评估制导工具误差的Bayes估计的多层先验模型研究

针对运载火箭试验评估的特点,认为制导工具误差系数统计特性参数中,方差比较稳定,但均值可能随时间缓慢变化.由此,将制导工具误差系数的先验模型从一层先验改进为两层先验,建立了能利用跟踪数据、阵地测试数据等试验资源的多参数Bayes模型来评定制导工具系统误差.并应用经验Bayes模型改善了落点精度的评估.

Multilayer Prior Model of Bayes Method for Evaluation of Guidance Instrumentation Systematic Error

Due to high cost of test evaluation of launch vehicle, operational test programs have been declined. As far as the test and evaluation of launch vehicle in small sample size is concerned, this paper points out the mean value of the test statistical parameters of guidance instrumentation systematic error may change. Therefore, the prior distribution is improved and described by two layer prior distribution. Then a new multi parameter Bayesian method is put forward to evaluate guidance instrumentation systematic error by making full use of trajectory tracking data, test information and so on. Also the empirical Bayesian method is applied to improve the precision of evaluate the impact point error.