基于选择偏好异质性的多式联运服务网络协同优化研究

In the organization process of intermodal transport, due to the low collaborative efficiency of different modes of transport in the transshipment links, the connection of multiple transport modes is not smooth, and the integrated transport service system has not yet formed. Meanwhile, as customers pay more attention to the quality of transport services, traditional service network design methods have been difficult to meet with customer individualized and diversified service requirements, which have restricted the exertion of comparative advantages of different modes of transport and is one of the key issues to be solved in the current intermodal transport development process. Based on the data mining of customers’ transport service choice behavior, the project constructs a choice behavior model for intermodal transport customers based on Regret Theory and elaborates the formation mechanism of customers' choice behavior. Furthermore, considering the effects of latent variables on customers' choice preferences, an endogenous market segmentation method is proposed to analyze the heterogeneity of choice preferences of different types of customers for transport price, transport time, and transport reliability, and to measure their Values of Time (VOT) and Values of Reliability (VOR). Finally, taking VOT and VOR as components of the generalized cost of the service network, a collaborative optimization model for intermodal service network is constructed with constraints of transshipment synchronization, arrival and departure time window, and the transport capacity limitation, to provide a theoretical basis for the improvement of collaborative efficiency of intermodal transport and the construction of integrated transport services system.

在多式联运组织过程中，由于不同运输方式在中转环节的协同效率低，导致多方式衔接不畅，一体化运输服务体系尚未形成。同时，随着客户对运输服务质量的愈加看重，传统的服务网络设计方法，已难以满足客户个性化、多元化的服务需求，制约了不同运输方式比较优势的发挥，是当前多式联运发展过程中亟待解决的关键问题之一。本项目在对客户的运输服务选择行为进行数据挖掘的基础上，构建基于遗憾理论的多式联运客户选择行为模型，阐述客户的选择行为形成机理；进而，考虑隐性变量对客户选择偏好的影响，提出内生性市场细分方法，分析不同类型客户对运输价格、运输时间、运输可靠性的选择偏好异质性，并测算其时间价值、可靠性价值；最后，将时间价值与可靠性价值作为服务网络广义成本的构成要素，构建考虑中转同步约束、到发时刻约束、运输能力约束的多式联运服务网络协同优化模型，为提高多式联运的协同效率、构建一体化运输服务体系提供理论依据。

本项目针对考虑选择偏好异质性的多式联运服务网络协同优化问题，在传统服务网络设计理论的基础上，从需求偏好异质性的角度，提出根据客户在服务价值方面所表现的异质性，整合不同运输方式的服务特点，并进行差异化的服务类型设计。研究过程中，首先通过对客户选择行为模式的分析，构建了考虑随机效用最大化与随机遗憾最小化的选择行为模型，弥补了既有研究的不足，并对RUM模型、RRM模型的偏好参数估计给出了定量模型。同时，提出了基于选择偏好异质性的内生性市场细分方法，相对传统以货物外在特征进行划分的方法而言，内生性市场细分方法能够刻画样本中存在的偏好异质性。进而，结合案例分析，对西南地区的运输客户进行问卷调查，计算并分析得到样本中存在2类特征群体。其中，第一类客户(占比60.82%)的时间价值为1.99元/吨小时，可靠性价值为5.32元/吨小时；第二类客户(占比39.18%)的时间价值为0.81元/吨小时，可靠性价值为0.99元/吨小时，验证了两类客户在时间价值与可靠性价值方面有明显的差异性。最后，提出了改进的服务网络优化模型，并利用改进后的模拟退火算法求解。计算结果显示，在考虑选择偏好异质性的情况下，33.3%的运输需求得到了更优的运输服务，且总成本下降了6%，验证了考虑选择偏好异质性的情况下，服务网络的设计结果相对传统服务网络设计模型的结果更优。