Abstract
Optimal inspection schemes based on dispersed nonconformity count data are derived to properly discriminate between satisfactory and unsatisfactory batches. The Conway-Maxwell-Poisson distribution is adopted to describe the stochastic behavior of the number of nonconformities per sampled unit. A mixed integer nonlinear programming problem is stated in order to determine the lot sampling plan with a minimal sample size and limited producer and consumer risks. Explicit approximations of the smallest number of units to be tested per lot and the maximum tolerable nonconformity score are presented. A Monte Carlo simulation approach is then used to find the most powerful decision rule for lot disposition. In case of over-dispersion, the suggested perspective allows the practitioners to greatly reduce the required sample size for lot sentencing. The developed methodology is applied to the manufacturing of glass for illustrative and comparative purposes.
March 31, 2017 at 09:27PM
http://ift.tt/2nnoJOC
from
http://ift.tt/2nnoJOC
No comments:
Post a Comment