基于旅客出行偏好的客运通道内不同客运方式分担率研究

Various travel modes coexist in the same passengers’ corridor. The passengers have the characteristics of heterogeneity because of the differences of their travel purpose, level of income, ect. Besides, there are differences between these travel modes in the level-of-service attributes, which is indeterminate and partial for travelers since they lack of travel experiences and information. It results in variance between their expected perception and actual perception. Therefore, a complex decision is shown as multi-level and phase-wise one with feedback loop in long term for a large number of travelers. The individual behavior characteristic of heterogeneous passenger is used to emerge the macro characteristics of different travel modes. The individual attributes (i.e. heterogeneous, bounded rationality, differences in preferences) are generally not taken account in the traditional studies on travel choice and passenger shares. The paper focuses on four different travel modes in one passengers’ corridor, including high speed railway, traditional railway, civil aviation and motorway. Based on the travel preferences of heterogeneous individuals, we find out the key elements which affect the decision of travel choice with the rough set approach. The travel choice decision is modeled with the idea of bounded rationality across individuals based on Multi-Agent of comprehensive decision-making mechanism, which combines the expected utility function with complete rationality. Also, the paper explores the group effect in macroscopic scale, that is to say, the passenger shares in different modes. It will guide travelers with better level of service in their choice decision, through analyzing the response sensitivity of the key elements of the decision.

多种客运方式在客运通道内并存，旅客因出行目的、收入水平等差异导致其出行偏好具有异质性特点。同时，由于旅客缺乏出行经验而不能完全掌握各种客运方式的服务差异，使其预期与实际出行感知存在偏差，在多因素耦合作用下表现为多层次、分阶段、具有循环反馈机制的复杂决策，通过宏观涌现最终决定不同客运方式的分担率。传统旅客出行方式选择与分担率研究大多不考虑旅客在出行方式选择过程中异质性、有限理性和出行偏好差异，本研究针对客运通道内高铁、普铁、民航和公路四种客运方式，从异质旅客出行偏好出发，运用粗糙集方法筛选影响旅客出行方式选择的关键因素；引入“有限理性”思想，结合传统基于期望效用理论的“完全理性”出行方式选择方法，构建综合决策机制下基于Multi-Agent的旅客出行方式选择模型；探究客运通道内旅客在群体作用的宏观涌现，即不同客运方式的分担率；分析出行方式选择中关键影响因素的灵敏度，更好地服务于旅客出行引导。

高速铁路的出现，以其快速、便捷、舒适等优势为旅客提供了一个新的出行选择，在丰富传统的综合运输体系的同时也为其他客运方式带来了冲击与挑战，对旅客的出行方式选择行为产生了影响。旅客因出行目的、收入水平等差异导致其出行偏好具有异质性和旅客实际出行感知存在偏差，使得旅客在多种客运方式并存选择行为在多因素耦合作用下表现为多层次、分阶段、具有循环反馈机制的复杂决策，通过宏观涌现最终决定不同客运方式的分担率。.本研究以旅客出行偏好为研究对象，筛选出异质性旅客出行选择影响因素并计算权重，构建了完全理性与有限理性综合作用下的旅客出行方式选择模型。以旅客个体决策结果作为初始模糊选择概率，建立了基于Multi-Agent的旅客出行方式选择与运量分担仿真模型，并在此基础上对旅客出行方式选择影响因素进行了灵敏度分析。本研究的主要研究结果如下：.（1）具有不同出行属性的旅客对出行需求影响因素的权重偏向具有不同的特点。所有旅客对便捷性的关注程度较高；商务出差旅客比较关注舒适、快速、服务周到性和便捷性。旅游或探亲访友旅客更关注舒适、便捷和准时性。务工旅客和学生旅客则主要关注便捷性，低收入旅客最关注经济、便捷和安全性，高收入旅客关注快速、便捷、舒适性。.（2）Multi-Agent的运量分担仿真模型以北京-武汉客运通道内旅客为研究对象设定仿真参数，对日常、周末及高峰期时段旅客的出行方式选择行为进行实证研究得到了在北京-武汉客运通道内不同交通方式的旅客分担率。.（3）通过灵敏度分析深度挖掘了旅客出行选择影响因素的作用差异。对于日常出行旅客灵敏度较高的是票价、准时性、服务性和便捷性因素；对于周末出行旅客灵敏度较高的是票价、准时性和服务性因素；对于高峰期出行旅客灵敏度较高的是票价、准时、安全因素。.研究结论能够更好地服务于旅客出行引导的同时，对于建立资源配置合理、客流分担均衡的综合运输通道具有现实意义。