A model for time of day and mode choice using error components logit

Abstract

11 The severity of road congestion not only depends on the relation between traffic volumes and network 12 capacity, but also on the distribution of car traffic among different time periods during the day. A new error 13 components logit model for the joint choice of time of day and mode is presented, estimated on stated 14 preference data for car and train travellers in The Netherlands. The results indicate that time of day choice 15 in The Netherlands is sensitive to changes in peak travel time and cost and that policies that increase these 16 peak attributes will lead to peak spreading. In the Netherlands, the Dutch National Model System for traffic and transport (LMS) has been 21 used since 1990 to predict the responses of travellers to a wide range of developments, such as 22 changing travel times (e.g. from congestion) or the imposition of time-dependent road user 23 charging. One of the results of these simulations has been that the choice of when to travel (time of 24 day choice) greatly affects the amount of congestion on the road network and that policies aiming 25 at spreading out peak travel can be effective instruments to relieve congestion. U N C O R R E C T E D P R O O F 26However, these results rely to a large extent on a time of day choice sub-model within the 27 Dutch National Model System, which is more than 10 years old. Moreover, this sub-model uses a 28 rather simple and restrictive specification: only three time periods are distinguished within a 29 working day, there are no links between the outward and inward leg of the same tour, and the 30 model is multinomial logit (MNL). Since then, congestion has increased considerably, casting 31 doubt about whether the old specifications will still hold, while modelling capabilities also im-32 proved. 33In this paper, a new model for the joint choice of mode and time of day is presented and es-34 timated on new stated preference data. The model is not restricted to shifts between large time 35 periods and follows the error components logit (EClogit; also called mixed MNL) specification. 36 Using this type of model, one can take account of the different degrees of substitution between 37 time periods (e.g. greater substitution between nearby periods than between periods further away 38 from each other) and between time of day alternatives and alternative modes. It is a tour-based 39 model, in which outbound time of travel, duration of the activity at the destination and mode 40 choice are determined simultaneously. 41This new model was developed to become the basis of a new time of day choice sub-module of 42 the Dutch National Model System. It also covers public transport users, whereas the old module 43 only referred to the time of day choice of car drivers. Most empirical studies into the choice of time of day have considered only the demand of 52 travellers for travel at different points of time or periods in time (mostly using discrete time pe-53 riods) for given travel time and/or travel cost. Impacts on congestion and feedback to ...

On the development of time period and mode choice models for use in large scale modelling forecasting systems

“…However, when we consider the potential impacts of the selection of the SP data, the fact that PRISM also considers destination switching and several modes other than car driver and public transport, and the fact that it may generally be expected that SP responses are more elastic than RP responses, then such a difference is to be expected. Additionally, the forecasting results of de Jong et al (2003), who apply MMNL models to the Dutch SP data analysed in this paper, seem to be consistent with the NL results shown here.…”

“…Each of these data collection exercises formed part of larger urban, regional or national model development projects. These are, in chronological order of the collection date: the APRIL model for London (Bates & Williams, 1993;Polak & Jones, 1994), work in The Netherlands (de Jong et al, 2003) and the PRISM model developed for the West Midlands region of the United Kingdom (RAND Europe, 2004). In the remainder of this paper, these three datasets will be referred to respectively as the APRIL, Dutch and PRISM datasets.…”

“…These models are contributing substantially to our understanding of how travellers make time-of-day decisions (cf. de Jong et al, 2003;Hess et al, 2006). The resulting models, however, tend to be far too complex and far too data intensive to be of use for application in large scale model forecasting systems, where socio-economic detail is limited and detailed scheduling information is rarely available in the model implementation structure.…”

Random covariance heterogeneity in discrete choice models

“…de Jong et al, 2003). An initial Revealed Preference (RP) survey was conducted to select a sample of respondents for the follow up Stated Preference (SP) survey.…”

The Fuzzy Lattice Order Decision-Making Method for Travel Mode Choice Based on The Travel Time Reliability

“…The literature [4], [5] respectively give the traffic mode model and satisfaction criteria based on the value of travelers' behavior. All the traffic mode choice method above is based on the assumption which the travel time is reliable.…”