基于过程熵的供应商多元加工能力评价

在供应链管理中,当需要考察的供应商产品质量特性多于两个时,就需要对其多元加工过程的稳定性进行评定.利用信息熵原理和多元加工过程的抽样数据,首先计算得到单一质量特性加工过程的熵及其过程能力指数,然后基于熵权法,对单变量过程能力指数进行加权处理,从而得到多元加工过程能力指数的计算模型.为了减少单次抽样所带来的随机误差,利用Bootstrap再抽样技术,对多元过程能力指数的概率分布进行仿真模拟,获得了多元能力指数的经验分布及其大致的置信区间,为更有效地评价供应商多元质量特性加工能力提供了概率依据.最后以某汽车集团的曲轴供应过程过例,对三个候选供应商的多元加工过程进行评价和选择.

Evaluation on the Multivariate Process Capability of Suppliers Based on the Entropy of the Process

In the supply chain management,when there are more than two quality characteristics need to be considered for the supplier's products,it involves the evaluation of the multivariate process stability of the suppliers.This paper firstly made use of the information entropy principle and the sampling data from multivariate process of the supplier,to compute the process entropy of each quality characteristic and its process capability index.Then the multivariate process capability index(MPCI) was deduced by synthesizing the univariate process capability indices with the entropy method.To avoid the sampling errors of single sampling process,the stochastic distribution of the MPCI was simulated based on the bootstrap sampling technique,with the approximate confidence intervals of the MPCI also obtained,which could provide probability bases for more efficient evaluation on the multivariate process of the suppliers.Finally an example of axle supply process was taken for analysis,which involved the evaluation and selection among three candidates.