外弹道测量（简称外测）事后处理的原始数据中包含着由弹上、地面测量设备等产生的多种多样的系统误差和随机误差，外测数据质量与精度的实质就是对其噪声的准确分离。文中针对原始数据的特征，建立对其小波包分解后的各高频门限权值融合，并利用要求的数据精度值控制权函数的噪声分离模型，得到σ控制的数据和各种噪声的分离。实际应用表明，此方法对多类任务数据的处理都取得了满意的结果。

In the original data of trajectory measurement,there exist all kinds of system error and random error that derive from missile-body,devices,as well as air environment.The data's quality and accuracy depend on the precise noise departure closely.Aiming at these characters of original data,this paper upbuilds a kind of noise departure model via wavelet packet and soft threshold syncretizing,meanwhile,the relevant scaled functions are controlled by the required accuracy value so that we can obtain the needed accuracy and noise.The application in practice proves that the method is valid and satisfactory in multi-kind tasks.

TN957.51