Stochastic data-driven optimization for multi-class dynamic pricing and capacity allocation in the passenger railroad transportation

Abstract

As for any passenger transportation service provider, pricing and capacity management are two critical tools for the profitability of a passenger railroad service provider: pricing affects the demand for the services and the capacity management sets the availability of the services in advance. In this study, an expert system is developed as a decision support tool for a passenger railroad service provider’s integrated pricing and capacity management problem, which has great significance for the success of the service provider. Considering the demand uncertainty, we first formulate the integrated pricing and capacity management problem as a stochastic nonlinear integer programming (SNLIP) model. This model includes dynamic pricing and dynamic capacity allocation decisions for multiple service classes over a planning horizon in order to maximize profit. Also, several key characteristics of the passenger railroad service operations are captured in the model. Due to inherent demand uncertainty as well as the dynamic nature of the problem, a fast and efficient solution approach is needed. Therefore, a simulation-based procedure embedded in a simulated annealing method is proposed to solve the model. Several real-life cases from Fadak Five-Star Trains (an Iranian luxurious passenger railroad service provider) are presented to demonstrate the model and the solution approach. The results of the case studies show the operational and profitability impacts of using the proposed decision support tool as well as its potential capabilities for practical use by other service providers.