Application of Data Clustering to Railway Delay Pattern Recognition

Table 1

Review of previous uses of univariate statistics in railway operation analysis.

	Environment		Technique					Purpose
	Real operation	Simulation	Distribution fitting	Test significance	Regression analysis	Percentile sorting	Visual inspection
Goverde et al. (2001)	X		X					Distributions of primary and secondary delays
Sipilä (2010)		X		X				Comparison running time supplement strategies
Lindfeldt (2010)	X	X	X					Distributions of primary delays from real operation for simulation
Olsson and Haugland (2004)	X				X			Factors affecting punctuality
Gorman (2009)	X				X			Factors that generate delays on single track lines
Lindfeldt (2010)	X				X			Factors that increase delays on line segments
Lindfeldt and Sipilä (2014)		X		X				Travel times with different operation models, with/without free freight operation
Shih et al. (2014)		X			X			Factors affecting average delay per train-mile
Schittenhelm and Richter (2009)	X					X	X	Visual inspection of quantile-based representation of deviations and change in deviation
Richter (2010)	X					X		Delay tabular representation and sorting train service performance
Peterson (2012)	X					X		Rolling average delay for specific train services
Andresson et al. (2011)	X						X	Assessment of effectiveness of running time supplements
Andresson et al. (2013)	X						X	Identification of critical points for robustness
van Oort et al. (2015)	X						X	Delay profiles, headway profiles