目前航空装备测试性试验存在试验样本量的确定缺少理论依据、故障检测率(FDR)单侧置信下限评估结果不稳定的问题,影响试验结果的可信度.为了使评估结果以较大的概率分布在较小的范围内,通过结合正态分布特性和单侧置信下限估计算法进行推导和计算,得到影响单侧置信下限评估的因素是试验样本量和被测产品的故障检测水平,给出了不同故障检测水平和评估结果偏差要求下试验样本量的最低取值.通过仿真分析,验证了给出的最低取值是保证评估结果稳定性的充分条件.在此基础上提出一种可以得到稳定评估结果的测试性试验方案确定方法,并通过实际案例对该方法进行验证.%To determine test plan in aviation equipment testability test,there are problems of lack of theoretical basis and instability of testability parameter evaluation results of the one-side lower confidence limits,which affects the credibility of the test results.In order to make the evaluation result distributed in a smaller scale with a large probability,the mathematical analysis and calculation were performed by combining the normal distribution characteristics and the one-side lower confidence limits estimation algorithm,which found out that the factors affecting the results of the one-side lower confidence limits evaluation were the sample size and the fault detection level of the tested product.The minimum values of sample size under different fault detection levels and deviations were given.Simulation analysis shows that the minimum values is a sufficient condition for the stability of the evaluation results.On this basis,a method for determining the test plan that can obtain stable evaluation results is proposed,and it is verified through the actual case.
展开▼
