Unsupervised machine learning methodologies for identification of transversal imbalanced loads in freight railway vehicles

Abstract

Imbalanced loads in freight railway vehicles pose significant risks to vehicle running safety as well as track integrity, increasing the likelihood of derailments and increasing track wear rate. This study presents a robust machine learning-based methodology designed to detect and classify transverse imbalances in freight vehicles using dynamic rail responses. The proposed approach employs wayside monitoring systems with accelerometers and strain gauges, integrating advanced feature extraction methods, including principal component analysis, log-mel spectrograms, and multi-feature-based techniques. The methodology enhances detection accuracy by normalizing features to eliminate environmental and operational variations and employing data fusion for sensitive index creation. It is capable of distinguishing between different severity levels of imbalanced loads across various wagon types. By simulating scenarios with typical European freight wagons, the study demonstrates the effectiveness of the approach, offering a valuable tool for railway infrastructure managers to mitigate risks associated with imbalanced loads. This research contributes to the field by providing a scalable, non-invasive solution for real-time monitoring and safety enhancement in freight rail operations.

Keywords

Freight traffic loads / Imbalanced vertical loads / Wayside condition monitoring / Train–track interaction / Artificial intelligence

References

[1] Bernal E, Spiryagin M, Cole C. Onboard condition monitoring sensors, systems and techniques for freight railway vehicles: a review. IEEE Sens J, 2019, 19(1): 4-24

[2] Chong SY, Lee J-R, Shin H-J. A review of health and operation monitoring technologies for trains. Smart Struct Syst, 2010, 6(9): 1079-1105

[3] Montenegro PA, Carvalho H, Ribeiro D, et al.. Assessment of train running safety on bridges: a literature review. Eng Struct, 2021, 241: 112425

[4] Alves V, Meixedo A, Ribeiro D, et al.. Evaluation of the performance of different damage indicators in railway bridges. Procedia Eng, 2015, 114: 746-753

[5] D-RAIL (2012) Summary report and database of derailments incidents

[6] Ulianov C, Defossez F, Vasić Franklin G, et al.. Yannis G, Cohen S, et al.. Overview of freight train derailments in the EU: causes, impacts, prevention and mitigation measures. Traffic safety, 2016, Hoboken, Wiley

[7] European Railway Agency (2012) Prevention and mitigation of freight train derailments at short and medium terms (ERA/REP/02-2012/SAF)

[8] UIC (2022) Loading Guidelines Code of practice for the loading and securing of goods on railway wagons Volume 1 Principles

[9] Transportation Technology Center Inc (2010) General Rules Covering Loading of Carload Shipments of Commodities in Closed Cars

[10] Railway C. General rules covering loading and reinforcement in railway freight wagons, 2006, Beijing, China Railway Press

[11] Cheng LR, Harrison H, Union G et al (2007) Effects of improper loading in heavy haul operations. In: IHHA Specialist Technical Session (STS), Kiruna, Sweden

[12] Liu X, Rapik Saat M, Barkan CPL. Freight-train derailment rates for railroad safety and risk analysis. Accid Anal Prev, 2017, 98: 1-9

[13] Zhang D, Tang Y, Sun Z, et al.. Optimising the location of wagon gravity centre to improve the curving performance. Veh Syst Dyn, 2022, 60(5): 1627-1641

[14] Zhang D, Tang Y, Peng Q, et al.. Effect of mass distribution on curving performance for a loaded wagon. Nonlinear Dyn, 2021, 104(3): 2259-2273

[15] Mosleh A, Costa PA, Calçada R. A new strategy to estimate static loads for the dynamic weighing in motion of railway vehicles. Proc Inst Mech Eng Part F J Rail Rapid Transit, 2020, 234(2): 183-200

[16] Pintão B, Mosleh A, Vale C, et al.. Development and validation of a weigh-in-motion methodology for railway tracks. Sensors, 2022, 22(5): 1976

[17] Su J, Wang X, Chen R et al (2012) Design and research on monitoring system of overload and unbalanced load of freight cars. In: World Automation Congress 2012, Puerto Vallarta, Mexico June 24–28, 201, IEEE, pp 1–4

