The Application of a Maximum Likelihood Approach to an Accelerated Life Testing with an Underlying Three-Parameter Weibull Model
Volume 4, Number 3, July 2008 - Paper 4 - pp. 233 - 241
DANIEL I. DE SOUZA, JR.Fluminense Fed. University, Civil Engineering Dept., Grad. Program, Niterói, RJ, Brazil & North Fluminense State University, Industrial Engineering Dept., Campos, RJ, Brazil
(Received on November 28, 2006)
Accelerated life testing provides timely information about the life distribution of products, components and materials. Accelerated testing is attained by subjecting the products or parts to testing conditions much higher than these products or parts are likely to experience under normal use conditions. The three-parameter Weibull model is commonly used to represent the underlying failure-time distribution for electronic products, in which the minimum lives of these products are different from zero. The standard maximum likelihood method for estimating the parameters of the three parameter Weibull model can have problems since the regularity conditions are not met (see Murthy et al. ; Blischke ; Zanakis and Kyparisis ). In this study, we will develop an accelerated life testing model in which the underlying sampling distribution is the three-parameter Weibull model. We will be assuming a linear acceleration condition. To estimate the shape, scale and minimum life of the three-parameter Weibull model we will use a maximum likelihood approach for censored failure data. To overcome the "no-regularity" problem resulting from the above mentioned, we will apply a modification proposed by Cohen et al. . An example will illustrate the application of the proposed accelerated life testing model.
Click here to download the paper.
Please note : You will need Adobe Acrobat viewer to view the full articles.