基于随机加权的捷联惯组历次测试数据验前分布研究

摘要

针对捷联惯性测量组合(捷联惯组)历次测试数据小样本的特点,在总体分布参数形式未知的情况下,根据已有的先验信息,提出了通过随机加权法,获得捷联惯组历次测试数据总体参数的验前分布.结合当前样本信息,利用贝叶斯方法得到捷联惯组历次测试数据的验后分布.将统计推断建立在验后分布基础之上,可以减 小小样本情况下的统计分析误差.%Aimming at the small sample of strapdown inertial measurement unit (SIMU) successive test data, a random weighting method is used to get the prior distribution function of population's parameter with the known prior information under the situation that the parameters' format of general distribution is unknown. Bayesian method is adopted to determine the posterior distribution by using the prior information, population information and sample information. The illustration shows that this method can greatly improve the accuracy of statistic inference, and overcome the shortage of small sample size.