ROBUST AND ADAPTIVE VOLT-VAR AND DEMAND RESPONSE STRATEGIES FOR ACTIVE DISTRIBUTION NETWORKS UNDER COMPLEX UNCERTAINTIES
Loading...
Authors
Najafi, Soroush
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Alternative Title
Abstract
The rapid proliferation of distributed energy resources (DERs)—such as photovoltaic (PV) systems, battery energy storage systems (BESS), and flexible loads, including residential and commercial buildings with heating, ventilation, and air conditioning (HVAC) systems, as well as electric vehicle charging stations (EVCS)—has created tremendous opportunities for grid services while also introducing unprecedented operational challenges for active distribution networks (ADNs). This dissertation comprehensively addresses these emerging complexities through the development and validation of innovative optimization frameworks, integrating Volt-VAR optimization (VVO) and demand response programs (DRP) in ADNs, particularly within the context of commercial buildings. Recognizing the critical importance of accurate uncertainty representation, this dissertation adopts a Gaussian Mixture Model-based Chance-Constrained Optimization (GMM-CCO) approach. This methodology effectively characterizes complex, non-Gaussian uncertainties prevalent in PV generation, load fluctuations, and potential extreme conditions, thereby significantly enhancing the robustness and resilience of operational strategies. The GMM-CCO enables ADNs to withstand both typical forecast deviations and rare, high-impact scenarios, ensuring continuous and reliable performance without compromising operational efficiency. Further advancing the field of ADNs control strategies, this dissertation introduces an adaptive Q-V droop control methodology underpinned by an offline Extremum-Seeking (ES) algorithm. This novel approach leverages local Thevenin equivalent estimations, performed autonomously by edge processors at inverter nodes, thereby avoiding intrusive real-time network perturbations. The ES algorithm dynamically calibrates droop settings offline, producing stable, optimized reactive power injections. This significantly mitigates voltage fluctuations, improves power quality, and extends the operational lifespan of inverter-based systems. The decentralization of control processes reduces communication burdens, facilitating practical deployment in large-scale ADNs. The methodologies proposed in this dissertation have been rigorously validated through comprehensive simulations on widely recognized IEEE benchmark systems, including the IEEE 13-node, IEEE 37-node, IEEE 69-node, and IEEE 123-node test feeders. These diverse simulation environments confirm substantial improvements in key operational metrics, highlighting significant reductions in voltage deviations, network energy losses, and overall operational costs. The scalability and adaptability of the proposed solutions are evident from their consistent performance across various network sizes and configurations, demonstrating practical applicability and effectiveness in real-world scenarios. Additionally, this dissertation emphasizes targeted demand-side flexibility from commercial buildings. By strategically focusing on high-value commercial loads—such as HVAC and EV charging facilities—the dissertation offers practical insights into overcoming implementation barriers associated with broader demand response programs. This targeted approach simplifies coordination, reduces operational complexity, and maximizes the impact of demand-side management strategies, thereby enhancing both economic and operational outcomes for distribution system operators (DSOs). Overall, the proposed frameworks and methodologies offer robust and scalable solutions for the integrated management of voltage regulation and demand-side flexibility in modern ADNs. By leveraging advanced statistical uncertainty modeling, adaptive control techniques, and targeted demand response strategies, this dissertation significantly contributes to enhancing grid resilience, reliability, and operational efficiency, providing valuable insights and practical solutions to effectively manage the evolving landscape of distribution networks.