An improved YOLOv7 for the state identification of sliding chairs in railway turnout

Abstract

The sliding chairs are important components that support the switch rail conversion in the railway turnout. Due to the harsh environmental erosion and the attack from the wheel vibration, the failure rate of the sliding chairs accounts for up to 10% of the total failure number in turnout. However, there is little research carried out in the existing literature to diagnose the deterioration states of the sliding chairs. To fill out this gap, by utilizing the images containing the sliding chairs, we propose an improved You Only Look Once version 7 (YOLOv7) to identify the state of the sliding chairs. Specifically, to meet the challenge brought by the small inter-class differences among the sliding chair states, we first integrate the Convolutional Block Attention Module (CBAM) into the YOLOv7 backbone to screen the information conducive to state identification. Then, an extra detector for a small object is customized into the YOLOv7 network in order to detect the small-scale sliding chairs in images. Meanwhile, we revise the localization loss in the objective function as the Efficient Intersection over Union (EIoU) to optimize the design of the aspect ratio, which helps the localization of the sliding chairs. Next, to address the issue caused by the varying scales of the sliding chairs, we employ K-means++ to optimize the priori selection of the initial anchor boxes. Finally, based on the images collected from real-world turnouts, the proposed method is verified and the results show that our method outperforms the basic YOLOv7 in the state identification of the sliding chairs with 4% improvements in terms of both mean Average Precision@0.5 (mAP@0.5) and F1-score.

Keywords

Railway turnout / Sliding chairs / State identification / Object detection

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This work was supported by the National Key R&D Program of China (2021YFF0501102), the National Natural Science Foundation of China (52372308, U2368202, U1934219, 52202392, 52022010, U22A2046, 52172322, and 62271486).

References

[1] W. Ji, Y. Zuo, R. Fei, et al., An adaptive fault diagnosis model for railway single and double action turnout. IEEE Trans. Intel. Transp. Syst. (2022).

[2] Y. Cao, Y. An, S. Su, et al., A statistical study of railway safety in China and Japan 1990–2020. Accid. Anal. Prev., 175 (2022) 106764.

[3] F. Wang, Y. Cao, L. Tan, et al., A reasoning diagram based method for fault diagnosis of railway point system. High.-Speed Railw., 1 (2023), pp. 110-119.

[4] Y. Cao, Y. Ji, Y. Sun, et al., The fault diagnosis of a switch machine based on deep random forest fusion. IEEE Intel. Transp. Syst. (2022).

[5] F. Wang, T. Tang, J. Yin, et al., A signal segmentation and feature fusion based RUL prediction method for railway point system, IEEE, Maui (2018), pp. 2303-2308.

[6] L.H. Zhao, Q. Lu. Method of turnout fault diagnosis based on grey correlation analysis. J. China Railw. Soc., 36(2)(2014), pp. 69-74.

[7] Y. Cao, Y. Sun, G. Xie, et al., A sound-based fault diagnosis method for railway point machines based on two-stage feature selection strategy and ensemble classifier. IEEE Trans. Intel. Transp. Syst., 23(8)(2021), pp. 12074-12083.

[8] F. Wang, M.R. Gahrooei, Z. Zhong, et al., An augmented regression model for tensors with missing values. IEEE Trans. Autom. Sci. Eng., 19(4)(2021), pp. 2968-2984.

[9] Y. Zhang, Q. Xie. Railway turnout detection method based on machine vision. Chin. Comput. Appl. Softw., 32 (1) (2015), pp. 225-228.

[10] F. Wang, A. Wang, T. Tang, et al., SGL-PCA: health index construction with sensor sparsity and temporal monotonicity for mixed high-dimensional signals. IEEE Trans. Autom. Sci. Eng., 20(1)(2023), pp. 372-384.

[11] F. Wang, Z. Zhong, G. Wang, et al., A penalized convolution model for oil leakage detection in electrohydraulic railway point systems. IEEE Trans. Instrum. Meas., 70 (2020), pp. 1-9.

[12] Y. Luo. Research on the switch slide baseplates under the conditions of dry friction. J. Shanghai Tiedao Univ., 19 (1998), pp. 37-42.

