Reliability Estimation for Hybrid System under Constant-stress Partially Accelerated Life Test with Progressively Hybrid Censoring

摘要

Background: Reliability analysis for the systems with masked data had been studied bymany scholars. However, most researches focused on a system that is either series or parallel only,and the component in the system is mainly exponential or Weibull. In engineering practice, it is oftenseen that the structure of a system is a combination of series and parallel system, and other typesof components are also used in the system. So it is important to study the reliability analysis of hybridsystems with modified Weibull components.Objective: For the hybrid system with masked data, the constant stress partial accelerated life test isperformed under type-II progressive hybrid censoring. These data from life test are used to estimateunknown parameters and reliability index of system. The research results will not only provide theoreticalbasis and reference for system reliability assessment but also favor the patents on partial acceleratedlife test.Methods: Maximum likelihood estimates of unknown parameters are investigated with the numericalmethod. The approximate confidence intervals, and bootstrap confidence intervals for parametersare constructed by the asymptotic theory and the bootstrap method, respectively.Results: Maximum likelihood estimations of unknown parameters and reliability index of systemare derived. The approximate confidence intervals and bootstrap confidence intervals for unknownparameters are proposed. The performance of estimation of unknown parameters and reliability indexare evaluated numerically through Monte Carlo method.Conclusion: The performance on maximum likelihood estimation method is effective and satisfying.For the confidence intervals of parameters, bootstrap method outperforms the approximate method.