基于NGO-BP的制导炮弹高空风场参数辨识研究  

Estimation of high-altitude wind parameters of guided projectile based on NGO-BP

作　　者：何虎 易文俊[1] HE Hu;YI Wenjun(National Key Laboratory of Transient Physics,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学瞬态物理全国重点实验室,南京210094

出　　处：《兵器装备工程学报》2024年第11期90-95,共6页Journal of Ordnance Equipment Engineering

基　　金：国家自然科学基金项目(62203191);国防科技重点实验室基金项目(2022JCJQLB06105);江苏省高等学校基础科学(自然科学)研究面上项目(22KJB590001)。

摘　　要：高空风场对制导炮弹弹道诸元有较大影响,为了提高控制精度,如何快速准确获取高空风场数据从而修正气象参数对弹道特性的影响是当前研究的热点。本文中创新性的使用北方苍鹰优化算法(NGO)优化反向传播(BP)神经网络的原始权值,对高空风场参数按照高度进行辨识研究。通过采集的高空风场参数和基于实验的弹道数据进行仿真分析,使用误差统计学评估法和仿真弹道数据结果分析两种方法对辨析出的高空风场参数精度分别进行评估,分析结果表明:利用NGO-BP神经网络法相对于反向传播神经网络法对高空风场参数辨识精度上有明显提高。The high-altitude wind field has a significant impact on the ballistic characteristics of guided shells.In order to improve control accuracy,how to quickly and accurately obtain high-altitude wind field data and correct the impact of meteorological parameters on ballistic characteristics is currently a hot research topic.This article innovatively uses the Northern Goshawk Optimization Algorithm(NGO)to optimize the original weights of the Backpropagation(BP)neural network,and conducts identification research on high-altitude wind field parameters according to height.By collecting high-altitude wind field parameters and conducting simulation analysis based on experimental trajectory data,the accuracy of the identified high-altitude wind field parameters was evaluated using two methods:error statistics evaluation method and simulation trajectory data result analysis.The analysis results showed that the accuracy of identifying high-altitude wind field parameters using the NGO BP neural network method was significantly improved compared to the backpropagation neural network method.

关 键 词：北方苍鹰优化算法 反向传播(BP)神经网络 高空风场参数 弹道分析 误差分析 

分 类 号：TJ85[兵器科学与技术—武器系统与运用工程] TP391.9[自动化与计算机技术—计算机应用技术]