摘要
针对目前复杂装备测试性评估中的小子样问题,在对经典评估方法和传统Bayes评估方法进行分析的基础上,提出了一种基于多源信息融合的测试性评估方法。该方法综合考虑了多源信息的可靠度和可信度,较为合理地解决了多源信息对复杂装备测试性水平反映的可靠程度问题,以及融合求解后验分布时,因多源信息和现场试验信息的异总体特性而导致的小子样现场试验信息被淹没的问题。以某型火控系统测试性评估为例进行了研究,并与经典评估方法、传统Bayes评估方法进行了比较。结果表明,基于多源信息融合的测试性评估方法所得到的结果更为合理,在处理现场试验信息为小子样的问题时,更具优势。
Aiming at the problems of classical evaluation methods and traditional Bayesian evaluation method,a testability evaluation method was proposed based on multi-source information fusions for small sample problem of current complex equipment testability evaluations.The reliability and credibility of multi-source informations were comprehensively considered,and the problems of the reliability of multi-source informations on the testing level of complex equipment were solved,as well as the problems that small sample field test informations were submerged due to different characteristics of multi-source informations and field test informations when the fusions were solved after the posterior distributions.Finally,the testability evaluation of a certain type of fire control system was taken as an example,and compared with the classical evaluation method and the traditional Bayesian evaluation method.The results show that the testability evaluation method based on mult-source information fusions is more reasonable and has more advantages when dealing with field test informations for small samples.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2017年第23期2875-2879,共5页
China Mechanical Engineering
关键词
测试性评估
小子样
多源信息
可靠度
可信度
testability evaluation
small sample
multi-source information
reliability
credibility
