交通运输系统工程与信息 ›› 2024, Vol. 24 ›› Issue (5): 259-267.DOI: 10.16097/j.cnki.1009-6744.2024.05.024

• 系统工程理论与方法 • 上一篇    下一篇

高铁列车动态定价与预售时段划分综合优化方法

许景,邓连波*,刘华儒,胡心磊   

  1. 中南大学,交通运输工程学院,长沙 410075
  • 收稿日期:2024-05-26 修回日期:2024-08-14 接受日期:2024-08-20 出版日期:2024-10-25 发布日期:2024-10-23
  • 作者简介:许景(1998- ),女,山西大同人,博士生。
  • 基金资助:
    湖南省自然科学基金(2023JJ30703, 2023JJ40784)。

Joint Optimization of Dynamic Pricing and Pre-sale Period Division for High-speed Trains

XU Jing, DENG Lianbo*, LIU Huaru, HU Xinlei   

  1. School of Traffic and Transportation Engineering, Central South University, Changsha 410075, China
  • Received:2024-05-26 Revised:2024-08-14 Accepted:2024-08-20 Online:2024-10-25 Published:2024-10-23
  • Supported by:
    Natural Science Foundation of Hunan Province, China (2023JJ30703, 2023JJ40784)。

摘要: 立足高铁收益亟需提高和灵活化市场票价体制实施的背景,本文考虑需求在预售期内各天的波动性和差异性,以及预售时段划分方案对铁路收益的影响,研究高铁列车动态定价与预售时段划分的综合优化问题。为预售期内每天构建独立的弹性需求函数,考虑列车能力约束、需求约束及票价递增约束等条件,建立高铁列车动态定价与预售时段划分综合优化大规模非线性模型。根据模型特点,设计双层遗传—模拟退火算法求解,将优化问题分为外层预售时段划分、内层动态定价与票额分配的双层优化问题,内外两层分别采用遗传算法和模拟退火算法求解。最后,采用一个数值算例验证优化模型和求解算法的有效性,并探讨不同预售时段数量下的划分结果。结果表明,随预售时段数量的增加,预售期的划分主要集中在后半段;预售时段数量为5时,优化后,该数值算例的收益提高了约1.21%。

关键词: 铁路运输, 动态定价, 双层遗传—模拟退火算法, 高铁列车, 时段划分, 席位分配

Abstract: Based on the need to enhance high-speed rail revenue and implement a flexible market ticket pricing system, this paper focuses on the joint optimization of dynamic pricing and pre-sale period division considering the demand fluctuations and differences on each day during the booking horizon, as well as the impact of the pre-sale period division on railway revenue. Separate elastic demand functions are constructed for each day. A large-scale nonlinear model is developed to optimize the dynamic pricing and pre-sale period division for high-speed trains in consideration of the train capacity constraints, demand constraints, and price-related constraints. To solve the optimization problem, a bi-level genetic-simulated annealing algorithm is designed according to the model's properties. The optimization problem is divided into an outer-level pre-sale period division problem and an inner-level dynamic pricing and seat allocation problem, which are solved by genetic algorithm and simulated annealing algorithm, respectively. At last, a numerical instance is provided to evaluate the effectiveness of the optimization model and solution algorithm, and the results for different numbers of pre-sale period are discussed. The results indicate that as the number of period increases, the division of the booking horizon primarily concentrates on the latter half. For a case with five periods, the optimized revenue increased by approximately 1.21%.

Key words: railway transportation, dynamic pricing, bi-level genetic-simulated annealing algorithm, high-speed trains, pre-sale period division, seat allocation

中图分类号: