Modeling Accelerated Life Tests with Dependent Failure Modes
In many life tests, it is not uncommon to observe multiple dependent failure modes. Since a full understanding of the underlying mechanism of such dependence can be extremely difficult, adequate physical models are rarely available. Hence, this research investigates the statistical modeling for constant-stress ALT with multiple dependent failure modes. A statistical ALT model is firstly constructed. The stochastic dependence between different failure modes of a testing unit is established by a Gamma frailty model. It is shown that the proposed model incorporates the ALT model with independent failure modes as its limiting case, and thereby, a hypothesis testing for dependence is naturally motivated.
Accelerated Life Test Design of Experiments Competing Risks Frailty
XIAO LIU
Mechanical and Industrial Engineering, Qatar University, Doha, Qatar Industrial and Systems Engineering, Rutgers University, Piscataway, NJ, USA
国际会议
北京
英文
195-200
2011-06-20(万方平台首次上网日期,不代表论文的发表时间)