Exploring train passengers’ arrival rate using smart card data
2015 (English)Conference paper, Abstract (Other academic)
In the context of public transport, passengers’ arrivals at a station can indicate their perception of the transit system’s reliability, as well as the extent to which they consult timetables in journey planning. Moreover, the arrival rate is an essential element in determining passenger waiting time at astation, which itself is a critical component of the level of service in public transit journeys.Despite the significance of the passenger arrival rate, researchers often make simplistic assumptions about passenger arrivals in the majority of the literature. In high frequency routes, for instance, passengers are usually assumed to arrive uniformly at the stops without consulting the timetable; in low frequency routes, the arrivals are usually assumed to be mainly concentrated around the scheduled departure times. These assumptions can be validated with on-site manual data collection, but these can be costly and intractable to apply to network-level studies.Consequently, the primary objective of this paper is to develop a more comprehensive methodology forcapturing passengers’ arrival distributions, using the numerous boarding records available from transit smart card data. To serve this purpose, boarding data from a selected set of train stations are extracted from the smart card dataset of Southeast Queensland in Australia. This study analyses how passenger arrival patterns vary as a function of service frequencies. It is hypothesized that there is a threshold where a random arrival regime gradually transforms into a timetable coordination arrival pattern. Furthermore,this threshold is a random variable that varies in the population and results with mixed regimes for intermediate ranges of headways. The dataset includes services with headways ranging from 5 to 60 minutes and hence facilitates this analysis. Furthermore, sinceno real-time information is available to passengers in Southeast Queensland, it is assumed that passengers only consult fixed timetables to arrive at stops. Therefore, the passenger arrival rate is assumed to be a function of the given headway.
Place, publisher, year, edition, pages
Transport Systems and Logistics
IdentifiersURN: urn:nbn:se:kth:diva-183028OAI: oai:DiVA.org:kth-183028DiVA: diva2:906773
18th Euro Working Group on Transportation, EWGT 201514-16 July 2015, Delft, The Netherlands
QC 201603292016-02-252016-02-252016-03-29Bibliographically approved