Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks

Mert Kesici; Can Berk Saner; Yusuf Yaslan; V. M. Istemihan Genc

doi:10.23919/ELECO47770.2019.8990370

Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks

Mert Kesici, Can Berk Saner, Yusuf Yaslan, V. M. Istemihan Genc

Department of Computer Engineering

Istanbul Technical University

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

2 Citations (Scopus)

Abstract

Transient instabilities triggered by critical faults can lead to rapidly developing blackouts in power systems. With the data driven methods using synchrophasor measurements collected from PMUs, it is possible to predict the instabilities just after the clearance of a critical fault. In this study, the performance of a classifier to be used for an early prediction of transient instability is enhanced by modifying the loss function used during the offline training phase. This study proposes to assign weights to the terms in the binary crossentropy loss function associated with each class, as misclassification of different classes generally causes different costs to the utility and its customers. The proposed method is able to determine the optimum values of the weights according to a pre-selected tolerance value, which represents how far the accuracy is away from being acceptable. The efficacy of the proposed method is examined on the 127-bus WSCC test system.

Original language	English
Title of host publication	ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	141-145
Number of pages	5
ISBN (Electronic)	9786050112757
DOIs	https://doi.org/10.23919/ELECO47770.2019.8990370
Publication status	Published - Nov 2019
Event	11th International Conference on Electrical and Electronics Engineering, ELECO 2019 - Bursa, Turkey Duration: 28 Nov 2019 → 30 Nov 2019

Publication series

Name	ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering

Conference

Conference	11th International Conference on Electrical and Electronics Engineering, ELECO 2019
Country/Territory	Turkey
City	Bursa
Period	28/11/19 → 30/11/19

Bibliographical note

Publisher Copyright:
© 2019 Chamber of Turkish Electrical Engineers.

Funding

This work was supported by the Scientific and Technological Research Council of Turkey (TUBITAK) under grant number 118E184.

Funders	Funder number
TUBITAK	118E184
Türkiye Bilimsel ve Teknolojik Araştirma Kurumu

Access to Document

10.23919/ELECO47770.2019.8990370

Cite this

Kesici, M., Saner, C. B., Yaslan, Y., & Istemihan Genc, V. M. (2019). Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks. In ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering (pp. 141-145). Article 8990370 (ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/ELECO47770.2019.8990370

Kesici, Mert ; Saner, Can Berk ; Yaslan, Yusuf et al. / Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks. ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 141-145 (ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering).

@inproceedings{c6e39ceeec2444a78c48c81b4c18d342,

title = "Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks",

abstract = "Transient instabilities triggered by critical faults can lead to rapidly developing blackouts in power systems. With the data driven methods using synchrophasor measurements collected from PMUs, it is possible to predict the instabilities just after the clearance of a critical fault. In this study, the performance of a classifier to be used for an early prediction of transient instability is enhanced by modifying the loss function used during the offline training phase. This study proposes to assign weights to the terms in the binary crossentropy loss function associated with each class, as misclassification of different classes generally causes different costs to the utility and its customers. The proposed method is able to determine the optimum values of the weights according to a pre-selected tolerance value, which represents how far the accuracy is away from being acceptable. The efficacy of the proposed method is examined on the 127-bus WSCC test system.",

author = "Mert Kesici and Saner, {Can Berk} and Yusuf Yaslan and {Istemihan Genc}, {V. M.}",

note = "Publisher Copyright: {\textcopyright} 2019 Chamber of Turkish Electrical Engineers.; 11th International Conference on Electrical and Electronics Engineering, ELECO 2019 ; Conference date: 28-11-2019 Through 30-11-2019",

year = "2019",

month = nov,

doi = "10.23919/ELECO47770.2019.8990370",

language = "English",

series = "ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering",

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

pages = "141--145",

booktitle = "ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering",

address = "United States",

}

Kesici, M, Saner, CB, Yaslan, Y & Istemihan Genc, VM 2019, Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks. in ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering., 8990370, ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering, Institute of Electrical and Electronics Engineers Inc., pp. 141-145, 11th International Conference on Electrical and Electronics Engineering, ELECO 2019, Bursa, Turkey, 28/11/19. https://doi.org/10.23919/ELECO47770.2019.8990370

Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks. / Kesici, Mert; Saner, Can Berk; Yaslan, Yusuf et al.
ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering. Institute of Electrical and Electronics Engineers Inc., 2019. p. 141-145 8990370 (ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering).

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

TY - GEN

T1 - Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks

AU - Kesici, Mert

AU - Saner, Can Berk

AU - Yaslan, Yusuf

AU - Istemihan Genc, V. M.

PY - 2019/11

Y1 - 2019/11

N2 - Transient instabilities triggered by critical faults can lead to rapidly developing blackouts in power systems. With the data driven methods using synchrophasor measurements collected from PMUs, it is possible to predict the instabilities just after the clearance of a critical fault. In this study, the performance of a classifier to be used for an early prediction of transient instability is enhanced by modifying the loss function used during the offline training phase. This study proposes to assign weights to the terms in the binary crossentropy loss function associated with each class, as misclassification of different classes generally causes different costs to the utility and its customers. The proposed method is able to determine the optimum values of the weights according to a pre-selected tolerance value, which represents how far the accuracy is away from being acceptable. The efficacy of the proposed method is examined on the 127-bus WSCC test system.

AB - Transient instabilities triggered by critical faults can lead to rapidly developing blackouts in power systems. With the data driven methods using synchrophasor measurements collected from PMUs, it is possible to predict the instabilities just after the clearance of a critical fault. In this study, the performance of a classifier to be used for an early prediction of transient instability is enhanced by modifying the loss function used during the offline training phase. This study proposes to assign weights to the terms in the binary crossentropy loss function associated with each class, as misclassification of different classes generally causes different costs to the utility and its customers. The proposed method is able to determine the optimum values of the weights according to a pre-selected tolerance value, which represents how far the accuracy is away from being acceptable. The efficacy of the proposed method is examined on the 127-bus WSCC test system.

UR - http://www.scopus.com/inward/record.url?scp=85080924130&partnerID=8YFLogxK

U2 - 10.23919/ELECO47770.2019.8990370

DO - 10.23919/ELECO47770.2019.8990370

M3 - Conference contribution

AN - SCOPUS:85080924130

T3 - ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering

SP - 141

EP - 145

BT - ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 11th International Conference on Electrical and Electronics Engineering, ELECO 2019

Y2 - 28 November 2019 through 30 November 2019

ER -

Kesici M, Saner CB, Yaslan Y, Istemihan Genc VM. Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks. In ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering. Institute of Electrical and Electronics Engineers Inc. 2019. p. 141-145. 8990370. (ELECO 2019 - 11th International Conference on Electrical and Electronics Engineering). doi: 10.23919/ELECO47770.2019.8990370

Cost Sensitive Class-Weighting Approach for Transient Instability Prediction Using Convolutional Neural Networks

Abstract

Publication series

Conference

Bibliographical note

Funding

Access to Document

Other files and links

Fingerprint

Cite this