[13] X. Li, Y. Gao, J. Xing, et al., Wear reduction mechanism of graphite and MoS₂ in epoxy composites. Wear, 257(3-4)(2004), pp. 279-283.

[14] Y. Wang, Y. Gao, Y. Li, et al., Research on synergistic lubrication effect of silver modified Cu-Ni-graphite composite. Wear, 444 (2020) 203140.

[15] X. Cai, C. Li, P. Wang. Effect on the scant displacement of switch rail induced by friction of slide baseplate. China Railw. Sci., 28(1)(2007), pp. 8-12.

[16] R. Girshick, J. Donahue, T. Darrell, et al., Rich feature hierarchies for accurate object detection and semantic segmentation, CVPR’14 Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, ACM (2014) 580–587.

[17] R. Girshick. Fast R-CNN. IEEE International Conference on Computer Vision (ICCV), IEEE (2015), pp. 1440-1448.

[18] S. Ren, K. He, R. Girshick, et al., Faster R-CNN: towards real-time object detection with region proposal networks. Adv. Neural Inf. Proc. Syst. (2015), p. 28.

[19] W. Liu, D. Anguelov, D. Erhan, et al., SSD: Single shot multibox detector. Lect. Notes Comput. Sci., 9905 (2016), pp. 21-37.

[20] J. Redmon, S. Divvala, R. Girshick, et al., You only look once: unified, real-time object detection, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, (2016) 779–788.

[21] J. Redmon, A. Farhadi, YOLO9000: better, faster, stronger, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2017) 7263–7271.

[22] J. Redmon, A. Farhadi, YOLOv3: An incremental improvement, arXiv: 1804.02767, (2018).

[23] A. Bochkovskiy, C.-Y. Wang, H.-Y.M. Liao, YOLOv4: Optimal speed and accuracy of object detection, arXiv: 2004.10934, (2020).

[24] C. Li, L. Li, H. Jiang, et al., YOLOv6: A single-stage object detection framework for industrial applications, arXiv: 2209.02976, (2022).

[25] C.-Y. Wang, A. Bochkovskiy, H.-Y.M. Liao, et al., YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, arXiv: 2207.02696, (2022).

[26] Y. Cao, Z. Chen, T. Wen, et al., Rail fastener detection of heavy railway based on deep learning. High. Speed Railw., 1(1)(2022), pp. 63-69.

[27] F. Wang, A. Wang, T. Tang, et al., SGL-PCA: Health index construction with sensor sparsity and temporal monotonicity for mixed high-dimensional signals. IEEE Trans. Autom. Sci. Eng., 20(1)(2022), pp. 372-384.

[28] F. Wang, J. Du, Y. Zhao, et al., A deep learning based data fusion method for degradation modeling and prognostics. IEEE Trans. Reliab., 70(2)(2020), pp. 775-789.

[29] Z. Niu, G. Zhong, H. Yu. A review on the attention mechanism of deep learning. Neurocomputing, 452 (2021), pp. 48-62.

[30] G. Wang, X. Chen, F. Zhang, et al., Experience: understanding long-term evolving patterns of shared electric vehicle networks. MobiCom’19: The 25th Annual International Conference on Mobile Computing and Networking, ACM (2019) 1–12.

[31] S. Woo, J. Park, J-Y. Lee, et al., CBAM: convolutional block attention module. computer vision – ECCV 2018. Lecture Notes in Computer Science, Springer, Cham (2018), pp. 3-19.

[32] Z. Yang, X. Wang, J. Li. EIoU: an improved vehicle detection algorithm based on vehiclenet neural network. J. Phys.: Conf. Ser., 1924 (2021) 012001.

[33] J.A. Hartigan, M.A. Wong. Algorithm AS 136: a k-means clustering algorithm. J. R. Stat. Soc. Ser. C. (Appl. Stat. ), 28(1)(1979), pp. 100-108.

[34] D. Arthur, S. Vassilvitskii, K-means++: the advantages of careful seeding, SODA’07: Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms, Stanford, (2006) 1027–1035.