多部门机构下的生产规划与资源配置

From the viewpoint of globalization, service industry is becoming the backbone for promoting the economic and social development. In recent years, research on service science has drawn great academic attentions. In real-world operations, many service suppliers, such as financial organizations or supermarket chains, consist of a group of homogeneous decision making units (DMUs) that consume the same set of resources to produce the same set of products. As a strong support for decision making, it always attracts management in how to improve performance and optimize operation patterns based upon efficiency evaluation. This study mainly involves two aspects of resource reallocation and production planning, specifically including an equitable allocation of fixed costs and available resources, and the formulation of output goals which makes a full use of the current production capacity. This study intends to combine the basic ideas from game theory and time series analysis with data envelopment analysis (DEA), in order to develop programming models fitting real settings that optimize input-output patterns and equitably allocate fixed costs, available resources, and output targets. During the above process, the following will be highly considered, including the structural features and competing modes within a multi-level network systems, as well as special input/output factors.

从全球化视野出发，服务产业日益成为推动经济发展、产业升级、社会进步和民生改善的主力军，对于服务科学领域的相关研究也得到了广泛的关注。在实际运作过程中，很多服务产业的提供者如金融机构、连锁超市等由一些同质的决策单元来构成其运营体系，消耗同种类的投入资源以制造同种类的终端产品。作为管理决策的有力支持，如何在效率评价的基础上改善绩效表现与优化运营模式是管理者的兴趣重点，在整个服务业都具有广泛的应用背景。主要涉及资源配置和生产规划问题，具体包括固定成本的合理分摊、现有资源的优化再分配、以及充分发挥产能的产量目标制定等。本项目以数据包络分析为基本研究方法，借鉴博弈论和时间序列分析等领域的基本思想，结合多层次网络系统的结构特征和竞争模式，考虑特定投入/产出指标的特殊性，建立契合实际应用环境的规划模型，以实现评价和优化投入-产出模式、合理设计成本分摊机制、优化投入资源配置和规划产出水平的目的。

从全球化视野出发，服务产业日益成为推动经济发展、产业升级、社会进步和民生改善的主力军，对于服务科学领域的相关研究也得到了广泛的关注。作为管理决策的有力支持，如何在效率评价的基础上改善绩效表现与优化运营模式是管理者的兴趣重点，在整个服务业都具有广泛的应用背景。本项目在数据包络分析的理论框架下，研究了具有非常规投入产出指标组合的效率测度和绩效改进问题，分别建立针对有界离散型数据和与量纲无关的指标权重约束下的评价模型，有效拓展了数据包络分析方法的适用场景。结合多层次网络系统的结构特征和竞争模式，解决了并型结构下整体和子阶段效率的测度问题，并将其进一步推广至子阶段拥有差异化投入产出指标集的一般情形，打破传统先决条件、极大拓展了应用范围。在效率分析理论框架下，基于方向性距离函数提出一种将全国碳减排总目标在省际进行合理分摊的交互式迭代算法；同时对Malmquist-Luenberger生产率指数测算模型进行改善，解决了M-L指数的无解问题，运用于展现1999至2012年间我国各省份在环境因素制约下经济增长绩效的动态发展趋势，为实现资源优化配置提供科学的决策支持。