Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18

With the increasing proportion of distributed generation in power systems and the widespread adoption of new energy vehicles, more and more devices that utilize power electronic technology are being integrated into the power grid. This has led to a continuous rise in the level of supraharmonic emiss...

Full description

Saved in:
Bibliographic Details
Main Authors: Yangyang Zhu, Fei Zhong, Jiaqi Gao, Yuntai Cao, Xiaotian Wang, Zhihong Jiang
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:Results in Engineering
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2590123025003664
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850205066976296960
author Yangyang Zhu
Fei Zhong
Jiaqi Gao
Yuntai Cao
Xiaotian Wang
Zhihong Jiang
author_facet Yangyang Zhu
Fei Zhong
Jiaqi Gao
Yuntai Cao
Xiaotian Wang
Zhihong Jiang
author_sort Yangyang Zhu
collection DOAJ
description With the increasing proportion of distributed generation in power systems and the widespread adoption of new energy vehicles, more and more devices that utilize power electronic technology are being integrated into the power grid. This has led to a continuous rise in the level of supraharmonic emissions, and the supraharmonics associated with the integration of renewable energy into the grid have become one of the urgent power quality issues that need to be addressed in the power grid. The research analyzes why traditional methods for detecting supraharmonics cannot fundamentally eliminate spectral leakage. An improved ResNet18 transfer learning model is proposed to address the characteristics of supraharmonics in renewable energy grid integration. This model eliminates spectral leakage and provides a comparative analysis of convolutional neural networks and commonly used transfer learning detection models. The experimental results show that the improved IR_ResNet18 model significantly outperforms the original ResNet18 model and other models in terms of accuracy. The detection accuracy of the improved ResNet18 model is 3.2 % higher than that of the original ResNet18 model, and even more noticeably, its accuracy exceeds that of other models, reaching as high as 99.9 %. The improved ResNet18 model demonstrates greater reliability when processing new data, offering stronger generalization capability, better model stability, and enhanced robustness.
format Article
id doaj-art-89484afe1a9942a0a1e9db34fbc06b2d
institution OA Journals
issn 2590-1230
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series Results in Engineering
spelling doaj-art-89484afe1a9942a0a1e9db34fbc06b2d2025-08-20T02:11:11ZengElsevierResults in Engineering2590-12302025-03-012510428110.1016/j.rineng.2025.104281Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18Yangyang Zhu0Fei Zhong1Jiaqi Gao2Yuntai Cao3Xiaotian Wang4Zhihong Jiang5School of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaCorresponding author.; School of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaSchool of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaSchool of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaSchool of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaSchool of Electrical and Information Engineering, Changchun Institute of Technology, Changchun 130025, PR ChinaWith the increasing proportion of distributed generation in power systems and the widespread adoption of new energy vehicles, more and more devices that utilize power electronic technology are being integrated into the power grid. This has led to a continuous rise in the level of supraharmonic emissions, and the supraharmonics associated with the integration of renewable energy into the grid have become one of the urgent power quality issues that need to be addressed in the power grid. The research analyzes why traditional methods for detecting supraharmonics cannot fundamentally eliminate spectral leakage. An improved ResNet18 transfer learning model is proposed to address the characteristics of supraharmonics in renewable energy grid integration. This model eliminates spectral leakage and provides a comparative analysis of convolutional neural networks and commonly used transfer learning detection models. The experimental results show that the improved IR_ResNet18 model significantly outperforms the original ResNet18 model and other models in terms of accuracy. The detection accuracy of the improved ResNet18 model is 3.2 % higher than that of the original ResNet18 model, and even more noticeably, its accuracy exceeds that of other models, reaching as high as 99.9 %. The improved ResNet18 model demonstrates greater reliability when processing new data, offering stronger generalization capability, better model stability, and enhanced robustness.http://www.sciencedirect.com/science/article/pii/S2590123025003664SupraharmonicsResNet18Renewable energyGrid integrationTransfer learning
spellingShingle Yangyang Zhu
Fei Zhong
Jiaqi Gao
Yuntai Cao
Xiaotian Wang
Zhihong Jiang
Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
Results in Engineering
Supraharmonics
ResNet18
Renewable energy
Grid integration
Transfer learning
title Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
title_full Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
title_fullStr Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
title_full_unstemmed Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
title_short Research on Supraharmonic detection in renewable energy grid Inte-gration based on improved ResNet18
title_sort research on supraharmonic detection in renewable energy grid inte gration based on improved resnet18
topic Supraharmonics
ResNet18
Renewable energy
Grid integration
Transfer learning
url http://www.sciencedirect.com/science/article/pii/S2590123025003664
work_keys_str_mv AT yangyangzhu researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18
AT feizhong researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18
AT jiaqigao researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18
AT yuntaicao researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18
AT xiaotianwang researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18
AT zhihongjiang researchonsupraharmonicdetectioninrenewableenergygridintegrationbasedonimprovedresnet18