AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Novel distributed coordination control for DC microgrid: Modeling, Analysis and Optimization

Authors

;

Term

3. term

Publication year

2017

Submitted on

Pages

14

Abstract

Denne afhandling adresserer udvikling af ny, distribueret koordinationsstyring for ø‑koblede DC‑mikronet med fokus på modellering, analyse og design af sekundære styrealgoritmer. Udgangspunktet er en hierarkisk struktur med droop‑baseret primær styring, hvis kendte begrænsninger er spændingsafvigelser og netophed i strømdeling ved ikke‑negligerbare linjeimpedanser. Der afledes detaljerede, gennemsnitlige tilstandsrumsmodeller af indre reguleringssløjfer, droop‑styring og DC‑topologi for et multi‑konverter system, og følsomheden over for strukturelle og regulatoriske parametre analyseres. En central del er en afvejning mellem præcis strømdeling og spændingsregulering, som bruges til at vejlede valg af styreparametre. På basis heraf udformes en containment‑baseret spændingskoordinering, der holder mikronettets spænding inden for et foreskrevet interval, samt en dynamisk consensus‑baseret regulator, der opnår relativt præcis strømdeling via nabobaseret kommunikation. PI‑parametre tunes ved hjælp af egenværdikurver, og de afledte modeller verificeres ved sammenligning med MATLAB/Simulink, hvor de foreslåede distribuerede controllere demonstreres. Den distribuerede arkitektur forbedrer samtidig robusthed og skalerbarhed, idet hver omformer kun udveksler information med sine naboer. (Detaljerede kvantitative resultater ligger uden for dette uddrag.)

This thesis develops a novel distributed coordination control strategy for islanded DC microgrids, focusing on modeling, analysis, and secondary controller design. The work builds on a hierarchical architecture with droop‑based primary control, whose limitations include steady‑state voltage deviation and poor current sharing in the presence of line impedances. Detailed averaged state‑space models of the inner control loops, droop control, and the multi‑converter DC microgrid are derived, and sensitivity to structural and controller parameters is examined. A key contribution is a trade‑off analysis between tight voltage regulation and accurate current sharing that guides controller parameter selection. Based on this, a containment‑based voltage coordinator is proposed to keep the bus voltage within a prescribed range, and a dynamic consensus‑based current regulator is designed to achieve relatively accurate current sharing using neighbor‑to‑neighbor communication. PI gains are tuned via eigenvalue locus analysis, and the mathematical models are validated against MATLAB/Simulink implementations, where the proposed distributed controllers are demonstrated. The distributed architecture enhances reliability and scalability since each converter exchanges data only with its neighbors. (Detailed quantitative results are beyond this excerpt.)

[This summary has been generated with the help of AI directly from the project (PDF)]