Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln
It is difficult to identify the working conditions of the rotary kilns due to the harsh environment in the kilns. The flame images of the firing zone in the kilns contain a lot of working condition information, but the flame image data sample size is too small to be used to fully extract the key fea...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-08-01
|
Series: | Entropy |
Subjects: | |
Online Access: | https://www.mdpi.com/1099-4300/24/9/1186 |
_version_ | 1797488707755835392 |
---|---|
author | Yuchao Hu Weihua Zheng Xin Wang Bin Qin |
author_facet | Yuchao Hu Weihua Zheng Xin Wang Bin Qin |
author_sort | Yuchao Hu |
collection | DOAJ |
description | It is difficult to identify the working conditions of the rotary kilns due to the harsh environment in the kilns. The flame images of the firing zone in the kilns contain a lot of working condition information, but the flame image data sample size is too small to be used to fully extract the key features. In order to solve this problem, a method combining transfer learning and attention mechanism is proposed to extract key features of flame images, in which the deep residual network is used as the backbone network, the coordinate attention module is introduced to capture the position information and channel information on the branch of feature graphs, and the features of flame images obtained are further screened to improve the extraction ability. At the same time, migration learning is performed by the pre-trained ImageNet data set, and feature migration and parameter sharing are realized to cope with the training difficulty of a small sample data size. Moreover, an activation function Mish is introduced to reduce the loss of effective information. The experimental results show that, compared with traditional methods, the working condition recognition accuracy of rotary kilns is improved by about 5% with the proposed method. |
first_indexed | 2024-03-10T00:06:10Z |
format | Article |
id | doaj.art-52693a5c07c34b41bbd95ecf8cf19e68 |
institution | Directory Open Access Journal |
issn | 1099-4300 |
language | English |
last_indexed | 2024-03-10T00:06:10Z |
publishDate | 2022-08-01 |
publisher | MDPI AG |
record_format | Article |
series | Entropy |
spelling | doaj.art-52693a5c07c34b41bbd95ecf8cf19e682023-11-23T16:07:25ZengMDPI AGEntropy1099-43002022-08-01249118610.3390/e24091186Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary KilnYuchao Hu0Weihua Zheng1Xin Wang2Bin Qin3School of Electrical & Information Engineering, Hunan University of Technology, Zhuzhou 412007, ChinaSchool of Electrical & Information Engineering, Hunan University of Technology, Zhuzhou 412007, ChinaSchool of Electrical & Information Engineering, Hunan University of Technology, Zhuzhou 412007, ChinaSchool of Electrical & Information Engineering, Hunan University of Technology, Zhuzhou 412007, ChinaIt is difficult to identify the working conditions of the rotary kilns due to the harsh environment in the kilns. The flame images of the firing zone in the kilns contain a lot of working condition information, but the flame image data sample size is too small to be used to fully extract the key features. In order to solve this problem, a method combining transfer learning and attention mechanism is proposed to extract key features of flame images, in which the deep residual network is used as the backbone network, the coordinate attention module is introduced to capture the position information and channel information on the branch of feature graphs, and the features of flame images obtained are further screened to improve the extraction ability. At the same time, migration learning is performed by the pre-trained ImageNet data set, and feature migration and parameter sharing are realized to cope with the training difficulty of a small sample data size. Moreover, an activation function Mish is introduced to reduce the loss of effective information. The experimental results show that, compared with traditional methods, the working condition recognition accuracy of rotary kilns is improved by about 5% with the proposed method.https://www.mdpi.com/1099-4300/24/9/1186rotary kilnflame imageworking condition recognitiondeep learningtransfer learningcoordinate attention mechanism |
spellingShingle | Yuchao Hu Weihua Zheng Xin Wang Bin Qin Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln Entropy rotary kiln flame image working condition recognition deep learning transfer learning coordinate attention mechanism |
title | Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln |
title_full | Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln |
title_fullStr | Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln |
title_full_unstemmed | Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln |
title_short | Working Condition Recognition Based on Transfer Learning and Attention Mechanism for a Rotary Kiln |
title_sort | working condition recognition based on transfer learning and attention mechanism for a rotary kiln |
topic | rotary kiln flame image working condition recognition deep learning transfer learning coordinate attention mechanism |
url | https://www.mdpi.com/1099-4300/24/9/1186 |
work_keys_str_mv | AT yuchaohu workingconditionrecognitionbasedontransferlearningandattentionmechanismforarotarykiln AT weihuazheng workingconditionrecognitionbasedontransferlearningandattentionmechanismforarotarykiln AT xinwang workingconditionrecognitionbasedontransferlearningandattentionmechanismforarotarykiln AT binqin workingconditionrecognitionbasedontransferlearningandattentionmechanismforarotarykiln |