Line planning and differential pricing for high-speed parallel trains

Abstract

The rapid development of China's high-speed railway (HSR) has fundamentally altered intercity passenger travel patterns. Escalating construction and operational expenses have resulted in persistent financial losses for HSR operating companies. This paper introduces a nonlinear optimization approach for HSR operations, focusing on line planning problem (LPP) and differential pricing problem (DPP). It takes into account the boundedly rational choice behavior of stochastic users. Taking into account the differences in passenger choice behavior, latent class analysis (LCA) is used to classify passengers. Three influencing factors, namely departure time, travel time and fare, were selected to establish a utility model of parallel trains with three reference points, and the perceived cost of parallel trains by heterogeneous passengers was characterized through the prospect value. A bi-level planning model has been developed to optimize both lines and fares. The primary objective is to maximize the profits of the HSR operating companies while reducing the perceived costs to passengers. Operational lines and frequencies are generated using possible line sets (PLSs) as a foundation. To identify the most effective lines and fare strategies, a dual approach is employed, combining particle swarm optimization (PSO) with sensitivity analysis (SA). In order to validate the practicality and efficacy of the model and algorithm, a numerical case study based on the Nanning-Guangzhou (NG) HSR is presented.