The inverse Gaussian (IG) process is recently proposed as a flexible family of models for degradation modeling. This paper investigates Bayesian analysis of IG process model with random effects for degradation modeling. Novel features of Bayesian analysis are the natural manners for incorporating subjective information and pooling of random effects information between test specimens. An IG process model with random effects named as the random drift IG process model is investigated using the Bayesian method. In addition, a Bayesian χ2 goodness-of-fit test is developed for this Bayesian analysis. The applicability of the Bayesian method for degradation analysis with this IG process model is demonstrated with a classic example.

This content is only available via PDF.
You do not currently have access to this content.