Procedure for combining ﬁeld measurements and machine learning to quantify impact of different track parameters on ride quality of railway tracks

Abstract

Ride quality in terms of vibration is a fundamental factor affecting passengers’ satisfaction. Every year, passenger carriers invest significantly in various aspects of their system, including track and infrastructure, to improve ride quality. The assessment of ride quality and understanding the extent of the impact of different parameters on its magnitude is essential for railway operators to make informed decisions regarding capital expenditures. This paper presents a methodology for using machine learning techniques to find the correlation between various parameters (including train speed, weather conditions, presence of track features, and composition of the track substructure) and ride quality (quantified using measurements from accelerometers mounted on a rail car). The statistical model was developed using field measurements collected over a 50 km section of VIA Rail’s track in Canada. This paper describes the collected field data, the development of the statistical model, and discusses the importance of each parameter on the accuracy of the model.

Keywords

Ride quality train speed track features machine learning statistical model regression analysis

Get full access to this article

View all access options for this article.

References

Davis

Boundy

. Transportation energy data book. Edition 37.2. Oak Ridge, TN: Oak Ridge National Laboratory, 2019.

Railway Association of Canada. Rail trends. 2018.

Amdor-Jimenez

Mohammadi

Nasiri

Level of comfort and safety in railway transit. In: 4th international conference on transportation information and safety, Banff, Canada, 8–10 August 2017.

Transport Canada. 2012. Rail transportation, www.tc.gc.ca/eng/policy/anre-menu-3020.htm (2012, accessed 19 December 2019).

Roghani

Hendry

MT.

Quantifying the impact of subgrade stiffness on track quality and the development of geometry defects. J Transp Eng, Part A: Systems 2017; 143: 04017029.

Hyslip

Fractal analysis of railway track geometry. Transport Res Record 2002; 1785: 50–57.

Selig

Substruture maintenance management. In: International heavy haul conference, Cape Town, South Africa, 6–10 April 2019.

Transport Canada. Rules respecting track safety. Ottawa: Transport Canada, 2011.

Zhang

Sun

, et al. Track irregularities estimation based on the vibration of car-body. In: IEEE 2nd advanced information technology, electronic and automation control conference. Chongqing, 25–26 March 2017, pp.1369–1372. Piscataway: IEEE.

10.

Lee

Choi

Kim

, et al. A mixed filtering approach for track condition monitoring using accelerometers on the axle box and bogie. IEEE Trans Instrument Measure 2012; 61: 749–758.

11.

Kawasaki

, and and Youcef-Toumi

Estimation of rail irregularities. In: Proceedings of the 2002 American control conference 2002, Anchorage, AK, USA, 8–10 May 2002. pp.360–366. Piscataway: IEEE.

12.

Toma

Liu

Jahagirdar

, et al. Track monitoring using ride quality measurements. In: Proceedings of the 2019 world congress of rail research 2019, Tokyo, Japan, 28 October–1 November 2019.

13.

ISO 2631-11997. Mechanical Vibration and shock evaluation of human exposure to whole body vibration – Part 1 general requirements.

14.

Roghani

Hiedra-Cobo

Charbachi

Studying the impact of freeze thaw cycles on performance of railway tracks. In: International heavy haul conference, 2019, Narvik, Norway, 11–14 June 2019, pp.897–902.

15.

Roghani

Kashani

Wiljanen

BR.

Using ground penetrating radar to evaluate freeze thaw condition of railway substructure. In: Transportation research board conference, 2020, Washington, DC, 12–16 January 2020.

16.

Silvast

Nurmikolu

Wiljanen

, et al. Identifying frost susceptible areas on Finnish railways using the ground penetrating radar technique. J Rail Rapid Transit 2013; 227: 3–9.

17.

Pylkkanen

Nurmikolu

Causes, effects and control of seasonal frost action in railways. Sci Cold Arid Regions 2011; 5: 363–367.

18.

Hyslip

Olhoeft

Smith

, et al. Ground penetrating radar for railroad track substructure evaluation. Washington, DC: Federal Railroad Administration Office of Research and Development, US Dept. of Transportation, 2005.

19.

Ebersohn

Selig

Use of track geometry measurements for maintenance planning. Transportation Research Record No. 1470, 1994.

20.

Kleinbaum

Kupper

Muller

, et al. Applied regression analysis and other multivariable methods. Belmont, CA: Duxbury Press, 1988.

21.

Russell

SJ.

Norvig

P. Artificial intelligence: a modern approach. 3rd ed. Upper Saddle River: Prentice Hall, 2010.

22.

James

Witten

Hastie

, et al. An introduction to statistical learning: with applications in R. New York: Springer, 2017, pp.303–336.

23.

Hastie

Tibshirani

Friedman

JH.

The elements of statistical learning. 2nd ed. New York: Springer, 2009, pp.337–384.

24.

Chen

Guestrin

Xgboost: a scalable tree boosting system. In: Proceedings of the 22nd international conference on knowledge discovery and data mining, San Francisco, CA, 13–17 August 2016, pp.785–794. New York: ACM.

25.

xgboost developers. 2019. XGBoost Documentation – xgboost 1.0.0-SNAPSHOT documentation, https://xgboost.readthedocs.io/en/latest/ (2019, accessed 7 January 2020).