基于GoogLeNet Inception V3模型的火电用钢金相组织智能识别

doi:10.11896/cldb.23050138

Abstract
Figure/Table
References
Related Citation (15)

Download:

Full Text ( PDF ) (12699KB)
Export: BibTeX | EndNote (RIS)

Abstract The steel used in thermal power units is mostly alloy steel, which have complex metallographic structures and have been used in high-temperature and high-pressure environments for a long time, resulting in structural aging and other phenomena. The manual recognition of metallographic organization is easily influenced by subjective factors of the testing personnel, with large fluctuations in recognition accuracy and poor repeatability of results. This study utilized metallographic organization images to establish a dataset of metallographic organization for thermal power steel based on relevant standards, and divided it into training, validation, and testing sets. Based on the GoogLeNet Inception V3 model, three models of original, transfer learning and fine tuning transfer learning were established, and the training set training model was used. The model with the best generalization ability on the verification set was taken as the final model. The test set was combined with the confusion matrix to comprehensively evaluate the performance of the built model. The results show that the recognition accuracy of the three models for the metallographic structure of steel used in thermal power is 96.6%, 93.0%, and 92.4%, respectively, which can accurately identify complex metallographic structures. The original Inception V3 model has the best performance, with an average of 96.6%, 96.6%, 96.6%, and 99.2% for accuracy, positive predictive value, true positive rate, and true negative rate in the test set. This study provides a new method for intelligent recognition of complex metallographic structures.

Key words： microstructure intelligent recognition deep learning convolutional neural network computational material science

Published: 10 September 2024 Online: 2024-09-30

CLC number:

TG142.1+5

Fund:Inner Mongolia Power Research Institute Branch Self Funded Technology Project in 2022(2022-ZC-05) and the National Natural Science Foundation of China (52061037).

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	ZHANG Yanfei
	ZHANG Yongzhi
	GONG Weiwei
	LIU Xiao
	WEI Zhigang

Cite this article:

ZHANG Yanfei,ZHANG Yongzhi,GONG Weiwei, et al. Intelligent Recognition of Microstructure of Steel in Thermal Power Unit Based on GoogLeNet Inception V3 Model[J]. Materials Reports, 2024, 38(17): 23050138-7.

URL:

http://www.mater-rep.com/EN/10.11896/cldb.23050138 OR http://www.mater-rep.com/EN/Y2024/V38/I17/23050138

1 Decost B L, Holm E A. Computational Materials Science, 2015, 110, 126.
2 Chowdhury A, Kautz E, Yener B, et al. Computational Materials Science, 2016, 123, 176.
3 Naik D L, Sajid H U, Kiran R. Metals, 2019, 9(5), 546.
4 Tsutsui K, Terasaki H, Uto K, et al. Materials Today Communications, 2020, 25, 1.
5 Wu W H, Lee J C, Wang Y M. Sensors, 2020, 20(19), 1.
6 Yang X, Shi C, Cao H, et al. American Chemical Society Omega, 2023, 8(6), 5300.
7 Kusche C, Reclik T, Freund M, et al. Plos One, 2019, 14(5), 1.
8 Hu F K. Microstructure identification of high strength steel materials based on texture recognition and deep convolution neural network. Master's Thesis, Huazhong University of Science & Technology, China, 2019 (in Chinese).
胡方康. 基于纹理识别和深度卷积神经网络的高强钢材料显微组织识别. 硕士学位论文, 华中科技大学, 2019.
9 He H Z. Automatic identification of steel metallographic structure based on deep learning. Master's Thesis, Wuhan Institute of Technology, China, 2022 (in Chinese).
何惠珍. 基于深度学习的钢材金相组织自动化识别. 硕士学位论文, 武汉工程大学, 2022.
10 Zhang Y Z, Xin Q Z, Kong X M, et al. Hot Working Technology, 2021, 50(12), 68 (in Chinese).
张永志, 辛全忠, 孔祥明, 等. 热加工工艺, 2021, 50(12), 68.
11 Zhang Y Z, Li X Y, Xin Q Z, et al. Materials Reports, 2022, 36(12), 160 (in Chinese).
张永志, 李旭英, 辛全忠等. 材料导报, 2022, 36(12), 160.
12 Dimiduk D M, Holm E A, Niezgoda S R. Integrating Materials and Manufacturing Innovation, 2018, 7(3), 157.
13 Decost B L, Francis T, Holm E A. Acta Materialia, 2017, 133, 30.
14 Adachi Y, Taguchi M, Hirokawa S. Tetsu-to-Hagane, 2016, 102(12), 722.
15 Krizhevsky A, Sutskever I, Hinton G E, et al. In:NIPS'12:Proceedings of the 25th International Conference on Neural Information Processing Systems. Nevada, 2012, pp.1097.
16 Simonyan K, Zisserman A. Computer Science, DOI:10.48550/arXiv.1409.1556.
17 Szegedy C, Liu W, Jia Y, et al. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, 2015, pp.1.
18 Ioffe S, Szegedy C. In:International Conference on Machine Learning. Lille, 2015, pp.448.
19 Szegedy C, Vanhoucke V, Ioffe S, et al. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, 2016, pp.2818.
20 Shin H C, Roth H R, Gao M, et al. IEEE Transactions on Medical Imaging, 2016, 35(5), 1285.
21 Zhang Y Z, Xin Q Z, Wang Y L, et al. Materials Reports, 2021, 35(24), 152 (in Chinese).
张永志, 辛全忠, 王永亮, 等. 材料导报, 2021, 35(24), 152.
22 Foody G M. Remote Sensing of Environment, 2002, 80(1), 185.