一种弹道外测数据的野值处理方法  

A Data Processing Method on Outliers for Trajectory Tracking Measurement

作　　者：李坤坤 曹锐 胡亚雄 LI Kun-kun;CAO Rui;HU Ya-xiong(Northwest Institute of Nuclear Technology,Xi’an 710024,China)

机构地区：[1]西北核技术研究院,西安710024

出　　处：《火力与指挥控制》2021年第6期145-148,共4页Fire Control & Command Control

摘　　要：为抑制弹道外测数据中异常值的影响,设计了一种基于最小二乘趋势项建模与一阶差分还原的联合野值处理方法。该方法基于最小二乘原理建立了外测数据趋势项多项式的一般模型;利用一阶差分处理化斑点型野值为孤立型野值,在简化野值类型的同时又降低了野值的处理难度;最后引入肖维涅准则进行野值检验剔除,并积分还原得到便于后续分析的外测数据。仿真算例结果对比表明,该方法简便易行,在不破坏原始数据有用信息前提下,可抑制复杂类型野值的影响,取得较好的处理效果。In order to suppress the effects of outliers in the trajectory tracking measurement,a joint data processing method based on curve fitting trend by the least square and first order difference is designed.The method firstly establishes a polynomial model for the trend of the measurements based on the least square principle.Then it uses a method of first-order difference to process the trend-removed measurements,which simplifies the outlier types and decreases process difficulty degree of outliers.Finally,the chauvigne criterion is introduced to eliminate outliers from the differenced data,and get data for future analysis.The simulation results show the convenience of implementation and effectiveness of the method.

关 键 词：外测数据 野值 趋势项 最小二乘 差分 

分 类 号：TJ013.2[兵器科学与技术—兵器发射理论与技术]