Formulation of Rectifier Numerical Average-Value Model for Direct Interface With Inductive Circuitry

The computational cost for the simulation of detailed models of machine-rectifier systems is expensive because of repetitive diodes switching. Average-value models (AVMs) of machine-rectifier systems have been developed that can alleviate the computational burden by neglecting the details of the swi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on energy conversion 2019-06, Vol.34 (2), p.741-749
Hauptverfasser: Zhang, YuQi, Cramer, Aaron M.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The computational cost for the simulation of detailed models of machine-rectifier systems is expensive because of repetitive diodes switching. Average-value models (AVMs) of machine-rectifier systems have been developed that can alleviate the computational burden by neglecting the details of the switching of each individual diode while retaining the average characteristics. This paper proposes an alternative formulation of numerical AVMs of machine-rectifier systems, which makes direct use of the natural dynamic impedance of the rectifier without introducing low-frequency approximations or algebraic loops. By using this formulation, direct interface of the AVM is achieved with inductive circuitry on both the ac and dc sides allowing traditional voltage-in, current-out formulations of the circuitry on these sides to be used with the proposed formulation directly. This numerical AVM formulation is validated against an experimentally validated detailed model and compared with previous AVM formulations. It is demonstrated that the proposed AVM formulation accurately predicts the system's low-frequency behavior during both steady and transient states, including the cases where previous AVM formulations cannot predict accurate results. Both run times and numbers of time steps needed by the proposed AVM formulation are comparable to those of existing AVM formulations and significantly decreased compared with the detailed model.
ISSN:0885-8969
1558-0059
DOI:10.1109/TEC.2018.2874546