[18] Maraini D, Shahidi P, Hopkins BM et al (2014) Development of a bogie-mounted vehicle on-board weighing system. In: Proceedings of the 2014 Joint Rail Conference. Colorado Springs, Colorado, USA, April 2–4, 2014

[19] Lan C, Yang Z, Liang X, et al.. Experimental study on wayside monitoring method of train dynamic load based on strain of ballastless track slab. Constr Build Mater, 2023, 394: 132084

[20] Cortis D, Giulianelli S, Malavasi G, et al.. Self-diagnosis method for checking the wayside systems for wheel-rail vertical load measurement. Transp Probl, 2018, 12(491-100

[21] Allotta B, Gaburri G, Innocenti A et al (2014) An innovative high speed weigh in motion system for railway vehicles. Paper presented at 2014 IEEE/ASME 10th international conference on mechatronic and embedded systems and applications (MESA), Senigallia, Italy, September 10–12, 2014. IEEE, pp1–7

[22] Zhou W, Abdulhakeem S, Fang C, et al.. A new wayside method for measuring and evaluating wheel-rail contact forces and positions. Measurement, 2020, 166: 108244

[23] Ding Y, Cheng L (2018) High speed overload and imbalance load system in china. In: Proceedings of the 2018 Joint Rail Conference. Pittsburgh, PA, USA, April 18–20, 2018

[24] Pan J, Li W, Dai X (2013) Train overload and unbalanced load detection based on FBG gauge. In: Proceedings of the SPIE Fourth Asia Pacific Optical Sensors Conference, Wuhan, China

[25] Chang Q, Han M (2011) Study on a combined method of Butterworth high-pass filter and Canny edge detector in the application of detecting cargo loading position on railway vehicles. In: Proceedings of 2011 international conference on computer science and network technology, Harbin, December 24–26, 2011. IEEE, pp 1959–1963

[26] Mohammadi M, Mosleh A, Vale C, et al.. An unsupervised learning approach for wayside train wheel flat detection. Sensors, 2023, 23(41910

[27] Mosleh A, Meixedo A, Ribeiro D, et al.. Automatic clustering-based approach for train wheels condition monitoring. Int J Rail Transp, 2023, 11(5639-664

[28] Guedes A, Silva R, Ribeiro D, et al.. Clustering-based classification of polygonal wheels in a railway freight vehicle using a wayside system. Appl Sci, 2024, 14(93650

[29] Silva R, Guedes A, Ribeiro D, et al.. Early identification of unbalanced freight traffic loads based on wayside monitoring and artificial intelligence. Sensors, 2023, 23(31544

[30] Belotti V, Crenna F, Michelini RC, et al.. Wheel-flat diagnostic tool via wavelet transform. Mech Syst Signal Process, 2006, 20(81953-1966

[31] Mosleh A, Meixedo A, Ribeiro D, et al.. Early wheel flat detection: an automatic data-driven wavelet-based approach for railways. Veh Syst Dyn, 2023, 61(6): 1644-1673

[32] Jorge T, Magalhães J, Silva R, et al.. Early identification of out-of-roundness damage wheels in railway freight vehicles using a wayside system and a stacked sparse autoencoder. Veh Syst Dyn, 2025, 63(2): 232-257

[33] Magalhães J, Jorge T, Silva R, et al.. A strategy for out-of-roundness damage wheels identification in railway vehicles based on sparse autoencoders. Railw Eng Sci, 2024, 32(4): 421-443

[34] Zhang Q, Peng J, Tian K, et al.. Advancing ultrasonic defect detection in high-speed wheels via UT-YOLO. Sensors, 2024, 24(5): 1555

[35] Trilla A, Bob-Manuel J, Lamoureux B, Vilasis-Cardona X. Integrated multiple-defect detection and evaluation of rail wheel tread images using convolutional neural networks. Int J Progn Health Manag, 2021

[36] Xing Z, Zhang Z, Yao X, et al.. Rail wheel tread defect detection using improved YOLOv3. Measurement, 2022, 203: 111959

