The phenomenon of hunting motion in the process of high-speed train operation will greatly reduce the ride comfort, and may even cause serious safety accidents such as train derailment. According to the image characteristics of high-speed train wheel-rail contact, a train hunting motion monitoring method based on You Only Look Once (YOLO) neural network and machine vision is proposed. The method can detect the high-speed train hunting motion state and parameters, and send an alarm signal when the relative wheel-rail displacement exceeds the threshold. The wheel-rail contact image during high-speed operation is obtained by installing a camera system at the bottom of the high-speed train bogie. A detection model of wheel-rail contact parameters is then established and the key position of wheel-rail contact is identified. Combined with the image processing algorithms including the Canny edge detection and Hough linear transform methods, the relative wheel-rail displacement during the process of high-speed train hunting motion is calculated. The accuracy and reliability of the model is validated by field measured data. The results show that the proposed method can accurately detect the high-speed train hunting motion under normal illumination, overexposure and dark conditions. The image processing speed is up to 60 frames per second, which can meet the needs of real-time detection in the high-speed railway application environment.
WangCChenJLiB, et al.Impact of high-speed railway construction on spatial patterns of regional economic development along the route: a case study of the Shanghai–Kunming high-speed railway. Soc Econ Plann Sci2023; 87: 101583.
2.
WangQLiDZengJ, et al.A diagnostic method of freight wagons hunting performance based on wayside hunting detection system. Measurement2024; 227: 114274.
3.
ZhaiW. Vehicle–track coupled dynamics models. Springer, 2020, pp. 17–149.
4.
WanTHTsangCWHuiK, et al.Anomaly detection of train wheels utilizing short-time fourier transform and unsupervised learning algorithms. Eng Appl Artif Intell2023; 122: 106037.
5.
OukhellouLDebiollesADenœuxT, et al.Fault diagnosis in railway track circuits using Dempster–Shafer classifier fusion. Eng Appl Artif Intell2010; 23(1): 117–128.
6.
Peinado GonzaloAHorridgeRSteeleH, et al.Review of data analytics for condition monitoring of railway track geometry. IEEE Trans Intell Transport Syst2022; 23(12): 22737–22754.
7.
GhiasiRKhanMASorrentinoD, et al.An unsupervised anomaly detection framework for onboard monitoring of railway track geometrical defects using one-class support vector machine. Eng Appl Artif Intell2024; 133: 108167.
8.
ZhaiGXuYSpencerBF. Bidirectional graphics-based digital twin framework for quantifying seismic damage of structures using deep learning networks. Struct Health Monit2025; 24(1): 86–110.
9.
ZhaiGSpencerBFYanJ, et al.Coupled data/physics-driven framework for accurate and efficient structural response simulation. Eng Struct2025; 327: 119636.
10.
PengLZhengSLiP, et al.A comprehensive detection system for track geometry using fused vision and inertia. IEEE Trans Instrum Meas2020; 70: 1–15.
11.
ZhengDLiLZhengS, et al.A defect detection method for rail surface and fasteners based on deep convolutional neural network. Comput Intell Neurosci2021; 2021: 2565500–2565515.
12.
WeiDWeiXTangQ, et al.RTLSeg: a novel multi-component inspection network for railway track line based on instance segmentation. Eng Appl Artif Intell2023; 119: 105822.
13.
PiresACVianaMCAScaramussaLM, et al.Measuring vertical track irregularities from instrumented heavy haul railway vehicle data using machine learning. Eng Appl Artif Intell2024; 127: 107191.
14.
RopalkarOSYadavOPZambareC, et al.Improved estimation of rail–wheel contact forces from instrumented wheel-set data through higher harmonic cancellation and a back-propagation neural network scheme. Eng Appl Artif Intell2023; 119: 105811.
15.
SunJMeliECaiW, et al.A signal analysis based hunting instability detection methodology for high-speed railway vehicles. Veh Syst Dyn2021; 59(10): 1461–1483.
16.
FangSYangLTangJ, et al.Visual measurement of lateral relative displacement of wheel-rail of high-speed train under earthquake. Eng Struct2024; 305: 117736.
17.
WangJLingLDingXet al. The influence of aerodynamic loads on carbody low-frequency hunting of high-speed trains. International Journal of Structural Stability and Dynamics. 2022; 22(13): 2250145.
18.
ZhangWZengJLiY. A review of vehicle system dynamics in the development of high-speed trains in China. Int J Dyn Control2013; 1: 81–97.
19.
ChongSYLeeJRShinHJ. A review of health and operation monitoring technologies for trains. Smart Struct Syst2010; 6(9): 1079–1105.
20.
WuPZhangFWangJ, et al.Review of wheel-rail forces measuring technology for railway vehicles. Adv Mech Eng2023; 15(3): 16878132231158991.
21.
SunJMeliESongX, et al.A novel measuring system for high-speed railway vehicles hunting monitoring able to predict wheelset motion and wheel/rail contact characteristics. Veh Syst Dyn2023; 61(6): 1621–1643.
22.
