Fitting the Weibull log-linear model to accelerated life-test data

摘要

The Weibull log-linear model is a widely-used acceleratednlife-test model in reliability engineering. The standard deviation ofnlog(life), s, was often assumed to be a constant or stress-independent.nHowever, theoretical and experimental research results suggest that, innmany cases, s is stress-dependent. The data analysis via the MLE methodnmust be performed numerically, because of the complexity of the modelnand many unknown parameters being involved. The commonly-used methodsnoften fail to converge when the starting point is not close to thensolution, especially for censored data. Generally, no easy-to-usensoftware is available for the Weibull log-linear model. To facilitatenthis process, an efficient algorithm is presented in this paper, tonobtain the MLE of the model parameters from test data (with or withoutncensoring) for both stress-independent and stress-dependent models. Thenvalidity and effectiveness of this procedure are illustrated withnnumerical examples. The method is numerically stable, and easy tonimplement and program