关键词: 运筹学, 效率分析, 网络数据包络分析, 外源投入型两阶段制造系统

Abstract: Manufacturing process evaluation is an important part to improve the efficiency of the manufacturing system. Data Envelopment Analysis(DEA)is an effective tool to measure the relative efficiencies of the manufacturing system. However, the traditional DEA models ignore the production information embedded in the intermediate products as well as the allocation information of various inputs among the individual sub-process and can't reflect the real underlying production technology. To solve this problem, we have proposed a DEA-efficiency measurement and decomposition models of two-stage manufacturing system with exogenous inputs under both constant and variable returns to scale. The modeling method does not need the pre-specified aggregation way of sub-effciencies into the system efficiency. The evaluation results for the iron&steel manufacturing system show that the model can do more to find inefficient decision-making units(DMUs)and possible potential for performance improvement than the traditional approaches, and help decision makers to detect the origination of inefficiency for those inefficient DMUs. The numerical example proves the reasonability of our method. This study provides an analytical method for the efficiency measurement and decomposition of such complex manufacturing system.

Key words: operational research, efficiency analysis, network data envelopment analysis, two-stage manufacturing system with exogenous inputs