基于动态门限与自适应插值的外引导平滑算法

A range external data smoothing algorithm based on dynamic threshold and adaptive interpolation

为了实现靶场光电经纬仪外引导数据的实时平滑,确保外引导数据驱动下光电经纬仪稳定获取图像,提出基于动态门限的野值处理算法和基于自适应插值的处理算法。对于外引导数据中的野值,提出一种利用影响函数动态计算样本方差的方式,构建动态判别门限实时完成野值的处理;对于外引导数据的插值处理,判断插值计算的连贯性,并将外引导的“卡顿”数据进行分类,采取不同的策略实时自适应插值。实验结果表明:基于动态门限的五点外推野值剔除方法,野值检测率平均在80%以上,同时虚警率较低;插值处理算法能同时应对卡顿与非卡顿的外引导数据,处理结果平滑连贯。算法已成功应用于海军某观测站的光电经纬仪,满足了靶场弹道测量系统的需要。

In order to realize the real-time smoothing of the external guidance data of the photoelectric theodolite and ensure the stable image acquisition by the photoelectric theodolite driven by the external guidance data,this paper proposes an outfield processing method based on the dynamic threshold and an adaptive interpolation processing algorithm.Firstly,for outliers in the external guidance data,a method of dynamically calculating sample variance using influence function is proposed,and a dynamic discriminant threshold is constructed to process outliers in real time.Secondly,for the interpolation processing of the external guide data,the coherence of the interpolation calculation is judged,and the external guide‘stuck’data is classified,and different strategies are adopted,so as to adapt the real-time interpolation according to the data situation.The experimental results show that the average outlier detection rate is over 80%and the false alarm rate is low by the five-point extrapolation outlier elimination method based on dynamic threshold.The interpolation processing algorithm can deal with both stuck and non-stuck external guidance data,and the processing results are smooth and coherent.At present,the algorithm has been successfully applied to the main control computer of the photoelectric theodolite in a naval observation station,which meets the needs of the ballistic measurement system of the shooting range.