Constant-partially accelerated life tests for Marshall–Olkin exponential series system with dependent masked data

Qianrao Fu,Yimin Shi
DOI: https://doi.org/10.1080/03610918.2016.1255967
2017-05-16
Abstract:ABSTRACT This paper considers the constant-partially accelerated life tests for series system products, where dependent M-O bivariate exponential distribution is assumed for the components. Based on progressive type-II censored and masked data, the maximum likelihood estimates for the parameters and acceleration factors are obtained by using the decomposition approach. In addition, this method can also be applied to the Bayes estimates, which are too complex to obtain as usual way. Finally, a Monte Carlo simulation study is carried out to verify the accuracy of the methods under different masking probabilities and censoring schemes.
What problem does this paper attempt to address?