首页  |  期刊简介  |  期刊荣誉  |  编委会  |  现任主编  |  投稿指南  |  下载中心  |  期刊征订
中文核心期刊
Ei Compendex收录期刊
中国科学引文数据库来源期刊
中文科技期刊数据库收录期刊
国际刊号:1004-6801
国内刊号:32-1361/V
用户登录
  E-mail:  
  密  码:  
  作者 审稿  
  编辑 读者  
期刊向导
联系方式
  • 主管:中华人民共和国工业
              和信息化部
  • 主办:南京航空航天大学
              全国高校机械工程
              测试技术研究会
  • 国际刊号:1004-6801
  • 国内刊号:32-1361/V
  • 地址:南京市御道街29号
  • 电话:025-8489 3332
  • 传真:025-8489 3332
  • E-mail:qchen@nuaa.edu.cn
  • 邮编:210016
基于测试性增长的指标动态评估方法
Indicator Dynamic Evaluation Method Based on Testability Growth
  
DOI:10.16450/j.cnki.issn.1004?6801.2021.06.024
中文关键词:  测试性  指标评估  测试性增长  层次Bayes网络  Fisher精确检验  最大熵模型  Gompertz模型
英文关键词:testability  indicator evaluation  testability growth  hierarchical Bayes network  Fisher's exact test  maximun entropy model  Gompertz model
基金项目:测试性;指标评估;测试性增长;层次Bayes网络;Fisher精确检验;最大熵模型;Gompertz模型
作者单位
王康1, 史贤俊1, 聂新华1, 李祺2, 陈垚3 (1.海军航空大学310教研室 烟台264001)(2.中国航天科工集团第三研究院 北京100074)(3.91776部队 北京100161) 
摘要点击次数: 15
全文下载次数: 3
中文摘要:
      针对现有测试性指标评估方法未能考虑装备研制阶段不同层次结构测试性水平动态增长的特性,导致测试性评估置信度不高的问题,提出一种测试性增长条件下基于层次Bayes网络模型的测试性指标动态评估方法。根据装备结构特征建立测试性指标动态评估的层次Bayes网络模型,并以测试性指标作为网络传递参数;考虑延缓纠正的测试性增长试验策略,给定测试性阶段序化增长约束条件,针对不同层次节点各阶段增长试验数据,采用单边Fisher精确检验法对测试性增长趋势进行检验,并基于检验结果确定增长阶段数;提出利用最大熵模型和改进Gompertz模型的先验参数估计方法,结合Bayes定理以及研制阶段各层次节点先验信息确定节点先验分布;进一步基于层次Bayes网络融合推理算法确定顶层节点测试性指标的后验分布,实现对装备测试性指标的动态评估,并通过案例进行验证。结果表明,该方法相较于直接运用Beta分布,具备更为准确合理的指标评估结论。
英文摘要:
      Aiming at the problem that the existing testability indicator evaluation method fails to consider the characteristics of the testability dynamic growth at different levels in the equipment development stage, which leads to the low confidence of the testability evaluation, under the testability growth conditions, an indicator dynamic evaluation method based on hierarchical Bayes network model is proposed. A hierarchical Bayes network model for testability indicator evaluation could be established by equipment structural characteristics with the testability indicators as transmission parameters. When considering the testability growth test strategy for delaying correction, the testability sequential growth constraints are given. Meanwhile, the testability data of each stage at different levels of nodes are used to check the testability growth trend by Fisher's exact test method, and the number of growth stages can be determined. Then a priori parameter estimation method using the maximum entropy model and the improved Gompertz model is proposed, so that the Bayes theorem and the prior information of the nodes at each level in the development stage could be used to determine the prior distribution. Finally, the posterior distribution of the top-level node testability indicator is determined based on the hierarchical Bayes network fusion reasoning algorithm in order to realize dynamic evaluation, and an actual case is given to prove the effectiveness of the method. The results show that the method has more accurate and reasonable indicator evaluation conclusion than direct application of Beta distribution.
查看全文  查看/发表评论  下载PDF阅读器
关闭

Copyright @2010-2015《振动、测试与诊断》

地址:南京市御道街29号        邮编:210016

电话:025-8489 3332      传真:025-8489 3332       E-mail:qchen@nuaa.edu.cn

您是本站第2771764位访问者 本站今日一共被访问95

技术支持:北京勤云科技发展有限公司