利用组合方法预测火箭弹飞行坐标的精度分析  

Accuracy analysis of rocket flight coordinate prediction using a combined model

作　　者：田珂 常华俊 段鹏伟 TIAN Ke;CHANG Huajun;DUAN Pengwei(Unit 63861 of the People’s Libration Army,Baicheng 137001,China)

机构地区：[1]中国人民解放军63861部队,吉林白城137001

出　　处：《兵器装备工程学报》2022年第S02期142-150,共9页Journal of Ordnance Equipment Engineering

摘　　要：火箭弹飞行过程比较复杂,使用单一模型很难准确预测出所有的坐标数据。为了最大限度提高预测精度,把雷达已测坐标作为样本数据建立合理的模型,对数据进行重构。同时利用ARIMA模型、二次多项式回归模型和一元线性回归模型预测缺失的坐标数据。ARIMA模型适合预测射程,二次多项式回归模型适合预测高程,一元线性回归模型适合预测横偏。实验结果表明,3个模型的预测误差均小于1‰,可以作为火箭弹飞行坐标的预测模型。Testing the flight coordinates of rockets is a common test subject.Due to the narrow radar beam,when the rocket flight state is abnormal,the radar is easy to lose the target.Therefore,choosing the measured coordinates of the radar as the sample data,establishing a reasonable model to predict the missing coordinate data,and reconstructing the data have become a necessary work.However,the first mock exam is difficult to predict all the coordinate data.In order to maximize the prediction accuracy,the combination of ARIMA model,quadratic polynomial regression model and univariate linear regression model is selected to predict the missing coordinate data.ARIMA model is suitable for predicting range,quadratic polynomial regression model is suitable for predicting elevation,and univariate linear regression model is suitable for predicting lateral deviation.The experimental results show that the prediction errors of the three models are less than 1‰,which can be used as the prediction model of rocket flight coordinates.

关 键 词：飞行坐标 ARIMA模型 二次多项式回归模型 一元线性回归模型 组合模型 

分 类 号：TJ71[兵器科学与技术—武器系统与运用工程]