固体推进剂燃速测试中异常数据的判断和处理方法

摘要

为了更精准地剔除固体推进剂燃速数据的异常数据,分析了格拉布斯法则和t检验法则,指出了格拉布斯法则的缺陷.通过两种方法对相同的测试数据进行实验,结果表明,t检验法则在保持格拉布斯法则精度的基础上可以剔除格拉布斯法则不能剔除的异常数据.证明t检验法则可以更精确地筛选异常数据,并在t检验法则的基础上提出了一种新的快速收敛的筛选算法.%In order to filter the abnormal data of burning rate data of solid propellant more accurately, the Grubbs and the T-test method are analyzed and the defect of the Grubbs is presented. The two methods are experimented with the same test data. The results show that with the accuracy of the Grubbs,the T-test method can filter the abnormal data which the Grubbs can not filter. Compared with the Grubbs,the T-test method is more accurate in filtering the abnormal data. A new filtering algorithm with rapid convergent is proposed based on the T-test method.