Fuzzy Logic Model of Mode Choice



Fuzzy Logic Model of Mode Choice

Authors

HOLLAND R, TORG, University of Newcastle upon Tyne, UK

Description

Various authorities see the provision of better quality public transport as a possible incentive to car ddvers to switch mode (DETR, 1998). Both real-time information and the provision of more modern vehicles are among the possible improvements that could

Abstract

Various authorities see the provision of better quality public transport as a possible incentive to car ddvers to switch mode (DETR, 1998). Both real-time information and the provision of more modern vehicles are among the possible improvements that could be made. The assessment of the impact of such measures is often predicted using one or more stages of a four-stage model of transportation.

Generalised cost models of mode choice and route assignment require that certain assumptions are made about the traveller. First, it is assumed that all tdp makers are stdctly rational utility maximisers, and all decision-makers are perfectly informed (about the trip). The random differences in perception of these costs allows the assignment of tdps between modes and routes to be calculated using a Iogit function. In reality, such an omniscient individual (homo econimus) would be required to gather and store large amounts of information, to perform very complex computations, and to make consistent decisions based on those computations. The generalised cost approach formalises the trade of between travel time and travel cost through the value of travel-time savings. Factors such as comfort and convenience are often referred to as "soE factors, and are bundled up in a modal constant, which can be expressed as a time or cost penalty for a given mode. Stated preference surveys can be used to determine both the value of travel time savings and the "worth" of the changes to the soft factors.

It is also assumed in generalised cost models that individuals have free choice, and that there is zero cost involved in searching for information and learning. The impetus for investigating a novel form of mode choice model was that when considedng the impact of changes in the quality of information provided to a decision-maker, the assumption of omniscient individuals may not be valid. Indeed, much work has been reported on the modelling the effect on in-vehicle information systems for car drivers.

Publisher

Association for European Transport