考虑动态时滞效应的两阶段网络DEA模型

A Two-Stage Network DEA Model Considering Dynamic Time Lag Effects

时滞效应广泛存在于两阶段生产的投入产出活动中。现有的估算时滞效应的方法不能反映各个决策单元(DMU)在时间跨度内每个阶段所采取的各种策略。本文建立了一个新的数据包络模型,通过加入一个可变时滞因子来评估具有时滞效应的两阶段动态网络系统的性能,以优化系统整体效率的方式对每个DMU进行重新建模。通过一个算例验证了现有方法由于其局限性在一定程度上低估了系统的整体效率测量,而提出的模型解决了这一问题,并能很好地表现了资源存量的动态变化,从宏观角度确定了资源的虚拟再分配,可以为动态网络系统中的资源合理配置提供解决方案,优化整体效率。

Time lag effects are widely present in the input-output activities of two-stage production. Existing treatments for estimating time lag effects do not reflect the various strategies deployed by individual decision-making units (DMUs) at each stage over a time horizon. In this paper, a new data envelopment model is developed to evaluate the performance of the two-stage dynamic network system with the time lag effect by incorporating a variable time lag factor. Individual DMUS is reprogrammed in a way that optimizes the overall efficiency of the system. It is verified that the existing methods underestimate the overall system efficiency measurement to some extent due to their limitations by an arithmetic example. However, the proposed model solves this problem and represents the dynamics of the resource stock very well. The proposed model identifies the virtual redistribution of resources from a macro perspective, which can provide solutions for the rational allocation of resources in dynamic network systems and optimize overall efficiency.