01556833

Component

Transportation networks are inherently uncertain due to random disruptions; meanwhile, real-time information potentially helps travelers make better route choices under such disruptions. This research estimates the first latent-class routing policy choice model using revealed preference (RP) data where travelers take routing policies in addition to fixed paths. A routing policy is defined as a decision rule applied at each link that maps possible realized traffic conditions to decisions on the link to take next. It represents a traveler’s ability to incorporate real-time information not yet available at the time of decision. A case study is conducted in Stockholm, Sweden and data for the stochastic time-dependent network are generated from hired taxi Global Positioning System (GPS) traces. A latent-class, latent-choice, latent-path, Policy Size Logit model is specified. Latent-class factor captures two classes of travelers, strategic travelers who follow routing policies and non-strategic travelers who follow fixed paths. A latent-choice factor is needed when estimating the model based on path observations because the choice of a routing policy is not observable and only its realized path is observed. A latent-path factor is needed when using GPS traces because they have relatively long gaps, and the observed path cannot be uniquely identified. The path and routing policy choice sets are generated and evaluated, and the latent-class routing policy choice model is estimated based on GPS traces. Results show that the probability of a traveler being strategic is significantly different from both 0 and 1 at 0.05 level, indicating that travelers are heterogeneous in terms of their ability and/or willingness to plan ahead and utilize real-time information. A benchmark fixed path choice model based on a transformed deterministic and static network is also estimated, and the lower final loglikelihood suggests the loss of explanatory power due to simplified assumptions of network stochasticity and travelers’ utilization of real-time information. This research demonstrates that an appropriate route choice model for uncertain networks should consider the underlying stochastic travel times and traveler heterogeneity in terms of real-time information utilization.

This paper was sponsored by TRB committee ADB20 Effects of Information and Communication Technologies (ICT) on Travel Choices. Alternate title: A Latent-Class Routing Policy Choice Model with Revealed Preference Data.

01550057

15-1963

English

16p

2015

Transportation Research Board 94th Annual Meeting

Digital/other

Figures; Maps; References; Tables

Case studies; Choice models; Global Positioning System; Logits; Real time information; Revealed preferences; Route choice; Travel time

Latent class models

Stockholm (Sweden)

Highways; Planning and Forecasting; I72: Traffic and Transport Planning

Transportation Research Board Annual Meeting 2015 Paper #15-1963

TRIS, TRB, ATRI

Dec 30 2014 12:42PM