For the gradual maturity of Bayesian survival analysis theory,as well as the defects of the traditional methods for storage reliability evaluation,the Bayesian survival analysis method is proposed to build regression models for reliability in the random truncated test.These models can reflect the influences of different environments on the ammunition storage lifetime.As an example,the common exponential distribution is used here,and Markov chain Monte Carlo(MCMC)method based on Gibbs sampling dynamically simulates the Markov chain of the parameters' posterior distribution.Also,the parameters' Bayesian estimations are calculated in the random truncated condition.The simulation results show that the proposed method is effective and directly perceived.
针对ARMA模型建模过程中模型识别和参数估计易受观测值异常点影响问题,构建了同时考虑加性异常点和更新性异常点的ARMA模型.运用基于Gibbs抽样的Markov Chain Monte Carlo贝叶斯方法,估计稳健ARMA模型参数,同步确定观测值中异常点的位置,辨别异常点类型.并利用我国人口自然增长数据进行仿真分析,研究结果表明:贝叶斯方法能够有效地识别ARMA序列的异常点.
