基于广义回归神经网络插值的雷达引信回波模拟  

Radar Fuse Echo Simulation Based on Generalized Regression Neural Network Interpolation

作　　者：王洋洋 曹菲 WANG Yangyang;CAO Fei(Rocket Force University of Engineering,Xi’an 710025,Shaanxi)

机构地区：[1]火箭军工程大学,陕西西安710025

出　　处：《火箭军工程大学学报》2024年第5期69-80,共12页Journal of Rocket Force University of Engineering

摘　　要：为了更加精确地对雷达引信回波信号进行模拟,以北京市密云区数字高程模型(Digital Elevation Model, DEM)作为地形研究数据进行插值分析和回波模拟,提出了使用广义回归神经网络(Generalized Regression Neural Network,GRNN)进行插值的方法。在插值后的DEM仿真地形上,考虑地形起伏对雷达引信回波遮挡的影响,仿真得到传统插值算法和GRNN插值算法下的雷达引信回波图,通过内插精度和回波分析了算法性能。结果表明:在地貌类型以平原、丘陵为主的地区,相较于传统算法,本文算法能够描绘地形细节,更加精确地模拟回波分布规律。In order to simulate the radar fuze echo signal more accurately,the interpolation and echo simulation were explored by taking the digital elevation model(DEM)of Miyun Dis-trict in Beijing as the terrain research data.Accordingly,an interpolation method using general-ized regression neural network(GRNN)was proposed.Furthermore,with the consideration of the effect of terrain undulation on radar fuze echo obstruction,radar fuze echo maps of the inter-polated DEM simulated terrain were obtained respectively by the traditional interpolation algo-rithm and the GRNN interpolation algorithm.Through the analysis of the interpolation accuracy and echo,the performance of the proposed method was explored.Results showed that in the area where the terrain type is mainly plains and hills,the bilinear interpolation algorithm and the GRNN algorithm can depict the details of the terrain and simulate the distribution of the echo more accurately.

关 键 词：广义回归神经网络 高程插值算法 精度评价 回波模拟 深度神经网络 数字高程模型 

分 类 号：TN973.3[电子电信—信号与信息处理] TJ434[电子电信—信息与通信工程]