Transformer fault diagnosis based on improved deep belief networks

Ming Xv; Zheng Cao; Jiawei Li

doi:10.1117/12.3010399

20 October 2023 Transformer fault diagnosis based on improved deep belief networks

Ming Xv, Zheng Cao, Jiawei Li

Proceedings Volume 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023); 1281422 (2023) https://doi.org/10.1117/12.3010399
Event: Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 2023, Chongqing, China

Abstract

Traditional transformer fault diagnosis methods have low accuracy and cannot be accurately classified. In order to improve the accuracy of transformer fault diagnosis, a SMOTE-RF-IHPO-DBN transformer fault diagnosis model is established. Firstly, the synthetic minority oversampling technique is used to balance the data. Then the random forest is used to reduce the dimensionality of the data to extract important information in the data and improve efficiency. Next, IHPO is used to optimize DBN, and finally use DBN to diagnose and classify fault types. The simulation results show that the accuracy of the proposed method is improved by 13.4%, 7.08%, 4.14% compared with the DBN, HPO-DBN and IHPO-DBN, respectively, which proves the effectiveness of the model.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Ming Xv, Zheng Cao, and Jiawei Li "Transformer fault diagnosis based on improved deep belief networks", Proc. SPIE 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 1281422 (20 October 2023); https://doi.org/10.1117/12.3010399

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available