This thesis investigates active control strategies for voltage balancing in a Flying Capacitor Multilevel Converter (FCMLC) with output filter operating under both DC/DC and DC/AC configurations. The study addresses a critical scenario often encountered in renewable energy and power storage systems: a sudden and unpredictable input voltage drop. In such cases, maintaining stable output voltage becomes essential, as voltage losses can compromise the performance of connected loads and the overall efficiency of the power conversion system. The primary objectives of this work are twofold: first, to ensure voltage balancing across the flying capacitors, crucial for minimizing waveform distortion and enhancing converter stability; and second, to maintain the desired output voltage level regardless of fluctuations in the DC input. To meet these goals, the thesis evaluates three control methods: Feedback Linearization, Sliding Mode Control, and an algorithm-based control, with a comparative analysis of the strengths and limitations of each approach. Feedback Linearization achieves high precision and robust closed-loop control across both configurations, Sliding Mode Control demonstrates strong robustness and accuracy in DC/DC applications but underperforms in DC/AC scenarios and Algorithm-Based Control provides computational efficiency and greater adaptability, but no robustness against disturbances. Managing capacitor voltage balancing is complex, as input voltage drops are unpredictable. For this reason in this thesis it has been developed a novel reference estimation technique that enables rapid capacitor voltage stabilization. Rather than waiting for DC voltage transients to end, this approach uses early-stage model identification to map the initial behavior of the DC voltage drop to its final stabilized value. This reference estimation method achieves over 99% accuracy, enhancing the converter’s response to sudden input variations by predicting and adjusting in real time. These findings contribute advancing the field of power electronics for dynamic energy systems.
In questa tesi vengono analizzate alcune strategie di controllo attivo per bilanciare le tensioni nei Convertitori Multilivello a Flying Capacitors (FCMLC) con filtro di uscita, operanti sia in configurazione DC/DC che DC/AC. In particolare è stato affrontato il problema delle cadute improvvise di tensione in ingresso, scenario comune nei sistemi ad energia rinnovabile e di accumulo. In questi casi, mantenere stabile la tensione di uscita è cruciale per evitare cali di prestazione nei carichi collegati e garantire l’efficienza del sistema di conversione. Gli obiettivi principali di questo studio sono due: assicurare il bilanciamento delle tensioni nei flying capacitors, fondamentale per ridurre la distorsione delle forme d’onda e migliorare la stabilità del convertitore e mantenere la tensione di uscita desiderata nonostante le fluttuazioni dell’ingresso. A tal fine, sono stati valutati e confrontati tre metodi di controllo: Feedback Linearization, Sliding Mode Control e un controllo basato su algoritmo. La feedback linearization si è dimostrata precisa e robusta in entrambe le configurazioni; lo sliding mode control offre ottime prestazioni in applicazioni DC/DC, ma è meno efficace in DC/AC. Il controllo basato su algoritmo, infine, garantisce maggiore efficienza computazionale e versatilità, ma essendo in anello aperto è meno robusto rispetto ai disturbi. Poiché le cadute di tensione in ingresso sono imprevedibili e richiedono un lungo transitorio prima di stabilizzarsi, è stato sviluppato un metodo in grado di prevedere il loro valore finale fin dall’inizio della caduta. In questo modo, le tensioni dei flying capacitors seguono un riferimento a scalino e si stabilizzano rapidamente senza attendere le oscillazioni dell’ingresso. Questa tecnica ha un'accuratezza superiore al 99% e consente una regolazione in tempo reale che migliora la rapidità di risposta del convertitore. I risultati ottenuti contribuiscono all’analisi e allo sviluppo dei sistemi di elettronica di potenza per applicazioni energetiche.
Advanced control strategies for active voltage balancing and load voltage control in flying capacitor multilevel converters
Farinazzo, Eleonora
2023/2024
Abstract
This thesis investigates active control strategies for voltage balancing in a Flying Capacitor Multilevel Converter (FCMLC) with output filter operating under both DC/DC and DC/AC configurations. The study addresses a critical scenario often encountered in renewable energy and power storage systems: a sudden and unpredictable input voltage drop. In such cases, maintaining stable output voltage becomes essential, as voltage losses can compromise the performance of connected loads and the overall efficiency of the power conversion system. The primary objectives of this work are twofold: first, to ensure voltage balancing across the flying capacitors, crucial for minimizing waveform distortion and enhancing converter stability; and second, to maintain the desired output voltage level regardless of fluctuations in the DC input. To meet these goals, the thesis evaluates three control methods: Feedback Linearization, Sliding Mode Control, and an algorithm-based control, with a comparative analysis of the strengths and limitations of each approach. Feedback Linearization achieves high precision and robust closed-loop control across both configurations, Sliding Mode Control demonstrates strong robustness and accuracy in DC/DC applications but underperforms in DC/AC scenarios and Algorithm-Based Control provides computational efficiency and greater adaptability, but no robustness against disturbances. Managing capacitor voltage balancing is complex, as input voltage drops are unpredictable. For this reason in this thesis it has been developed a novel reference estimation technique that enables rapid capacitor voltage stabilization. Rather than waiting for DC voltage transients to end, this approach uses early-stage model identification to map the initial behavior of the DC voltage drop to its final stabilized value. This reference estimation method achieves over 99% accuracy, enhancing the converter’s response to sudden input variations by predicting and adjusting in real time. These findings contribute advancing the field of power electronics for dynamic energy systems.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Farinazzo_Executive Summary.pdf
non accessibile
Descrizione: Executive Summary
Dimensione
2.32 MB
Formato
Adobe PDF
|
2.32 MB | Adobe PDF | Visualizza/Apri |
2024_12_Farinazzo.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
7.89 MB
Formato
Adobe PDF
|
7.89 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/230600