The Value of a Dynamic Information Device to the Individual User and to Traffic: a Probabilistic Model with Economic Analysis
F Leurent, T P Nguyen, Université Paris-Est, LVMT, ENPC, FR
Intended to analyze the value of a road information device to an equipped user, a model of route choice with two classes of users is worked out under assumptions of congestion, dynamic disruption, user perception and behavior.
The user of a road network chooses his route and/or departure time on the basis of his knowledge about the network state. Owing to the development of traffic measurement, data processing and telecommunication techniques, road information has kept improving both in scope, quality and availability. The disposal of sharp information enables the road user to make choices in a more opportunistic way, by adapting himself to the particular circumstances of the network during his trip.
The objective of the paper is to model the value of a road information device to a user, and more widely to the society. Put in other words, which profit may be derived from using a traffic information device or an information system and how would this profit compare to the cost of the device?
To gain insight into these issues, we work out a model of route choice with two classes of users respectively equipped and better informed with a dynamic component, or unequipped and less informed. We model the behavior of each class and also their interaction under specific assumptions of congestion, disruption and perception: the effective cost taking into account the congestion and the random disturbances; the subjective cost as perceived by a user on the basis of the information which is available to him. We simulate the model sensitivity with respect to two parameters: the total volume of demand and the rate of equipment, which determines the level of congestion and the population in each class. The individual (resp. collective) profit of using the information is addressed on the basis of specific indicators: the average cost undergone by each user class and the average cost to all the network users.
Much emphasis is put on the physical and economic significances of the random disturbances on the traffic conditions, thus inspiring our stochastic assumptions. This sharpness of interpretation distinguishes our model from previous approaches: the model of Maher and Hughes (1995) is shown to be irrelevant, as would be any "naive" model designed under commercial software for static traffic assignment.
KEYWORDS: traffic information; multiclass traffic assignment, Probit assignment model
Association for European Transport