Recreating passenger mode choice-sets for transport simulation: A case study of London, UK. Issue 1 (March 2018)

• Record Type: Journal Article
• Authors: Hillel, Tim; Elshafie, Mohammed Z E B; Jin, Ying
• Is Part Of: Proceedings of the Institution of Civil Engineers. Volume 171:Issue 1(2018:Mar.)
• Page Start: 29
• Page End: 42
• Publication Date: 2018-03
• Subjects: mathematical modelling -- transport management -- transport planning; Construction industry -- Technological innovations -- Periodicals; Construction industry -- Information technology -- Periodicals; Building -- Technological innovations -- Periodicals; Smart structures -- Periodicals; Smart materials in design -- Periodicals; 624.0285
• DOI: 10.1680/jsmic.17.00018
• Languages: English
• ISSNs: 2397-8759

Abstract

Abstract : Urban transport infrastructure is under increasing pressure from rising travel demand in many cities worldwide. It is no longer sustainable or even economically viable to cope with increased demand by continually adding capacity to transport networks. Instead, travel demand must be managed by encouraging passengers to adapt their travel behaviour. This approach necessitates a significantly deeper understanding of the seemingly random variations of passenger flows than is afforded by the current travel demand modelling techniques. This study presents a new modelling framework for predicting travel mode choice, through recreating and analysing the choice-set faced by the passenger at the time of day of their travel. A new data set has been developed by combining individual trip records from the London Travel Demand Survey (LTDS), with systematically matched trip trajectories alongside their corresponding mode alternatives from an online directions service and detailed estimates of public transport fares and car operating costs. The value of the data set is demonstrated by comparing two models of passenger mode choice based on stochastic gradient boosting trees, one using only the LTDS data and the other with our full data set. The models are then used to identify the key factors driving passenger mode choice.