改进YOLOv5s的地铁齿轮箱螺丝状态检测

doi:10.3778/j.issn.1002-8331.2406-0111

摘要/Abstract

摘要： 随着城市轨道交通的快速发展，地铁维护的安全性和效率变得尤为重要。其中，车底齿轮箱螺丝的状态直接关系到城轨车辆的安全运行，但传统的基于YOLOv5s检测方法在实时性方面存在明显不足。针对以上问题，提出一种基于改进YOLOv5s的齿轮箱螺丝状态检测算法。为保证模型实时性，采用了分组混洗卷积模块，显著减少了模型的计算量；针对传统空间金字塔池化技术未能充分考虑特征间差异性的问题，引入特征自适应融合的空间金字塔池化技术，提升了模型的性能；为提升单个卷积的表达能力，使用了一种新的特征提取单元，在保证轻量化的同时能够获取更多特征信息；为提高模型检测微小螺丝目标的能力，采用Focal Loss损失函数，提高了检测的准确率。实验结果表明，改进后的模型在mAP精度上比原始YOLOv5s算法提高了2.2个百分点，同时模型的参数和计算量分别降低2.9[×]106和8.1 GFLOPs，检测速度低至25 ms，提升了22%，证明了该方法有效性。

关键词: 齿轮箱, 螺丝检测, YOLOv5s, 特征融合, 小目标检测

Abstract: With the rapid development of urban rail transit, the safety and efficiency of subway maintenance have become particularly important. The condition of gearbox screws under the train is directly related to the safe operation of urban rail vehicles. However, traditional detection methods based on YOLOv5s have significant deficiencies in real-time performance. To address these issues, this paper proposes an improved YOLOv5s-based algorithm for detecting the condition of gearbox screws. Firstly, to ensure the real-time performance of the model, a group shuffle convolution module is used, which significantly reduces the calculation of the model. Secondly, to address the issue that traditional spatial pyramid pooling technology fails to fully consider the differences between features, a spatial pyramid pooling technique with feature adaptive fusion is proposed, enhancing the model’s performance. Next, to improve the representative ability of a single convolution, a new feature extraction unit is employed, which can acquire more feature information while ensuring lightweight design. Finally, to enhance the model’s ability to detect small screw targets, the Focal Loss function is used, improving detection accuracy. Experimental results show that the improved model increases mAP accuracy by 2.2 percen-tage points compared to the original YOLOv5s algorithm, while the parameters and computational load are reduced by 2.9[×]106 and 8.1 GFLOPs respectively, and the detection speed decreases to 25 ms, with a 22 percent improvement, demonstrating the effectiveness of this method.

Key words: gearbox, screw detection, YOLOv5s, feature fusion, small target detection

程光耀, 张文强, 王志城, 华路捷, 严世昌, 张涛. 改进YOLOv5s的地铁齿轮箱螺丝状态检测[J]. 计算机工程与应用, 2025, 61(20): 358-367.

CHENG Guangyao, ZHANG Wenqiang, WANG Zhicheng, HUA Lujie, YAN Shichang, ZHANG Tao. Screw Status Detection of Subway Gearbox Based on Improved YOLOv5s[J]. Computer Engineering and Applications, 2025, 61(20): 358-367.

