FDR验证试验的故障样本分配策略研究

Research on Fault Sample Allocation Strategy for Fault Detection Rate Verification Test

摘要:

故障率作为测试性验证试验故障样本分配的主要影响因素,针对一些情况下使得故障样本分配结果的合理性不足的问题,以故障检测率(Fault Detection Rate, FDR)作为验证指标,提出了一种考虑严酷度的样本故障模式选取方法。提出了基于模糊证据推理的故障模式严酷度排序解决方法。通过对相关标准中涉及的故障样本分配策略进行梳理,针对现行多因子综合加权比例分配方法不足之处,根据故障模式种类与验证样本量的数量关系,区分不同情况,借助预选样本集随机抽样、考虑严酷度的取整策略,以及动态概率调整,合理改善了故障率主体分配方案进行故障模式选取时样本分配集中不合理的状况。以某装备单元的FDR验证试验为例,验证了所提故障样本分配方法的可行性合理性。

Abstract:

Aiming at the problem that the failure rate is the main influencing factor of failure sample allocation in testability verification test, which makes the rationality of failure sample allocation result insufficient in some cases, taking the failure detection rate as the verification index, a sample failure mode selection method considering severity is proposed. A method of fault mode severity ranking based on Fuzzy evidential reasoning is proposed. By sorting out the fault sample allocation strategies involved in relevant standards, aiming at the shortcomings of the current multi factor comprehensive weighted proportional allocation method, distinguish different situations according to the quantitative relationship between the type of fault mode and the number of verification samples, with the help of random sampling of preselected sample set, rounding strategy considering severity, and dynamic probability adjustment, It reasonably improves the unreasonable condition of sample distribution concentration when the failure rate subject allocation scheme selects the failure mode. Taking the testability verification test of a certain unit as an example, the feasibility and rationality of the proposed fault sample allocation method are verified.