Sustainable regional rail system pricing using a machine learning-based optimization approach

Abstract

AbstractRegional transport pricing is indeed very vital in urban settings where the transportation network is spread out across large areas and can influence travel behavior and the sustainability of cities. Therefore, in addition to existing pricing systems, such as flat fare, distance-based fare, and zonal pricing, this study proposes a sustainable approach to regional rail system pricing using rent prices and a transportation affordability index. The proposed model aims to reduce commuters' overall travel distance in order to reduce air pollution and maintenance costs for public transportation vehicles. Rent-based pricing encourages people to rent houses in regions that shorten their travel distances and fill a gap in the literature on regional rail system pricing by dealing with the decentralization of the cities. A two-step clustering and non-linear optimization modeling approach are proposed based on face-to-face surveys with regional rail system passengers. For various clusters of stations, rent per income rates and rental-based ticket prices were obtained. Furthermore, a sensitivity analysis is conducted to evaluate different conditions of the affordability index and rent prices in the studied regions. Compared to the current pricing system, ticket revenues increased by 3.88% and 1.68% in rent-based pricing.