摘要
针对当前测试性验证试验设计方案均存在确定的故障样本量较大以及验证周期长的问题,提出了一种基于仿真数据的测试性验证序贯设计方案。首先,基于二项分布模型确定序贯设计的判决阈值;然后,在序贯试验下基于给定的样本量分配和故障模式选取方法,采用仿真的方式生成相应的故障模式并进行验证试验。同时,以序贯方式计算相应的概率比,通过其取值动态判定接受/拒收。结果表明,与传统方法相比,该方法与经典验证方法具备几乎相同的抽样特性,能达到有效减少测试性验证试验所需故障样本量的目的。
Aiming at the problems of large number of fault samples and long verification period in current design schemes of testability verification test, a sequential design scheme of testability verification based on simulation data was proposed.Firstly, the judgment threshold of sequential design was determined based on the binomial distribution model. Then, based on the sample size allocation and failure mode selection method of unit failure rate under sequential test, the corresponding failure mode was generated by simulation and the verification test was carried out. The sequential method calculated the corresponding probability ratio, and dynamically accepted or rejected it by its value. The results show that compared with the traditional method, this method has almost the same sampling characteristics as the classical verification method,and can effectively reduce the amount of fault samples required for the test verification.
作者
赵学远
王康
秦亮
王帅磊
ZHAO Xueyuan;WANG Kang;QIN Liang;WANG Shuailei(Naval Aviation University,Yantai Shandong 264001,China)
出处
《海军航空工程学院学报》
2019年第1期146-150,168,共6页
Journal of Naval Aeronautical and Astronautical University
关键词
测试性验证
仿真数据
故障样本量
序贯设计
testability verification
simulation data
fault sample size
sequential design
