Railway track irregularities are key factors that affect the safety of trains and the comfort of passengers. In this study, a railway track is separated into consecutive track maintenance units, which are used as research objects. We propose a novel framework, which is referred to as the tree-augmented naïve Bayes-track quality index (TAN-TQI), to identify possible underlying patterns or rules for predicting track irregularities based on the characteristic deterioration of track maintenance units. The prediction framework is validated using track irregularity data measured using track geometry cars. Our evaluation shows that the prediction performance of the TAN-TQI framework is better than that of the state-of-the-art method, the short-range prediction method for track quality index.
KawaguchiAMiwaMTeradaK. Actual data analysis of alignment irregularity growth and its prediction model. Q R RTRI2005; 46(4): 262–268.
2.
Veit P and Marschnig S. Sustainability in track: a precondition for high speed traffic. In: Proceedings of the 2010 joint rail conference, Urbana, IL, 27–29 April 2010, pp. 349–355. New York, NY: ASME.
3.
QuirogaLMSchniederEA heuristic approach to railway track maintenance scheduling. In: NingBBrebbiaCA (eds). Computers in railways XII: computer system design and operation in railways and other transit systems, Southampton, UK: Wit Press, 31 August – 2 September 2010, pp. 687–699.
4.
Jovanovic S. Railway track quality assessment and related decision making. In: The 2004 IEEE international conference on systems, man and cybernetics, volume 6, The Hague, The Netherlands, 10–13 October 2004, pp.5038–5043. Piscataway, NJ: IEEE Press.
5.
LiuRXuPWangF. Research on a short-range prediction model for track irregularity over small track lengths. J Transport Engng2010; 36: 1085–1091.
6.
XuPSunQLiuRWangF. A short-range prediction model for track quality index. Proc IMechE, Part F: J Rail Rapid Transit2011; 225: 277–285.
7.
XuPSunQLiuRWangF. Key equipment identification model for correcting milepost errors of track geometry data from track inspection cars. Transport Res, Part C: Emerg Technol2013; 35: 85–103.
8.
Meier-HirmerCSeneeARibouletGA decision support system for track maintenance. In: AllanJBrebbiaCARumseyAF (eds). Computers in railways X: computer system design and operation in the railway and other transit systems, Southampton, UK: Computational Mechanics Publications Ltd, 2006, pp. 217–226.
9.
Meier-HirmerCRibouletGSourgetFRoussignolM. Maintenance optimization for a system with a gamma deterioration process and intervention delay: application to track maintenance. Proc IMechE, Part O: J Risk Reliab2009; 223: 189–198.
10.
LeeTCJudgeGGZellnerA. Estimating the parameters of the Markov probability model from aggregate time series data, Amsterdam, The Netherlands: North-Holland, 1970.
11.
GolabiKShepardR. Pontis: A system for maintenance optimization and improvement of US bridge networks. Interfaces1997; 27: 71–88.
12.
Mishalani RG and Madanat SM. Infrastructure state transition probability computation using duration models. In: KCP Wang, S Madanat, S Nambisan and Spring G (eds) The seventh international conference on applications of advanced technologies in transportation. Boston, MA, 5–7 August 2002, pp.505–512. Reston, VA: ASCE Press.
13.
MishalaniRGMadanatSM. Computation of infrastructure transition probabilities using stochastic duration models. J Infrastruct Syst2002; 8: 139–148.
KobayashiKDoMHanD. Estimation of Markovian transition probabilities for pavement deterioration forecasting. KSCE J Civil Engng2010; 14: 343–351.
16.
KobayashiKKaitoKLethanhN. A statistical deterioration forecasting method using hidden Markov model for infrastructure management. Transport Res, Part B: Methodol2012; 46: 544–561.
17.
Bai L, Liu R, Sun Q, et al. Markov-based model for prediction of railway track irregularities. Proc IMechE, Part F: J Rail Rapid Transit. Epub ahead of print 18 September 2013. DOI: 10.1177/0954409713503460.
18.
Durango-CohenPL. A time series analysis framework for transportation infrastructure management. Transport Res, Part B: Methodol2007; 41: 493–505.
19.
WittenIHFrankEHallMA. Data mining: practical machine learning tools and techniques, Burlington, Massachusetts, USA: Morgan Kaufmann, 2011.
20.
BarratSTabboneS. A Bayesian network for combining descriptors: application to symbol recognition. Int J Doc Anal Recog2010; 13: 65–75.
21.
CanoJPerez-CortesJArlandisJTraining set expansion in handwritten character recognition. In: CaelliTAminADuinRW (eds). Structural, syntactic, and statistical pattern recognition, Heidelberg, Germany: Springer, 2002, pp. 548–556.
22.
De OliveiraLSCAndreãoRVSarcinelli-FilhoMThe use of Bayesian networks for heart beat classification. In: Amnir HussainIALS (ed.). Brain inspired cognitive systems 2008, Heidelberg, Berlin: Springer, 2010, pp. 217–231.
23.
FerreiraDOliveiraAFreitasA. Applying data mining techniques to improve diagnosis in neonatal jaundice. BMC Med Inform Decisions2012; 12: 143–143.
24.
MengshoelOJChaviraMCascioK. Probabilistic model-based diagnosis: an electrical power system case study. IEEE Trans Syst Man Cybern, A: Syst Hum2010; 40: 874–885.
25.
EN 13848-1: 2008. Railway applications - track - track geometry quality - part 1 characterisation of track geometry.
26.
The Railway Ministry of the People’s Republic of China. Rules of railway track maintenance, Beijing, People’s Republic of China: China Railway Publishing House, 2010.
27.
LuoLZhangGWuWChaiX. The control of track regularity state in wheel-rail system, Beijing, People’s Republic of China: China Railway Publishing House, 2006.
28.
ZhangW. The principles and applications of GJ-4 track geometry cars, Beijing, People’s Republic of China: China Railway Publishing House, 2001.
29.
EN 13848-5: 2008. Railway applications - track - track geometry quality - part 5 geometric quality levels.
30.
The Railway Ministry of People’s Republic of China. Track quality index management interim standards of existing railway, Beijing, People’s Republic of China: China Railway Publishing House, 2009.
31.
DudaROHartPE. Pattern classification and scene analysis, New York, NY: Wiley, 1973.
32.
Domingos P and Pazzani M. Beyond independence: conditions for the optimality of the simple Bayesian classifier. In: Saitta L (ed.) Proceedings of the 13th international conference machine learning, Bari, Italy, 3–6 July 1996, pp.105–112. Burlington, Massachusetts, USA: Morgan Kaufmann.
33.
PazzaniMJSearching for dependencies in Bayesian classifiers. In: FisherDLenzH (eds). Learning from data, Heidelberg, Germany: Springer, 1996, pp. 239–248.
34.
Keogh E and Pazzani M. Learning augmented Bayesian classifiers: a comparison of distribution-based and classification-based approaches. In: Heckerman D and Whittaker J (eds) Proceedings of the seventh international workshop on artificial intelligence and statistics, Fort Lauderdale, Florida, USA, 4–6 January 1999, pp.225–230. Burlington, Massachusetts, USA: Morgan Kaufmann.
35.
HallMFrankEHolmesG. The WEKA data mining software: an update. ACM SIGKDD Explorations Newsletter2009; 11: 10–18.
36.
LandisJRKochGG. The measurement of observer agreement for categorical data. Biometrics1977; 33: 159–174.
