面向复杂大群体的有限服务资源分配协调决策研究

According to the realistic background with expert numbers register of some large hospitals and train tickets purchasing during the Spring Festival, the project would study the coordinated decision-making method of limited service resource allocation based on the starting point of complex huge group characteristic classification.. Firstly, the project would design clustering method based on evaluating results of huge-scale service object under uncertainty environment, and the consistency aggregation approach, and also suppose consistency measurement model about different types of service object group. Based on the measurement value, the necessary coordinated group set would be built, and the priority order is formed at the same time. Secondly, we would study the key indicators of service forms affecting preference adjustment of various types of service object group, and establish the corresponding function expression, and form the group coordination satisfaction indexes. Thirdly, the paper would analyze the effect rule of the resource constraints and group satisfaction degree on key indicator adjustment of service form, and give the evolutionary game model of many kinds of service object groups to realize the effective matching between service forms and service demand..The project is expected to gain several innovative achievements with the aspects of huge-scale group clustering method, group evaluation consistency measure, the key indexes of influencing group preference adjustment, the evolutionary game model of many kinds of service object groups and etc. Relevant research results would contribute to resource allocation theory, and have guiding sense to improving limited resource allocation performance in service management.

项目以复杂大群体的特征分类为切入点，依托春运火车购票和大型医院专家号挂号等现实背景，研究面向复杂大群体的有限服务资源分配协调决策问题。项目研究不确定环境下基于大规模服务对象评价结果的群体聚类方法和评价信息一致性集结方法，设计不同类型服务对象群体评价一致性测度模型，并基于测度值对各类群体进行聚类，形成多个协调群体集合及其优先次序；提炼影响各类服务对象群体偏好调整的服务形式关键指标并建立相应的函数表达式，同时设计群体协调满意度指数，并剖析资源约束和群体满意度对服务形式关键指标调整的影响规律，建立多类服务对象群体演化博弈模型，以实现服务形式与服务需求的有效匹配。项目预期在大规模群体聚类方法设计、群体评价一致性测度模型构建、影响群体偏好调整的关键指标提炼、多类服务对象群体演化博弈模型设计等方面取得若干创新性成果。相关研究成果有助于丰富资源分配理论，对于提高服务管理领域有限资源分配绩效具有指导意义。

项目依托春运火车购票和大型医院专家号挂号等现实背景，研究面向复杂大群体的有限服务资源分配协调决策问题。项目主要研究成果分为如下几个方面：①复杂大规模服务对象的聚类模型设计：针对服务管理中大量消费主体评价信息聚类问题，结合模糊C均值（FCM）基本思想，提出了一种新的高斯核FCM聚类算法。②各类服务对象评价结果集结研究：基于前景理论中参考点思想，提出了确定参考点的方法，设计了大群体信息融合方法。③大规模服务对象评价一致性测度模型设计：一致性测度从评价效用值和方案排序值两个角度进行衡量，避免了单一角度所得结果的不准确性。④消费主体满意度的测量研究：以国内外目前关于满意度函数的研究为基础，构建了同一消费主体在不同分配方案下获得资源数量的满意度函数和同一方案下不同消费主体相互比较而获得的满意度函数。⑤基于满意度的资源分配决策模型：提出了基于前景理论的有限公共服务资源分配决策模型及求解算法。⑥准则值为区间数或者三角模糊数的资源分配决策模型：基于综合满意度函数表达式特点，将综合满意度函数以区间数或者三角模糊数的形式表示，保证了决策信息完整性，设计出切实有效的资源分配模型和方法。⑦群体交互式决策模型研究：设计了基于决策者信息分别集结的交互式群决策方法，有助于帮助决策者发现分歧并进行深入的交互沟通，从而提升群体决策绩效。相关研究成果有助于丰富资源分配理论，对于提高服务管理领域有限资源分配绩效具有指导意义。. 项目组认真开展研究，共发表论文(含录用)18篇，其中SCI论文3篇，EI论文7篇，CSSCI论文7篇，其他核心期刊1篇，完成了项目预定的计划。