PROBABILISTIC METHODS OF ZERO-DEFECTS CHECKING FOR NONHOMOGENEOUS SAMPLINGS

Abstract

A short outline of the modified probabilistic methods of zero-defects checking is given and the applicability of these methods for nonhomogeneous sampling is substantiated. The main task of this paper is  to  substantiate  the  proposed  modification  of  the  zero-defect  checking  for  obviously  heterogeneous sampling. The proposed modification of the classical zero-defects methods allows: essentially decrease required volumes of testing; reduce testing terms in comparison with repeated fulfillments of the classical zero-defects method of tests; visibly declined testing costs. The proposed modifications of the classical zero-defects  methods  are  fitted  to  work  for  all  cases  of  essentially  inhomogeneous  samples. Given substantiation  of  the  proposed  method  may  be  essentially used  also  for  the  modification  of  Bernoulli binomial testing scheme to reconcile it with the real situation in the accepting sampling. Our methods as a classical ones are based on application of Bernoulli's binominal testing scheme.