小波分析在外测数据滤波中的应用研究

为提高弹道估计精度,提出了一种基于小波分析的滤波方法,滤除外测数据中AR（自回归）模型的随机误差。分析讨论了滤波过程中的几个关键问题,提出了利用基于偏自相关系数截尾检验的方法来确定分解层数,然后采用GCV（广义交叉确认）准则来确定均方差意义下最优阈值的方法。本文提出的外测数据滤波方法计算简单,不需要估计噪声的方差。仿真结果表明,该方法能有效滤除外测数据中的AR噪声。

Research on Application of Wavelet Analysis in Tracking Data Filter

A filter method based on wavelet analysis is proposed to reduce the random deviation of tracking data from automatic regression（AR） model to improve the accuracy of trajectory estimation.Key issues in filtering are discussed.The number of analytical layers is defined with a method based on partial self-correlated coefficient censored verification and the optimal threshold value is defined with Generalized Cross Validation（GCV）.Computation of the tracking data filter is simplified as estimation of the variance of noise is no longer necessary.As shown by the result of simulation,the method effectively filters the noise of the AR model.