[37] Anagra I, Bahiuddin I, Priatomo HR, et al.. Detection of coal wagonload distributions based on geometrical features using extreme learning machine methods. Int J Inf Technol, 2024, 16(2939-947

[38] de Souza EF, Bragança C, Ribeiro D, et al.. Drive-by damage detection methodology for high-speed railway bridges using sparse autoencoders. Railw Eng Sci, 2024

[39] Meixedo A, Ribeiro D, Calçada R et al (2014) Global and local dynamic effects on a railway viaduct with precast deck. In: Civil-Comp Proceedings, Ajaccio, Corsica, France, April 8–11, 2014. Civil-Comp Press, pp 77

[40] Ghorbani H. Mahalanobis distance and its application for detecting multivariate outliers. Ser Math Inform, 2019, 34(3583-595

[41] Dervilis N, Antoniadou I, Barthorpe RJ, et al.. Robust methods for outlier detection and regression for SHM applications. Int J Sustain Mater Struct Syst, 2015, 2(1–2): 3

[42] Neves SGM, Montenegro PA, Azevedo AFM, et al.. A direct method for analyzing the nonlinear vehicle–structure interaction. Eng Struct, 2014, 69: 83-89

[43] Montenegro PA, Neves SGM, Calçada R, et al.. Wheel–rail contact formulation for analyzing the lateral train–structure dynamic interaction. Comput Struct, 2015, 152: 200-214

[44] UIC (2007) Code 517—Wagons—Suspension gear—Standardisation

[45] Silva R, Ribeiro D, Bragança C, et al.. Model updating of a freight wagon based on dynamic tests under different loading scenarios. Appl Sci, 2021, 11(2210691

[46] ANSYS Inc (2018) ANSYS^® Academic Research Mechanical Release 18.1

[47] Bragança C, Neto J, Pinto N, et al.. Calibration and validation of a freight wagon dynamic model in operating conditions based on limited experimental data. Veh Syst Dyn, 2022, 60(93024-3050

[48] European Committee for Standardization (2006) EN 13674-1: Railway applications—Track—Rail—Part 1: Vignole railway rails 46 kg/m and above

[49] Zhai W, Wang K, Cai C. Fundamentals of vehicle–track coupled dynamics. Veh Syst Dyn, 2009, 47(11): 1349-1376

[50] ERRI D 214 (1999) RP5—Rail bridges for speeds > 200 km/h: Numerical investigation of the effect of track irregularities at bridge resonance

[51] ERRI D 202 (1999) RP11—improved knowledge of forces in CWR track (including switches): Parametric study and sensitivity analysis of CWERRI Verification of an FEM for the buckling resistance of tracks with continuous welded rails (CWR)

[52] Wu Y-S, Yang Y-B. Steady-state response and riding comfort of trains moving over a series of simply supported bridges. Eng Struct, 2003, 25(2): 251-265

[53] UIC (2001) Code 774-3-R Track/Bridge interaction—recommendations for calculation

[54] Neto J, Montenegro PA, Vale C, et al.. Evaluation of the train running safety under crosswinds—a numerical study on the influence of the wind speed and orientation considering the normative Chinese Hat Model. Int J Rail Transp, 2021, 9(3): 204-231

[55] European Committee for Standardization (2020) EN 13848-2 Railway applications—Track— Track geometry quality— Part 2: Measuring systems—Track recording vehicles

[56] Hertz H (1882) Ueber die Berührung fester elastischer Körper (On contact between elastic bodies). In: Crelle A, Borchardt C, Schellbach (eds) Band 92. Berlin, Boston: De Gruyter, pp 156–171

[57] Kalker JJ (1996) Book of tables for the Hertzian creep-force law. Faculty of Technical Mathematics and Informatics, Delft University of Technology

[58] The MathWorks Inc (2020) MATLAB (R2020a). Natick, Massachusetts

Funding

Fundação de Amparo à Pesquisa do Estado de São Paulo(2022/13045-1)

Institute of Research and Development in Structures and Construction(UIDB/04708/2020)

RIGHTS & PERMISSIONS

The Author(s)