An AHP Based Day-to-day Route Choice Model: a Fuzzy Decision Making Approach
N Khademi, M Rajabi, A S Mohaymany, Iran University of Science and Technology, IR; S Zerguini, Universite Paris/LVMT, FR
This paper present a day-to-day route choice modeling process using Analytical Hierarchy Process (AHP) and the two main criteria:1) travel time and 2) predictability of travel time.
Modeling day-to-day route choice behavior has a critical role in predicting the state of the transport network. Although we do not repudiate the critical role of travel time variability in route choice behavior, we endeavor to construct a more descriptive model in this paper to understand the behavioral system in which a traveler responds endogenously to his/her subjective assessment due to the variability in travel time. Hence, in this study, we explore the impact of day-to-day travel time variability from the viewpoint of the traveler. In the context of day-to-day route swapping derived from our study, the impact of day-to-day travel time variability from the viewpoint of a traveler is explored; hence, the route choice behavior is tied mainly to the travel time prediction error. That is to say, we try to exploit the fact that travelers do not react based on amount of variability, but they do respond to the failure or success in predicting travel time. In our paper we construct an Analytical Hierarchy Process (AHP) model considering travel time and the predictability of travel time as the two main criteria in route choice modeling. The predictability of travel time is captured by the ability of the traveler to perform successive predictions for travel time in routine circumstances. The main distinction of this study with the previous studies in the domain of day-to-day route choice modeling is the manner of scaling and aggregating the two main route choice decision criteria, travel time and predictability of travel time. For this reason, the proposed method employs fuzzy AHP, a powerful and established prioritizing tool, for modeling traveler's decision-making. The fuzzy AHP contains the goal, decision criteria and decision alternatives. It ultimately calculates the final alternative weights with respect to the goal. In our decision problem, the fuzzy AHP synthesizes the predicted travel time and the perceived predictability measures for travel time for each route and then yields the final preferences of routes from the viewpoint of the travelers. The model is used in an iterative day-to-day fashion. In each day (iteration) the AHP is employed and the decisions of the travelers are modeled. Hence, the traffic volume in each route and in each day is calculated. Finally, a laboratory-like experiment is preformed to verify the validity of the proposed method through comparing the observed route choices and the modeled ones. In order to better model the decisions, the heterogeneity in the travelers' behavior is captured by considering optimism and pessimism for the system of travelers' subsequent feelings and intuition.
Association for European Transport