摘要:
利用Wald 的序贯概率比检验中接收产品时对应的批检验数:n1 *(0 , c11), n2 *(1 , c12),, nk *(k -1 , c1k), , 设计出一种改进型序贯检验(其中当c1t s 时(t =1 , , s), 取c1t *=c1t ;当c1t s 时, 取c1t* =s):n1*(0 , c11), n2 *(1 , c12*), , ns *(s -1 , s)。证明了当次品率较小(p p0)时, 改进型序贯检验的平均抽检个数N *(p)与序贯概率比检验的N(p)比较接近, 但抽检周期要小得多。表1 参6
Abstract:
Using sequential probability ratio test n1 *(0 , c11), n2 *(1 , c12), , nk *(k -1 , c1k), , design an ameliorate sequential test (if c1t s , assume c1t = c1t ;if c1t s , assume c1t* = s):n1 *(0 , c11 *), n2 *(1 ,c12*), , ns *(s -1 , s).If p p0 , average sample number of ameliorate sepuential test and sequential probability ratio test is proved little deference , but the test time is less .