Predictive modelling of running and dwell times in railway traffic

Abstract

Accurate estimation of running and dwell times is important for all levels of planning and control of railway traffic. The availability of historical track occupation data with a high degree of granularity inspired a data-driven approach for estimating these process times. In this paper we present and compare the accuracy of several approaches to model running and dwell times in railway traffic. Three global predictive model approaches are presented based on advanced statistical learning techniques: LTS robust linear regression, regression trees and random forests. Also local models are presented for a particular train line, station or block section, based on LTS robust linear regression with some refinements. The models are validated and compared using a test set independent from the training set. The applicability of the proposed data-driven approach for real-time applications is proved by the accuracy of the obtained estimates and the low computation times. Overall, the local models perform best both in accuracy and computation time.

Application of Data Clustering to Railway Delay Pattern Recognition

“…Compared to big data techniques proposed in the literature for other purposes in analysis of transport operation [11,23,24,32], this method relies on readily available data and does not need detailed knowledge on the infrastructure and occupation of individual blocking sections. It can therefore be scaled to different levels of detail or transferred to other modes of transportation where delay can be measured at fixed points on a given path, such as bus networks or air traffic.…”

“…Kecman and Goverde [32] apply big data techniques to predict running and dwelling times from actual operation data, based on records from block sections occupations. The study uses random forests of tree-based models, to predict nonlinear relations between input variables and process times, with sufficient robustness to outliers in the data, lowered risk of overfitting, and focus on real time application.…”

“…Typically, the track circuit boundaries do not correspond exactly to the platforms, and an offset is generated between the time recorded by the automatic system and the actual time a train arrives at the platform or departs. This is a rather common problem, and it is also reported in the Netherlands [32] and Norway [12]. For the Danish network, a correction factor was calculated by Banedanmark using statistical analyses of GPS positions of train trajectories in collaboration with the main rail operator, DSB.…”

Testing the Generality of a Passenger Disregarded Train Dwell Time Estimation Model at Short Stops: Both Comparison and Theoretical Approaches

“…To the best of our knowledge, so far, regression is the only approach that has been used in passenger disregarded models. Hansen et al [18] and Kecman and Goverde [19] estimated the train dwell time as a function of its arrival delay, which was derived from the track occupancy data of the Dutch railway. Li et al [2] proposed a train dwell time estimation model that improved the accuracy of these models.…”

Methodology for Determining Dwell Times Consistent with Passenger Flows in the Case of Metro Services

“…Statistical techniques were the first kind of approach proposed in the literature [2,3] and were borrowed from the bus service sector [4][5][6]. More recent contributions in this field are provided by [7][8][9][10]. However, broadly speaking, these methods are not generic enough to be applied in contexts other than those in which they were developed, and moreover, they provide no details about passenger behavioural rules when a train arrives.…”