基于MCMC稳态模拟的Weibull回归模型及其可靠性应用

讨论了贝叶斯加速失效模型族中应用最为广泛的Weibull回归模型，提出针对寿命服从weibull分布的产品，运用基于Gibbs抽样的MCMC方法动态模拟出参数后验分布的马尔可夫链，在失效率的先验分布为Gamma分布时，给出随机截尾条件下参数在Weibuil回归模型中的贝叶斯估计，提高了计算的精度。借助数据仿真说明了利用BUGS软件包进行建模分析的过程，证明了该模型在可靠性应用中的直观性与有效性。

Weibull Regression Model Based on MCMC and Its Application in Reliability

Weibull regression model was discussed,which is used widely in the family of Bayesian accelerated failure-time models.As for the productions whose life distributions belong to Weibull distribution,the MCMC method was brought forward based on Gibbs sampling to simulate dynamically the Markov Chain of the parameters' posterior distribution.From this,the parameters' Bayesian estimation of the Weibull regression model was given in the condition of the random truncated test and when the prior distribution of the failure rate belonged to the Gamma distribution,which improved the precision of the numeration.Also the data's simulation was utilized to show the process of setting the model by using the BUGS package.It proves the objectivity and validity of the model.