Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network
An arc fault is the leading cause of electrical fire. Aiming at the problems of difficulty in manually extracting features, poor generalization ability of models and low prediction accuracy in traditional arc fault detection algorithms, this paper proposes a fault arc detection method based on the f...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/16/13/4954 |
| _version_ | 1797591818003546112 |
|---|---|
| author | Xiang Gao Gan Zhou Jian Zhang Ying Zeng Yanjun Feng Yuyuan Liu |
| author_facet | Xiang Gao Gan Zhou Jian Zhang Ying Zeng Yanjun Feng Yuyuan Liu |
| author_sort | Xiang Gao |
| collection | DOAJ |
| description | An arc fault is the leading cause of electrical fire. Aiming at the problems of difficulty in manually extracting features, poor generalization ability of models and low prediction accuracy in traditional arc fault detection algorithms, this paper proposes a fault arc detection method based on the fusion of channel attention mechanism and residual network model. This method is based on the channel attention mechanism to perform global average pooling of information from each channel of the feature map assigned by the residual block while ignoring the local spatial data to enhance the detection and recognition rate of the fault arc. This paper introduces a one-dimensional depth separable convolution (1D-DS) module to reduce the network model parameters and shorten the time of single prediction samples. The experimental results show that the F1 score of the network model for arc fault detection under mixed load conditions is 98.07%, and the parameter amount is reduced by 46.06%. The method proposed in this paper dramatically reduces the parameter quantity, floating-point number and time complexity of the network structure while ensuring a high recognition rate, which improves the real-time response ability to detect arc fault. It has a guiding significance for applying arc fault on the edge side. |
| first_indexed | 2024-03-11T01:42:46Z |
| format | Article |
| id | doaj.art-2e56dc4cbd864844911b1df89c228be3 |
| institution | Directory Open Access Journal |
| issn | 1996-1073 |
| language | English |
| last_indexed | 2024-03-11T01:42:46Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj.art-2e56dc4cbd864844911b1df89c228be32023-11-18T16:28:14ZengMDPI AGEnergies1996-10732023-06-011613495410.3390/en16134954Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual NetworkXiang Gao0Gan Zhou1Jian Zhang2Ying Zeng3Yanjun Feng4Yuyuan Liu5School of Software Engineering, Southeast University, Nanjing 211189, ChinaSchool of Electrical Engineering, Southeast University, Nanjing 211189, ChinaState Grid Guangdong Electric Power Company, Guangzhou 510600, ChinaState Grid Guangdong Electric Power Company, Guangzhou 510600, ChinaSchool of Electrical Engineering, Southeast University, Nanjing 211189, ChinaSchool of Electrical Engineering, Southeast University, Nanjing 211189, ChinaAn arc fault is the leading cause of electrical fire. Aiming at the problems of difficulty in manually extracting features, poor generalization ability of models and low prediction accuracy in traditional arc fault detection algorithms, this paper proposes a fault arc detection method based on the fusion of channel attention mechanism and residual network model. This method is based on the channel attention mechanism to perform global average pooling of information from each channel of the feature map assigned by the residual block while ignoring the local spatial data to enhance the detection and recognition rate of the fault arc. This paper introduces a one-dimensional depth separable convolution (1D-DS) module to reduce the network model parameters and shorten the time of single prediction samples. The experimental results show that the F1 score of the network model for arc fault detection under mixed load conditions is 98.07%, and the parameter amount is reduced by 46.06%. The method proposed in this paper dramatically reduces the parameter quantity, floating-point number and time complexity of the network structure while ensuring a high recognition rate, which improves the real-time response ability to detect arc fault. It has a guiding significance for applying arc fault on the edge side.https://www.mdpi.com/1996-1073/16/13/4954arc fault detectionconvolutional neural networkresidual networkchannel attention mechanismone-dimensional depth separable convolution |
| spellingShingle | Xiang Gao Gan Zhou Jian Zhang Ying Zeng Yanjun Feng Yuyuan Liu Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network Energies arc fault detection convolutional neural network residual network channel attention mechanism one-dimensional depth separable convolution |
| title | Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network |
| title_full | Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network |
| title_fullStr | Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network |
| title_full_unstemmed | Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network |
| title_short | Fault Arc Detection Based on Channel Attention Mechanism and Lightweight Residual Network |
| title_sort | fault arc detection based on channel attention mechanism and lightweight residual network |
| topic | arc fault detection convolutional neural network residual network channel attention mechanism one-dimensional depth separable convolution |
| url | https://www.mdpi.com/1996-1073/16/13/4954 |
| work_keys_str_mv | AT xianggao faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork AT ganzhou faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork AT jianzhang faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork AT yingzeng faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork AT yanjunfeng faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork AT yuyuanliu faultarcdetectionbasedonchannelattentionmechanismandlightweightresidualnetwork |