An Optimization Decision Model of Urban Public Transit Fare Structures

Abstract

This paper investigates and compares several transit fare structures including flat fare, station-based, and mileage-based fare. For each fare structure, there exists an optimal fare pattern inducing to the highest total social benefit and total net profit. A bi-level model is then proposed to determine the optimal fare pattern. The upper-level problem intends to address the maximization of total social benefit from the perspective of transportation authority and the maximization of profit of transit authority, whereas the lower-level problem is a path-based stochastic transit assignment problem with elastic demand. A modified genetic algorithm is adopted to solve the bi-level model. A numerical example of a 9-node transit network is provided to illustrate that differentiated fare structures and have a better financial performance than flat fare.