A PRIORI EVALUATION OF PUBLIC TRANSPORTATION SERVICE
Résumé
This paper proposes one approach for evaluating the efficiency of an urban transportation service in multi actors context and in multicriteria framework. A priori analysis is intended for clarifying options, by reducing uncertainty and, more generally, by providing information about the alternatives in their specific context. Firstly, a diagnosis phase based on hierarchical ascending classification is performed for grouping actors in agreement with pertinent criteria that they have selected. The evaluation of the main impacts is based on the aggregate probability to use the service for different social categories. For it, stated preferences are collected from a representative sample size and an algorithm based on Fuzzy Logic allows us to obtain the probability to use a new transportation service. The robustness analysis of behavioural changes in agreement with external criteria (e.g. price of petrol) is studied using Taguchi's method. The search of an optimum solution consists in defining a combination of levels of criteria for the best satisfaction of the requirements and constraints of the ones (urban planners, traffic engineers) and the exigencies of the others (users, residents). For it, Doehlert's framework is used. A study case shows the main steps of our approach.