| 李昊,徐如远,朱耀,张生鹏,荣双龙,黄硕,祝文羲.威布尔分布场景下环境数据折算融合方法[J].装备环境工程,2025,22(5):42-49. LI Hao,XU Ruyuan,ZHU Yao,ZHANG Shengpeng,RONG Shuanglong,HUANG Shuo,ZHU Wenxi.Environmental Data Conversion and Fusion Method under the Weibull Distribution Scenario[J].Equipment Environmental Engineering,2025,22(5):42-49. |
| 威布尔分布场景下环境数据折算融合方法 |
| Environmental Data Conversion and Fusion Method under the Weibull Distribution Scenario |
| 投稿时间：2025-03-28 修订日期：2025-05-17 |
| DOI：10.7643/issn.1672-9242.2025.05.007 |
| 中文关键词: 环境数据折合 威布尔分布 寿命评估 最优线性无偏估计 Gehan-Wilcoxon检验 平均秩次法中图分类号:TB114.3 文献标志码:A 文章编号:1672-9242(2025)05-0042-08 |
| 英文关键词:environmental data conversion Weibull distribution life evaluation optimal linear unbiased estimation Gehan-Wilcoxon test average rank method |
| 基金项目:技术基础科研项目(JSZL2021601B001) |
|
|
| Author | Institution |
| LI Hao | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| XU Ruyuan | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| ZHU Yao | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| ZHANG Shengpeng | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| RONG Shuanglong | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| HUANG Shuo | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
| ZHU Wenxi | Aerospace Science & Industry Corp Defense Technology R&T Center, Beijing 100854, China |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 对威布尔分布场景下不同环境数据进行折合与融合，实现产品寿命可综合评价。方法 基于Nelson假设，构建威布尔分布场景下不同环境试验数据折合一般关系式，并采用最优线性无偏估计法对不同环境下的分布参数进行评估，并利用Γ函数最优线性无偏估计方法适用范围进行扩充，最终基于平均秩次法与扩充后的最优线性无偏估计法对折合与融合后的环境数据进行综合评价。以某产品在不同环境下定时截尾试验数据为示例，并采用Gehan-Wilcoxon检验方法对数据折算效果进行量化对比。结果 案例数据的结果显示，按照传统方法，仅采用威布尔分布尺度参数比值作为折合系数得到的折算数据与目标环境原始数据间的Gehan-Wilcoxon检验统计量|Z|为0.880 3，而本文所提方法得到的折算数据与目标环境原始数据间的Gehan-Wilcoxon检验统计量|Z|为0.344 5，表明后者得到的折算数据与目标环境原始数据的相似程度更高，说明本文所提方法的折合效果更好。结论 小样本场景下不同环境试验数据的形状参数估计结果会存在一定差异，采用数据折合关系的一般形式一定程度上能降低小样本场景对数据折算效果的影响。 |
| 英文摘要: |
| The work aims to achieve the comprehensive evaluation of product life in a Weibull distribution scenario through conversion and fusion of different environmental data. Based on the Nelson assumption, a general formula for the conversion of different environmental test data under a Weibull distribution scenario was constructed. The optimal linear unbiased estimation method was used to evaluate the distribution parameters under different environments. The applicability of the optimal linear unbiased estimation method was expanded by the Γ function. Ultimately, the comprehensive evaluation of the converted and fused environmental data was performed with the mean rank method and the expanded optimal linear unbiased estimation method. Finally, with the time-censored test data of a product under different environments as an example, the Gehan-Wilcoxon test method was used to quantitatively compare the data conversion effect. The results of the case data showed that according to the traditional method, the Gehan-Wilcoxon test statistic |Z| between the converted data obtained with only the ratio of the Weibull distribution scale parameters as the conversion coefficient and the original data of the target environment was 0.880 3. In contrast, the Gehan-Wilcoxon test statistic |Z| between the converted data obtained with the method proposed in this work and the original data of the target environment was 0.344 5, indicating that the converted data obtained by the latter was more similar to the original data of the target environment, demonstrating that the conversion effect of the method proposed in this work was better. Under small sample scenarios, there will be certain differences in the estimation results of the shape parameters of different environmental test data. Using the general form of the data conversion relationship can reduce the impact of small sample scenarios on the data conversion effect to a certain extent. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
