基于SMT方法的测试性验证试验方案设计

Design of Testability Verification Test Scheme Based on SMT Method

针对现有导弹测试性验证试验方案确定的故障样本量较大以及序贯类试验方案最大样本量上界无法确定的问题,提出一种基于序贯网图检验(SMT)方法的测试性验证试验方案。通过对序贯概率比检验(SPRT)的检验问题进行拆分,在确定检验点数目的基础上,提出一种检验点取值与最大样本量的优化求解方法。同时考虑最大样本量较大的情形,基于承制方风险和使用方风险设计了SMT方法的截尾策略。通过案例验证与经典验证方法以及SPRT方法进行对比分析。结果表明,该方法能控制验证试验的样本量且所确定的平均样本量也优于其他方法,导弹装备的测试性验证试验的实施具备实际指导意义。

In order to solve the problem that the fault sample size determined by the existing missile testability verification test scheme is large and the maximum sample size of the sequential test plan cannot be determined,a testability verification test based on sequential mesh test(SMT)method is proposed.Firstly,an optimal solution method for checkpoint value and maximum sample size was proposed by splitting the test problem of sequential probability ratio test(SPRT)and determining the number of checkpoints.Secondly,considering the case of large sample size,the censored strategy of SMT method was designed based on the risk of both the contractor and the user.Finally,the case verification was compared with the classical verification method and the SPRT method.The results show that the method can control the sample size of the verification test and the determined average sample size is better than other methods,and the method has practical guiding significance for the implementation of the testability of missile equipment.