基于Bootstrap方法的捷联惯组历次测试数据验前分布研究

The Prior Distribution Study of SIMU Successive Test Data Based on Bootstrap Method

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Bootstrap重采样技术,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,给出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。

Aiming at the small sample of strapdown inertial measurement unit （SIMU） successive test data, bootstrap method is used to get the prior distribution of population parameter of SIMU successive test data with the known prior information, when the parameter format of population distribution is unknown. According to the current sample information, posterior distribution is presented. Some statistical inferences can be made with the posterior distribution. It can reduce the error of statistical analysis in the case of small sample.