论文部分内容阅读
在成败型产品的可靠性评估中,为提高估计精度,经常利用历史数据来确定先验分布.但在工程实际中,历史样本和样本本质上属于不同的总体,这对可靠性评估结果有着显著影响.为此,采用相似系统分析确定历史样本和样本的相似程度,将其归纳为继承因子;然后,根据历史样本信息确定产品可靠性的历史后验,基于无信息先验得到产品可靠性的更新后验;最后通过继承因子,综合历史后验和更新后验,得到产品可靠性的融合后验,并在此基础上进行可靠性推断.该方法不仅充分利用了相似产品信息,而且突出了产品的独有特性.
In the reliability assessment of success or failure products, in order to improve the accuracy of estimation, historical data are often used to determine the prior distribution, but in engineering practice, historical samples and samples belong to different totality, which has a significant effect on the reliability assessment Therefore, similar system analysis is used to determine the degree of similarity between historical samples and samples, which can be summarized as inheritance factors. Then, historical posteriori of product reliability is determined based on historical sample information, product reliability is obtained based on no prior information Update the posteriori, finally get the fusion posteriori of the product reliability through the inheritance factor, the comprehensive historical posteriori and the updated posteriori, and then make the reliability inference based on it.It not only makes full use of similar product information, but also highlights Unique product features.