基于人工智能系统的地铁收益优化与分时定价理论研究

Research on Subway Revenue Optimization and Time-sharing Pricing Theory Based on Agent

目前,地铁在高峰期存在车厢内客流量大、过度拥挤等问题,且城市地铁运营收益状态存在较大改善空间。为此,本文从收益管理和拥挤定价角度出发,采用SP调查和数据挖掘等研究方法,在对乘客心理和行为分析的基础上,研究构建基于人工智能系统(Agent)仿真模型的地铁分时定价机制。尝试将收益管理应用于地铁定价策略中,为实现提高地铁服务性和经济性的双赢目标提供建议和对策。结果表明:3种不同收入人群乘客Agent的转移率随时间而变化,转移率因票价上涨而增加;而且地铁票价涨幅过大会导致一部分乘客转移到地面公交系统。可见地铁的分时定价能有效分散客流,降低满载率。

In order to effectively alleviate the problems of large passenger flow and overcrowded in the subway compartment during the rush hour, and improve the profit state of the urban subway operation.From the point of view of revenue management and congestion pricing, based on the analysis of passengers’ psychology and behavior, a time-sharing pricing mechanism based on Agent simulation model is developed by means of SP survey and data mining. This paper attempts to apply revenue management to the pricing strategy of metro, and provides suggestions and countermeasures for realizing the win-win goal of improving the service and economy of subway. The results show that the transfer rate of Agent in three different income groups varies with time, and the transfer rate increases with the increase of ticket price, and the increase of subway ticket price leads to the transfer of some passengers to the ground bus system. It can be seen that the time-sharing pricing of subway can effectively disperse the passenger flow and reduce the full load rate.