A Doubly Dynamic Assignment Model for Congested Urban Transit Networks



A Doubly Dynamic Assignment Model for Congested Urban Transit Networks

Authors

NUZZOLO A, CRISALLI U, University of Rome and RUSSO F, University of Reggio Calabria, Italy

Description

Dynamic path choice and assignment modelling for low frequency transit services, using a schedule-based approach, has received recently great attention in the literature (Nuzzolo and Russo, 1997; Nuzzolo et alii, 1997; Carraresi et alii, 1996; Florian, 19

Abstract

Dynamic path choice and assignment modelling for low frequency transit services, using a schedule-based approach, has received recently great attention in the literature (Nuzzolo and Russo, 1997; Nuzzolo et alii, 1997; Carraresi et alii, 1996; Florian, 1998), while less interest is given to high frequency services, which are the subject of this paper. One of the first within-day assignment models for high frequency services is the one introduced by Tong in 1986 (described in Wong and Tong, 1998), that uses a network model based on service timetables. Recently Hickmann and Wilson (1995) presented a within-day dynamic assignment model that allows consideration of time variation in supplied services, using a deterministic path choice model which takes into account information on bus arrival times at stops. Starting from this model, but with quite different results, Nuzzolo and Russo (1998) obtained stochastic within-day dynamic path choice models for regular and non-regular high frequency services, with or without information systems to users.

This paper presents a dynamic path choice model, based on user within-day utility updating and day-to-day cost learning processes, for high frequency services with information on waiting times. The model is used in a doubly dynamic assignment model that has been applied to a real-size urban transit network.

Publisher

Association for European Transport