PengXZengJWangJ, et al.Wayside wheelset lateral motion detection and vehicle hunting instability evaluation. Meas Sci Technol2023; 34(12): 125904.
23.
KulkarniRQazizadehABergM, et al.Vehicle running instability detection algorithm (VRIDA): a signal based onboard diagnostic method for detecting hunting instability of rail vehicles. Proc Inst Mech Eng F J Rail Rapid Transit2022; 236(3): 262–274.
24.
YeYGaoHHuangC, et al.Computer vision for hunting stability inspection of high-speed trains. Measurement2023; 220: 113361.
25.
YangT. Research on the wheel-rail contact relationship of trains based on image processing. Lanzhou Jiaotong University, 2020.
26.
MaZWangYSongZ. Image detection of relative lateral displacement between locomotive wheels and rails based on laser source. Journal of Graphics2017; 38(04): 623–628.
27.
LiangJSunJJiangY, et al.Advances and challenges in the hunting instability diagnosis of high-speed trains. Sensors2024; 24(17): 5719.
28.
ShiDŠabanovičERizzettoL, et al.Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications. Mech Syst Signal Process2022; 166: 108482.
29.
ShahidiPMarainiDHopkinsB, et al.Estimation of bogie performance criteria through on-board condition monitoring. Annual Conference of the PHM Society2014; 6(1): 648–657.
BochkovskiyAWangCYLiaoHYM (2020). Yolov4: optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934.
32.
DongXYanSDuanC. A lightweight vehicles detection network model based on YOLOv5. Eng Appl Artif Intell2022; 113: 104914.
33.
MinguellMGPanditR. TrackSafe: a comparative study of data-driven techniques for automated railway track fault detection using image datasets. Eng Appl Artif Intell2023; 125: 106622.
34.
TervenJCordova-EsparzaD (2023). A comprehensive review of YOLO architectures in computer vision: from YOLOv1 to YOLOv8 and YOLO-NAS. arXiv preprint arXiv:2304.00501.
35.
YuanZFangWZhaoY, et al.Research of insect recognition based on improved yolov5. J Artif Intell2021; 3(4): 145–152.
36.
AhmadTMaYYahyaM, et al.Object detection through modified YOLO neural network. Sci Program2020; 2020: 1–10.
37.
JiangPErguDLiuF, et al.A review of yolo algorithm developments. Procedia computer science2022; 199: 1066–1073.
38.
HorvatMGledecG (2022, September). A comparative study of YOLOv5 models performance for image localization and classification. In 33rd Central European Conference on Information and Intelligent Systems (CECIIS) (p. 349), Dubrovnik, Croatia, 2022.
39.
ThuanD.Evolution of yolo algorithm and Yolov5: the state-of-the-art object detention algorithm. Finland: MA: Oulu University of Applied Sciences, 2021.
40.
FoodyGMMcCullochMBYatesWB. The effect of training set size and composition on artificial neural network classification. Int J Rem Sens1995; 16(9): 1707–1723.
41.
ShuklaKNPotnisADwivedyP. A review on image enhancement techniques. Int J Eng Appl Comput. Sci2017; 2(07): 232–235.
42.
MathewMPMaheshTY. Leaf-based disease detection in bell pepper plant using YOLO v5. Signal Image Video Process2022; 16: 841–847.
43.
SargentDJ. Comparison of artificial neural networks with other statistical approaches: results from medical data sets. Cancer2001; 91(S8): 1636–1642.
44.
NixonMAguadoA. Feature extraction and image processing for computer vision. Academic Press, 2019.
45.
MaQDMaZJiC, et al.Artificial object edge detection based on enhanced Canny algorithm for high-speed railway apparatus identification. In: 2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI). Shanghai, China, 2017, October, pp. 1–6.
46.
MohammadpourMBahroudiAAbediM. Automatic lineament extraction method in mineral exploration using CANNY Algorithm and hough transform. Geotecton2020; 54(3): 366–382.
47.
LiHWangYZengJ, et al.Virtual point tracking method for online detection of relative wheel-rail displacement of railway vehicles. Reliab Eng Syst Saf2024; 246: 110087.
48.
ClarkeTAFryerJG. The development of camera calibration methods and models. Photogramm Rec1998; 16(91): 51–66.
49.
DuaneCB. Close-range camera calibration. Photogramm Eng1971; 37(8): 855–866.
50.
ZhangZ. A flexible new technique for camera calibration. IEEE Trans Pattern Anal Mach Intell2000; 22(11): 1330–1334.
51.
DaiXQuSHuangC, et al.The determination of limit wheel profile for hunting instability of railway vehicles using stacking feature deep forest. Eng Appl Artif Intell2023; 125: 106732.
52.
ShevtsovIYMarkineVLEsveldC. Optimal design of wheel profile for railway vehicles. Wear2005; 258(7-8): 1022–1030.
53.
LingLZhaoJZhangH, et al.Adaptive robust constraint-following control method for improving carbody hunting stability of high-speed trains, Veh Syst Dyn. 2024. Online ahead of print.
