基于多因子的机电设备测试性验证样本分配方案

摘要

In the testability verification test for electromechanical equipment ,the confidence level is usually low since the sample allocation method is not rational .To solve the problem,we proposed a new sample allocation method based on multi-factor by analyzing the key factors that have effect on the confidence level of conclusion .The sample size of each Unit Under Test ( UUT) was calculated out according to the relative ratio of integrated multi-factor,which could optimize the failure sampling structure and improve the representation of failure sampling set .Moreover,the single-step failure propagation algorithm was improved, and the improved algorithm can calculate failure propagation intensity more effectively .We made sample allocation by taking a certain stabilization and tracking platform as an example .The result showed that the scheme can optimize failure sampling structure and improve the confidence level of the testability verification test conclusion .%针对目前机电设备测试性验证试验中，因样本分配不合理导致试验结论置信度低的问题，通过分析影响试验结论置信度的4个关键因素即故障模式数、故障率、故障扩散强度总和与危害度，提出了基于多因子的样本分配方案。方案依据多因子综合相对比值，得出各UUT所分配的样本量，达到优化样本结构，提高样本集代表性的目的。对单步故障扩散算法进行了改进，改进后的算法可以更有效地计算故障扩散强度，并以某机电设备稳定跟踪平台为例进行了样本分配，比较分析抽样结果证明，应用该方案得到的样本结构更加合理，可以使试验结论有较好的置信度。