参考文献

[1] 张建超, 陈湘松, 郝如江, 等. 高速动车齿轮箱故障诊断技术研究综述[J]. 机械传动, 2023, 47(3): 133-140.
ZHANG J C, CHEN X S, HAO R J, et al. Review of fault diag-nosis technology for gearboxes of high-speed EMU[J]. Journal of Mechanical Transmission, 2023, 47(3): 133-140.
[2] 陈再刚, 刘禹清, 周子伟, 等. 轨道交通牵引动力传动系统动力学研究综述[J]. 交通运输工程学报, 2021, 21(6): 31-49.
CHEN Z G, LIU Y Q, ZHOU Z W, et al. Review of dynamics research on traction power transmission system of rail transit[J]. Journal of Traffic and Transportation Engineering, 2021, 21(6): 31-49.
[3] 叶宏鹏. 基于图像处理的动车轴端螺丝检测方法的研究[D]. 北京: 北京交通大学, 2018.
YE H P. Research on detection method of EMU axle-end bolts based on image processing[D]. Beijing: Beijing Jiaotong University, 2018.
[4] 路绳方. 复杂场景下动车底部螺栓丢失故障的自动检测[J]. 激光与光电子学进展, 2017, 54(11): 297-303.
LU S F. Automatic detection of bolt loss fault at the bottom of EMU in complex scenarios[J]. Progress in Laser and Optoelectronics, 2017, 54(11): 297-303.
[5] 费垚东, 李柏林, 范宏. TFDS中螺丝故障的自动识别算法研究[J]. 机械设计与制造, 2021(10): 97-101.
FEI Y D, LI B L, FAN H. Research on automatic identific-ation algorithm of bolt fault in TFDS[J]. Machinery Design & Manufacture, 2021(10): 97-101.
[6] DALAL N, TRIGGS B. Histograms of oriented gradientsfor human detection[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005: 886-893.
[7] PENG J, XUE H, WEI Z, et al. Integrating multi-network topology for gene function prediction using deep neural networks[J]. Brief Bioinform, 2021, 22(2): 2096-2105.
[8] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[9] REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 7263-7271.
[10] ZHANG J, LIU Y, LIU H. Distractor-aware visual tracking using hierarchical correlation filters adaptive selection[J]. Applied Intelligence, 2022, 52(6): 6129-6147.
[11] ZHANG J, JINX, SUN J, et al. Dual model learning combined with multiple feature selection for accurate visual trac-king[J]. IEEE Access, 2019, 7: 43956-43969.
[12] XIN N, KE G, LI W, et al. Feature refinement and filter network for person re-identification[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2020, 31(9): 3391-3402.
[13] 赵江平, 徐恒, 党悦悦. 基于改进Faster R-CNN的铁路客车螺丝检测研究[J]. 中国安全科学学报, 2021, 31(7): 82-89.
ZHAO J P, XU H, DANG Y Y. Research on screw detection of railway passenger cars based on improved faster R-CNN[J]. China Safety Science Journal, 2021, 31(7): 82-89.
[14] GIRSHICK R. Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision, 2015.
[15] 雷威, 杜森, 张皓泽, 等. 一种基于图像处理的列车车底螺丝丢失检测方法: CN202210067611.2[P]. 2022-05-07.
LEI W, DU S, ZHANG H Z, et al. A train undercarriage bolt loss detection method based on image processing: CN2022-10067611.2[P]. 2022-05-07.
[16] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proceedings of the European Conference on Computer Vision, 2016: 21-37.
[17] 邹一鸣, 李鹏, 林群煦, 等. 基于改进YOLOv5算法的地铁车辆转向架螺丝缺失检测[J]. 机械工程师, 2023(9): 50-54.
ZOU Y M, LI P, LIN Q X, et al. Detection of missing bolts in metro vehicle bogies based on improved YOLOv5 algorithm[J]. Mechanical Engineer, 2023(9): 50-54.
[18] Ultralytics. YOLOv5[EB/OL]. (2022-04-10)[2023-07-18]. https://github.com/ultralytics/yolov5.
[19] 牛鑫宇, 毛鹏军, 段云涛, 等. 基于YOLOv5s室内目标检测轻量化改进算法研究[J]. 计算机工程与应用, 2024, 60(3): 109-118.
NIU X Y, MAO P J, DUAN Y T, et al. Research on lightweight improved algorithm for indoor target detection based on YOLOv5s[J]. Computer Engineering and Applications, 2024, 60(3): 109-118.
[20] 何雨, 田军委, 张震, 等. YOLOv5目标检测的轻量化研究[J]. 计算机工程与应用, 2023, 59(1): 92-99.
