to evaluate OPF-based coordination using high-fidelity AC power flow simulations, thereby assessing model fidelity and practical effectiveness. Additionally, to ensure practical implementability of inverter-based control in real distribution systems, an uncertainty-aware Volt–VAR design approach is developed that explicitly incorporates both single-phase and three-phase inverter control structures while adhering to realizable control constraints. Finally, to address scalability challenges in complex scenarios and large-scale systems, a hybrid reinforcement learning (RL)–OPF framework is introduced for the coordinated control of flexible resources such as electric vehicles (EVs), where the RL layer reduces computational complexity in high-dimensional decision space, while the OPF layer ensures network constraint satisfaction and physical feasibility.

The proposed frameworks enable effective coordination of heterogeneous resources, improve voltage profiles, and mitigate phase unbalance while maintaining computational tractability and physical feasibility. The results demonstrate that combining physics-based optimization, implementable control design, and learning-based decision-making provides a scalable and practical solution for modern distribution networks. Overall, this work bridges the gap between theoretical optimization and real-world deployment, contributing to the reliable and efficient operation of future power systems with high DER integration.

Copyright Date

8-2026

Publication Statement

Copyright is held by the author. User is responsible for all copyright compliance.

Rights Holder

Hongxia Wang

Provenance

Received from ProQuest

File Format

application/pdf

Language

English (eng)

Extent

155 pgs

File Size

4.4 MB

Available for download on Sunday, November 05, 2028



Share

COinS