Probability of Detection models, for a specific NDE system, is very important to assess the probability at which a flaw of certain dimension can be detected. This paper proposes a fully model assisted probability of detection (MAPOD) with a Bayesian decision to classify the flaw size. Multiple correlated flaw features were integrated to the POD analysis. This was made by assigning the observation model directly to a multivariate Gaussian distribution. The mean and covariance of the distributions were empirically computed to form a Gaussian Mixture Model (GMM) to obtain the classical Hit/Miss POD curve. A risk analysis was performed on the probability of misclassification considering a 0/1 loss function. Classification decisions were made from the posterior estimates within the Bayesian framework. For the analysis, a finite element simulation study was performed for a flaw located at the sub-surface of a stainless steel (SS304) specimen.
IOS Press, Inc.
6751 Tepper Drive
Clifton, VA 20124
Tel.: +1 703 830 6300
Fax: +1 703 830 2300 firstname.lastname@example.org
(Corporate matters and books only) IOS Press c/o Accucoms US, Inc.
For North America Sales and Customer Service
West Point Commons
Lansdale PA 19446
Tel.: +1 866 855 8967
Fax: +1 215 660 5042 email@example.com