Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks

Mehrdad Biglarbegian; Mohammadreza Baharani; Namwon Kim; Hamed Tabkhi; Babak Parkhideh

doi:10.1109/ECCE.2018.8557565

Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks

Mehrdad Biglarbegian, Mohammadreza Baharani, Namwon Kim, Hamed Tabkhi, Babak Parkhideh

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

11 Scopus citations

Abstract

Reliability and operation of high-frequency Gallium Nitride (GaN) power converters are yet to be discovered. Coming with the reliability assessment and improving the life extension of power converters, the approach is to monitor semiconductor on-resistor changes as a precursor signature for diagnostic/prognostic. This paper presents a novel approach for hybrid condition-based prognostic and reliability monitoring of GaN devices. The proposed approach offers a multi-physics co-simulations solution for degradation fatigue modeling of the GaN power devices. With the availability of the most granular information deduced from the advanced devices, the paper develops deep learning based algorithms for online reliability in power electronics. The proposed algorithm is based on the prominent version of Recurrent Neural Network (RNN) named Long Short-Term Memory (LSTM). LSTM models are utilized for system training and simulation model calibrations, and eventually predicting the next states within the next time horizon.

Original language	English
Title of host publication	2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	7271-7277
Number of pages	7
ISBN (Electronic)	9781479973118
DOIs	https://doi.org/10.1109/ECCE.2018.8557565
State	Published - Dec 3 2018
Externally published	Yes
Event	10th Annual IEEE Energy Conversion Congress and Exposition, ECCE 2018 - Portland, United States Duration: Sep 23 2018 → Sep 27 2018

Publication series

Name	2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018

Conference

Conference	10th Annual IEEE Energy Conversion Congress and Exposition, ECCE 2018
Country/Territory	United States
City	Portland
Period	09/23/18 → 09/27/18

Funding

This research was supported by the National Science Foundation under Award No. 1610250. The authors also would like to thank Dr. Mohammad Biglarbegian at the University of Guelph, Energy Production and Infrastructure Center (EPIC), and ECE Department at UNC-Charlotte.

Keywords

Fault diagnostic
GaN semiconductor
High frequency DC-DC converter
Long short-term memory
Machine learning
Recurrent neural network
Reliability

Access to Document

10.1109/ECCE.2018.8557565

Cite this

Biglarbegian, M., Baharani, M., Kim, N., Tabkhi, H., & Parkhideh, B. (2018). Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks. In 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018 (pp. 7271-7277). Article 8557565 (2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ECCE.2018.8557565

@inproceedings{13d0d981d8864175b95824b73f43533e,

title = "Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks",

abstract = "Reliability and operation of high-frequency Gallium Nitride (GaN) power converters are yet to be discovered. Coming with the reliability assessment and improving the life extension of power converters, the approach is to monitor semiconductor on-resistor changes as a precursor signature for diagnostic/prognostic. This paper presents a novel approach for hybrid condition-based prognostic and reliability monitoring of GaN devices. The proposed approach offers a multi-physics co-simulations solution for degradation fatigue modeling of the GaN power devices. With the availability of the most granular information deduced from the advanced devices, the paper develops deep learning based algorithms for online reliability in power electronics. The proposed algorithm is based on the prominent version of Recurrent Neural Network (RNN) named Long Short-Term Memory (LSTM). LSTM models are utilized for system training and simulation model calibrations, and eventually predicting the next states within the next time horizon.",

keywords = "Fault diagnostic, GaN semiconductor, High frequency DC-DC converter, Long short-term memory, Machine learning, Recurrent neural network, Reliability",

author = "Mehrdad Biglarbegian and Mohammadreza Baharani and Namwon Kim and Hamed Tabkhi and Babak Parkhideh",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 10th Annual IEEE Energy Conversion Congress and Exposition, ECCE 2018 ; Conference date: 23-09-2018 Through 27-09-2018",

year = "2018",

month = dec,

day = "3",

doi = "10.1109/ECCE.2018.8557565",

language = "English",

series = "2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "7271--7277",

booktitle = "2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018",

}

Biglarbegian, M, Baharani, M, Kim, N, Tabkhi, H & Parkhideh, B 2018, Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks. in 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018., 8557565, 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018, Institute of Electrical and Electronics Engineers Inc., pp. 7271-7277, 10th Annual IEEE Energy Conversion Congress and Exposition, ECCE 2018, Portland, United States, 09/23/18. https://doi.org/10.1109/ECCE.2018.8557565

Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks. / Biglarbegian, Mehrdad; Baharani, Mohammadreza; Kim, Namwon et al.
2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 7271-7277 8557565 (2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks

AU - Biglarbegian, Mehrdad

AU - Baharani, Mohammadreza

AU - Kim, Namwon

AU - Tabkhi, Hamed

AU - Parkhideh, Babak

PY - 2018/12/3

Y1 - 2018/12/3

N2 - Reliability and operation of high-frequency Gallium Nitride (GaN) power converters are yet to be discovered. Coming with the reliability assessment and improving the life extension of power converters, the approach is to monitor semiconductor on-resistor changes as a precursor signature for diagnostic/prognostic. This paper presents a novel approach for hybrid condition-based prognostic and reliability monitoring of GaN devices. The proposed approach offers a multi-physics co-simulations solution for degradation fatigue modeling of the GaN power devices. With the availability of the most granular information deduced from the advanced devices, the paper develops deep learning based algorithms for online reliability in power electronics. The proposed algorithm is based on the prominent version of Recurrent Neural Network (RNN) named Long Short-Term Memory (LSTM). LSTM models are utilized for system training and simulation model calibrations, and eventually predicting the next states within the next time horizon.

AB - Reliability and operation of high-frequency Gallium Nitride (GaN) power converters are yet to be discovered. Coming with the reliability assessment and improving the life extension of power converters, the approach is to monitor semiconductor on-resistor changes as a precursor signature for diagnostic/prognostic. This paper presents a novel approach for hybrid condition-based prognostic and reliability monitoring of GaN devices. The proposed approach offers a multi-physics co-simulations solution for degradation fatigue modeling of the GaN power devices. With the availability of the most granular information deduced from the advanced devices, the paper develops deep learning based algorithms for online reliability in power electronics. The proposed algorithm is based on the prominent version of Recurrent Neural Network (RNN) named Long Short-Term Memory (LSTM). LSTM models are utilized for system training and simulation model calibrations, and eventually predicting the next states within the next time horizon.

KW - Fault diagnostic

KW - GaN semiconductor

KW - High frequency DC-DC converter

KW - Long short-term memory

KW - Machine learning

KW - Recurrent neural network

KW - Reliability

UR - http://www.scopus.com/inward/record.url?scp=85060308181&partnerID=8YFLogxK

U2 - 10.1109/ECCE.2018.8557565

DO - 10.1109/ECCE.2018.8557565

M3 - Conference contribution

AN - SCOPUS:85060308181

T3 - 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018

SP - 7271

EP - 7277

BT - 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 10th Annual IEEE Energy Conversion Congress and Exposition, ECCE 2018

Y2 - 23 September 2018 through 27 September 2018

ER -

Biglarbegian M, Baharani M, Kim N, Tabkhi H, Parkhideh B. Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks. In 2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 7271-7277. 8557565. (2018 IEEE Energy Conversion Congress and Exposition, ECCE 2018). doi: 10.1109/ECCE.2018.8557565

Scalable Reliability Monitoring of GaN Power Converter Through Recurrent Neural Networks

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this