Automated Flexible Modeling for Various Full-SiC Power Modules

Various types of SiC power module product have been developed in recent years. Therefore, the device model developing speed and the flexibility corresponding to each product are further required. In this article, a novel high-accuracy data-driven modeling method was developed using 1.2-, 3.3-, and 6...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on power electronics Vol. 38; no. 5; pp. 6094 - 6107
Main Authors: Nakashima, Junichi, Horiguchi, Takeshi, Mukunoki, Yasushige, Hagiwara, Makoto, Urakabe, Takahiro, Harada, Shigeki
Format: Journal Article
Language:English
Published: New York IEEE 01.05.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
Automation
Data models
Diodes
Dynamic characteristics
Error reduction
gallium nitride
Genetic algorithms
HEMTs
Heuristic algorithms
Linear programming
Mathematical models
Model accuracy
Modelling
Modules
MOSFET
Optimization
optimization methods
Parameters
Semiconductor device modeling
silicon carbide
Simulated annealing
simulation program with integrated circuit emphasis (SPICE)
Static models
Waveforms
ISSN:0885-8993, 1941-0107
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Various types of SiC power module product have been developed in recent years. Therefore, the device model developing speed and the flexibility corresponding to each product are further required. In this article, a novel high-accuracy data-driven modeling method was developed using 1.2-, 3.3-, and 6.5-kV full-SiC power module measurement data. A static model was rapidly developed based on an automated approach using an optimization algorithm instead of the general model parameter extracting method by using a fixed model equation. Furthermore, the simulated annealing method was used to minimize the objective function that is calculated based on the difference in dynamic characteristics by optimizing adjustment parameters to reduce the error. The parameter optimization method focuses on not only I D and V DS but also I G and V GS waveforms. As a result, the relative errors of di/dt, dv/dt, and switching loss in the wide current region decreased considerably while maintaining the consistency of the I G and V GS waveforms. The modeling method can be used to realize numerous full-SiC power module models easily. The proposed algorithm provides fast, automated, and high-accuracy modeling. Furthermore, the optimization of the model can facilitate considerable expansion of the full-SiC power module usage.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0885-8993
1941-0107
DOI:10.1109/TPEL.2023.3239597