An Entropy Measure of Flow Dominance for Predicting Operations Performance
提出一种基于熵的流程主导度度量（EFD），将系统中所有产品的路径信息整合为0-1之间的单一指标，实验表明该指标能显著预测制造系统的作业流时间、流时间标准差和在制品库存等绩效指标。
The flow of jobs within a system is an important operating characteristic that influences system performance. While the majority of previous studies on manufacturing performance consider product flows only as an implicit parameter of the design, we introduce an explicit measure of flow dominance based on entropy and test its efficacy in predicting the performance of manufacturing systems. In computing entropy flow dominance ( EFD ), we aggregate information embedded in the routings of all products within a system into a single measure. EFD is designed to indicate on a 0–1 scale the level of flow dominance, where 1 represents a pure flow shop and 0 represents a pure job shop. The result is a simple measure that provides managers a way to explain and predict complex phenomena. Our experimental results indicate that EFD is a statistically significant determinant of manufacturing system performance. Furthermore, the model including EFD as an independent variable accurately predicts manufacturing system performance as measured by job flow time, flow time standard deviation, and work in process. We note that the same results can also apply to service systems, such as the “back‐room” low‐contact type systems, that have similar characteristics as manufacturing systems.