基于Beta分布的最经济过程均值及检验规格限的确定  

Determining the Most Economic Process Mean and Inspection Specification Limits Based on a Beta Distribution

作　　者：张斌[1] 王海燕[2] 曹春正[1] 韩之俊[3] 

机构地区：[1]南京信息工程大学数理学院,江苏南京210044 [2]南京财经大学管理学院,江苏南京210046 [3]南京理工大学经济管理学院,江苏南京210096

出　　处：《系统工程》2008年第6期78-82,共5页Systems Engineering

基　　金：国家自然科学基金资助项目(70572024);江苏省"六大人才高峰"资助项目(06-D-036)

摘　　要：100%检验是对生产方和顾客都提供保护的一种检验方法,在工业生产或过程决策中起到非常重要的作用。考虑许多实际过程输出分布非正态的情况,本文以B eta分布作为过程输出分布,结合不对称二次质量损失函数,以生产方和顾客整个系统的损失最小为前提,提出最优检验策略。在全检决策中,建立了确定最优检验规格限及过程均值的经济模型。灵敏度分析反映了各参数对检验方案、最优过程均值确定及总损失的影响。数值计算结果说明本文所提模型对改进过程输出质量是有效的。100% inspection is a means of providing protection for both the producer and the consumer. It plays an important role in some industrial or other decision-making processes. In view of the non-normal distribution in practical output processes and on the premise of the minimum loss of the producer and the customer in the whole system, this paper treats a Beta distribution as a process output distribution and develops an optimal inspection strategy by using the asymmetric quadratic quality loss function. In the case of complete inspection decision-making, we propose an optimal model to deter- mine the optimal specification limits and process mean. The analysis of associated sensitivity shows that all the parameters have an effect on inspection plan, the determination of optimal process mean and the overall loss. Moreover, a numerical example illustrates the proposed models are effective at improving the quality of process output.

关 键 词：质量损失函数Beta分布 检验规格限 过程均值 

分 类 号：O213[理学—概率论与数理统计]