HE Y, TIAN J W, ZHANG Z, et al. Lightweight research of YOLOv5 target detection[J]. Computer Engineering and Applications, 2023, 59(1): 92-99.
[21] SUNANDINI G, SIVANPILLAI R, SOWNYA V, et al. Significance of atrous spatial pyramid pooling (ASPP) in Deeplabv3+for water body segmentation[C]//Proceedings of the 10th International Conference on Signal Processing and Integrated Networks, 2023: 744-749.
[22] QIU Y, LIU Y, CHEN Y, et al. A2SPPNet: attentive atrous spatial pyramid pooling network for salient object detection[J]. IEEE Transactions on Multimedia, 2023, 25: 1991-2006.
[23] JI J, LI S, LIAO X, et al. Semantic segmentation based on spatial pyramid pooling and multilayer feature fusion[J]. IEEE Transactions on Cognitive and Developmental Systems, 2023, 15(3): 1524-1535.
[24] TANI Y, YAMAKKKKAWA S, YAKUSHIJIN S, et al. Sema-ntic segmentation of spine and femur bone using atrous spatial pyramid pooling-based U-Net with fully connected CRF[C]//Proceedings of the IEEE International Conference on Systems, Man, and Cybernetics, 2023: 4967-4972.
[25] CHUNLONG Z, PEILE H, SHENGHUA W, et al. Pavement defect detection algorithm based on improved YOLOv7 complex background[J]. IEEE Access, 2024, 12: 32870-32880.
[26] 郭迎, 梁睿琳, 王润民. 基于CNN图像增强的雾天跨域自适应目标检测[J]. 计算机工程与应用, 2023, 59(16): 187-195.
GUO Y, LIANG R L, WANG R M. Cross-domain adaptive object detection based on CNN image enhancement in foggy conditions[J]. Computer Engineering and Applications, 2023, 59(16): 187-195.
[27] HU J, SHEN L, SUN G, et al. Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 7132-7141.
[28] WOO S, PARK J, LEE J, et al. CBAM: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision, 2018: 3-19.
[29] HOU Q, ZHOU D, FENG J. Coordinate attention for efficient mobile network design[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogn-ition, 2021: 13713-13722.
[30] DING X, ZHANG X, HAN J, et al. Diverse branch block: building a convolution as an inception-like unit[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 10886-10895.
[31] REN F, FEI J, LI H, et al. Steel surface defect detection using improved deep learning algorithm: ECA-SimSPPF-SIoU-YOLOv5[J]. IEEE Access, 2024, 12: 32545-32553.
[32] LI C. YOLOv6 v3.0: a full-scale reloading[J]. arXiv:2301. 05586, 2023.
[33] ZHAO X. A quality grade classification method for fresh tea leaves based on an improved YOLOv8x-SPPCSPC-CBAM model[J]. Scientific Reports, 2024, 14(1): 4166.
[34] 刘展威, 陈慈发, 董方敏. 基于YOLOv5s的航拍小目标检测改进算法研究[J]. 无线电工程, 2023, 53(10): 2286-2294.
LIU Z W, CHEN C F, DONG F M. Improved aerial small object detection algorithm based on YOLOv5s[J]. Radio Engineering, 2023, 53(10): 2286-2294.
[35] WANG C. Gold-YOLO: efficient object detector via gather-and-distribute mechanism[J]. arXiv:2309.11331, 2023.
[36] 赵楚, 段先华, 苏俊楷. 改进Faster RCNN的瓷砖表面瑕疵检测研究[J]. 计算机工程与应用, 2023, 59(14): 201-208.
ZHAO C, DUAN X H, SU J K. Research on ceramic tile surface defect detection by improved faster RCNN[J]. Computer Engineering and Applications, 2023, 59(14): 201-208.
[37] 刘艳萍, 刘甜. 改进的Cascade RCNN行人检测算法研究[J]. 计算机工程与应用, 2022, 58(4): 229-236.
LIU Y P, LIU T. Improved Cascade RCNN pedestrian dete-ction algorithm research[J]. Computer Engineering and Applications, 2022, 58(4): 229-236.
[38] 龚陈博, 南卓江, 陶卫. 基于改进Faster RCNN的PCB表面缺陷检测研究[J]. 自动化仪表, 2024, 45 (7): 99-103.
GONG C B, NAN Z J, TAO W. Research on PCB surface defect detection based on improved faster RCNN [J]. Automation Instrument, 2024, 45 (7): 99-103.