摘要
随着装备性能的提高和试验过程的复杂化,变动统计问题逐渐成为装备试验评估领域中亟待解决的重要问题,需要融合不同阶段、不同来源的异总体试验数据对产品研制、生产、使用过程中的性能变化过程进行恰当的建模与评估。首先分析了变动统计问题的主要特征,并简要回顾了变动统计问题在装备质量与性能指标评估领域中的研究历程。针对变动统计问题所面临的主要研究对象,分析了其中需要解决的若干关键问题。通过与其他几个相关研究领域的比较,提出了变动统计的基本内涵。最后,归纳提出了基本的变动统计方法:基于约束关系的多总体融合估计、基于线性模型的变动总体建模与预测和基于Bayes方法的多源验前信息融合,并对各类方法的特点与前景进行了讨论与展望。
With the advancement of equipment performance and increasing test complexity,the dynamic population statistics has become an important problem urgent to solve in the field of equipment test evaluation.Test data of diverse populations from different stages and sources should be synthesized to realize the modeling and evaluation of changing products performance during development,manufacturing and usage.Primary characteristics of dynamic population statistics are firstly analyzed,followed by a brief review of its research progress in the field of equipment quality and performance evaluation.For some central research subjects,several key problems are analyzed.Compared with other related research fields,the basic implication of the term is presented.Finally,three leading ways of dynamic population statistics constraint based multi-population integrated estimation,linear model based dynamic population forecasting and Bayesian based multi-source prior information fusion are presented,and their features and prospects are discussed and forecasted.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2010年第11期2427-2437,共11页
Journal of Astronautics
关键词
变动统计
装备试验评估
可靠性增长
武器性能
贝叶斯方法
Dynamic population statistics
Equipment test evaluation
Reliability growth
Weapon performance
Bayesian method
