需求族的动态管理及其服务资源配置优化研究

A demand family refers to a series of demand classes that have similar requirements and a few differences on the function, utility, experience from a product or service. The urgent needs and the conventional needs of the blood constitute a demand family with priorities while the ticket needs for three classes with different prices (first class, business class, economy class) constitute a demand family with value differentiation. Based on the practical background of a demand family, this project will start from the methodology of a demand family and focus on the demand family with priorities and the demand family with value differentiation. This project shall use the related theory of management science and describe the classification of demands and the dynamic changes of the classification. By using the big data related to the classification of demands, this project will investigate the factors that affect different demand classes and construct the data-driven predictive demand model. When the classification of demands is fixed, this project will separately consider the optimal allocation of service resources for the demand family with priorities and the one with value differentiation. When the classification of demands is changed, this project will reconstruct the data-driven predictive demand model and the dynamic function of service resources allocation policy. By comparing this new model with the old one, this project tries to find the differences between the data-driven predictive demand models and the optimal allocation policies of service resources. This project will provide a new theoretical perspective and an important decision-making reference for firms to rationally allocate service resources.

需求族是指对同一种产品或服务在功能、效用、体验等方面的要求大致相同但又存在些许差异的一系列需求类。对血液的紧急需求和常规需求构成了存在不同优先级的需求族，而对头等舱、商务舱、经济舱三个不同价格等级的机票需求则构成了存在价值差异的需求族。本项目基于需求族的实际背景，从需求族的分类方法入手，以存在优先序的需求族和存在价值差异的需求族为研究对象，运用管理科学的相关理论，对需求的分类以及分类的动态变化特性进行描述。利用需求族分类相关的大数据，深入探讨各类需求的影响因素，构建数据驱动的需求预测模型。当需求族的分类不变时，分别考虑存在优先序的需求族和存在价值差异的需求族的服务资源配置策略。当需求类的分类变化时，重新建立新的需求族预测模型和服务资源配置的动态方程，对比分析预测模型的改变和服务资源配置策略的差异。本项目从需求族的动态管理角度为企业合理配置服务资源提供新的理论视角和重要的决策参考依据。

本项目提出“需求族”概念，指的是对同一种产品或服务在功能、效用、体验等方面的要求大致相同但又存在些许差异的一系列需求类。围绕运用不同方法构建的产品/服务需求族，本项目旨在研究如何通过服务资源的优化配置实现需求满足。具体研究工作主要涵盖以下三方面：优先序需求族下的供应链优化、价值差异需求族下的在线服务策略以及感知价值差异需求族下的产品销售策略。（1）优先序需求族下的供应链优化：从理论上刻画了最优订货策略、最优在途库存加速策略以及最优库存分配策略的结构并给出了这三个最优策略所具备的单调性质，提出了三种不同的启发式策略，这部分成果发表于管理领域UTD24期刊Production and Operations Management上；（2）价值差异需求族下的在线服务策略： 基于用户学习效应和效用理论，建立了用户效用与用户学习之间的关系，分析了服务提供商的免费试用策略，包括最优的免费试用时间长度和试用结束后的服务定价，这部分成果发表于管理领域知名期刊IEEE Transactions on Engineering Management上；（3）感知价值差异需求族下的产品销售策略： 利用博弈论和委托代理理论分别构建了直销模型和意见领袖营销模型，得到了有启示意义的解析解，这部分成果以论文形式发表于国际期刊Journal of Cleaner Production上。