scientific article; zbMATH DE number 4003317

From MaRDI portal
Publication:4728070





[For the entire collection see Zbl 0615.00011.] Statistical inferences on the durability of a product may often have to be based on an analysis of failure data generated under an overstress or accelerated life test (ALT). The effectiveness of such inferences rests heavily on the validity of model assumptions concerning the life distribution and the effect of stress acceleration. In this article, the principal methodological approaches to ALT analysis are reviewed in light of plausibility of the model, flexibility of empirical fit and usefulness in practical application. These include parametric log-linear models, semi-parametric formulations based on proportional hazards or time transformation, and a reciprocal- linear regression model in the setting of a Brownian motion process for damage growth. Some theoretical considerations and practical issues of designing an ALT experiment are also discussed.




Cited in
(22)








This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4728070)