Parameter estimation for the weibull general log-linear model in accelerated life testing

Abstract

Uncertainty is commonly present in accelerated life test (ALT) data. The combination of a general log-linear model with a statistical distribution model—using the Weibull distribution in this paper—allows for addressing both multiple operational factors and uncertainty within a single model. This paper interprets the Weibull General Log-Linear (WGLL) model from the ALT perspective. Two parameter estimation methods—maximum likelihood estimation and least squares method—are proposed to address the parameter estimation problem. A simulation study demonstrates that both methods are effective, while the least squares method generally provides better results. The approaches proposed in this paper can be used to develop more general ALT models.