逐步Ⅰ型混合截尾下逆Lomax分布竞争失效产品的统计分析  

Statistical Analysis of Competing Failure Products for Inverse Lomax Distribution under Progressive Type-ⅠHybrid Censoring

作　　者：韩荣 蔡静 何剑 何飞 HAN Rong;CAI Jing;HE Jian;HE Fei(School of Data Science and Information Engineering,Guizhou Minzu University,Guiyang 550025,China)

机构地区：[1]贵州民族大学数据科学与信息工程学院,贵州贵阳550025

出　　处：《现代信息科技》2025年第1期76-81,87,共7页Modern Information Technology

基　　金：国家自然科学基金项目(11901134)。

摘　　要：在混合截尾样本数据下,对逆Lomax分布竞争失效产品的统计分析问题进行了研究。首先,基于逆Lomax分布竞争失效产品,利用极大似然理论,推导出未知参数的极大估计;并利用渐近似然理论确定未知参数的近似置信区间。其次,通过设定无信息先验为未知参数的先验分布,采用MH抽样算法求出参数的Bayes估计和HPD可信区间。最后,通过Monte Carlo模拟,计算出参数的均方误差(MSE)、平均绝对偏差(MAB)、平均区间长度(AL)以及覆盖率,并对两种估计方法进行了对比。实验结果表明,贝叶斯估计优于极大似然估计,在相同置信度下,基于Bayes估计的HPD平均可信区间长度优于MLE的近似置信区间平均区间长度。Under the Hybrid Censoring sample data,this paper studies the statistical analysis problem of competing failure products for inverse Lomax distribution.Firstly,based on the competing failure products for inverse Lomax distribution,it utilizes maximum likelihood theory to obtain the Maximum Likelihood Estimation of unknown parameters,and uses asymptotic likelihood theory to derive the Approximate Confidence Interval of unknown parameters.Secondly,by setting the uninformative prior as the prior distribution of the unknown parameters,the Bayes estimation and HPD Confidence Interval of the parameters are obtained by using the MH sampling algorithm.Finally,through Monte Carlo simulation,the Mean Squared Error(MSE),Mean Absolute Bias(MAB),Average Interval Length(AL),and coverage percentage of the parameters are calculated,and the two estimation methods are compared.The experimental results show that Bayes Estimation is superior to Maximum Likelihood Estimation,and under the same confidence level,the average confidence interval length of HPD based on Bayes estimation is better than the approximate confidence interval average interval length of MLE.

关 键 词：竞争失效 逆Lomax分布 极大似然估计 贝叶斯估计 MH抽样算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] O213.2[自动化与计算机技术—计算机科学与技术]