系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2011年
7期
1211~1221
,共null页
共享投入型两阶段生产系统 效率测度与分解 网络数据包络分析
共享投入型兩階段生產繫統 效率測度與分解 網絡數據包絡分析
공향투입형량계단생산계통 효솔측도여분해 망락수거포락분석
two-stage production systems with shared inputs; efficiency measurement and decomposition;network data envelopment analysis
“黑箱”视角下的标准数据包络分析(DEA)因忽略中间产品转化信息及投入要素在各子过程中的配置信息,无法体现决策单元的实际生产技术,自然不能提供有效的效率绩效测度.分别在固定规模报酬和可变规模报酬假设框架下,充分利用中间产品的转化信息与投入要素的配置信息,构建了存在共享投入的关联两阶段生产系统效率的网络DEA效率测度与分解模型.建模方法遵循客观评价的原则,无需事先主观设定子效率和系统效率之间的组合关系,且可通过事后分解确立二者之间的凸线性组合关系.此外,建模方法具有很好的扩展性,可被直接用来构建含有更加复杂转化结构的生产系统的效率测度与分解模型,为多阶段投资以及多阶段供应链运作等复杂生产系统的整体效率测度与分解提供了有效的分析途径.一个关于银行投资过程的实例被用来验证建模思想的合理性和优越性.
“黑箱”視角下的標準數據包絡分析(DEA)因忽略中間產品轉化信息及投入要素在各子過程中的配置信息,無法體現決策單元的實際生產技術,自然不能提供有效的效率績效測度.分彆在固定規模報酬和可變規模報酬假設框架下,充分利用中間產品的轉化信息與投入要素的配置信息,構建瞭存在共享投入的關聯兩階段生產繫統效率的網絡DEA效率測度與分解模型.建模方法遵循客觀評價的原則,無需事先主觀設定子效率和繫統效率之間的組閤關繫,且可通過事後分解確立二者之間的凸線性組閤關繫.此外,建模方法具有很好的擴展性,可被直接用來構建含有更加複雜轉化結構的生產繫統的效率測度與分解模型,為多階段投資以及多階段供應鏈運作等複雜生產繫統的整體效率測度與分解提供瞭有效的分析途徑.一箇關于銀行投資過程的實例被用來驗證建模思想的閤理性和優越性.
“흑상”시각하적표준수거포락분석(DEA)인홀략중간산품전화신식급투입요소재각자과정중적배치신식,무법체현결책단원적실제생산기술,자연불능제공유효적효솔적효측도.분별재고정규모보수화가변규모보수가설광가하,충분이용중간산품적전화신식여투입요소적배치신식,구건료존재공향투입적관련량계단생산계통효솔적망락DEA효솔측도여분해모형.건모방법준순객관평개적원칙,무수사선주관설정자효솔화계통효솔지간적조합관계,차가통과사후분해학립이자지간적철선성조합관계.차외,건모방법구유흔호적확전성,가피직접용래구건함유경가복잡전화결구적생산계통적효솔측도여분해모형,위다계단투자이급다계단공응련운작등복잡생산계통적정체효솔측도여분해제공료유효적분석도경.일개관우은행투자과정적실례피용래험증건모사상적합이성화우월성.
The black-box measurement of the transformation efficiency of the decision-making units (DMUs) in the context of the standard data envelopment analysis (DEA) can not effectively reflect the operational performance of DMUs. The reason is that it ignores the production information embedded in the intermediate products as well as the allocation information of various inputs among the individual sub-processes and can't reflect the real underlying production technology. In this paper, under both con- stant and variable returns to scale, we have proposed the network DEA-based efficiency measurement and decomposition models of the two-stage relational production systems with shared inputs. Our models make full use of the conversion information of intermediate products as well as the configuration information of inputs among component processes. Our modeling approach needn't the pre-specified aggregation way of sub-efficiencies into the system efficiency but can build an objective convex combination relationship between them in virtue of the ex-post decomposition. Furthermore, our modeling approach is extendable, which is easily extended to the network production systems with more complex transformation structures. This study provides an efficiency analytical tool for the efficiency measurement and decomposition of such empirical complex production systems as multi-stage investment processes and supply chain operation processes. Finally, an empirical case about bank's investment process is introduced to verify our models in terms of the rationality and superiority.