临近空间高超声速条件下雷达目标辨识的研究

发布时间：2018-05-30 03:08

本文选题：临近空间高超声速目标 + 雷达辨识　；参考：《北京交通大学》2017年硕士论文

【摘要】：临近空间高超声速飞行器是近年来的研究热点,临近空间通常被认为是高度界于普通航空飞行器飞行上界与航空飞行器飞行下界之间的空域,速度高于5马赫的飞行器通常被认为是高超声速飞行器。高超声速飞行器采用隐身外形设计,相比于传统类型航空飞行器,它的雷达散射横截面积小,高超声速飞行导致雷达回波信号距离走动、相位模糊问题剧烈,严重影响雷达探测、识别性能。开展针对临近空间超高声速目标的气动外形、运动特征和雷达电磁波散射特征等雷达成像识别方面的研究十分必要。本文针对临近空间高超声速目标雷达辨识理论、模型、融合识别方法开展了相关研究,主要内容包括:首先,介绍分析了雷达目标识别原理与机制,研究不同雷达技术对目标识别的应用范围与可用程度,结合现有高超声速飞行器计划试验机的设计思想明确该类目标识别的重难点与特征提取辨识的努力方向。其次,根据临近空间高超声速目标运动特点与雷达散射特征提出阶梯式组合的雷达识别方法。先利用远程窄带雷达获取的目标距离-多普勒特征提取目标高度、速度整体特征,“粗”识别整体目标,然后研究在宽带雷达获取一维距离像中目标双谱特征、参数化模型特征及ISAR(Inverse Synthetic Aperture Radar,逆合成孔径雷达)距离-瞬态多普勒信息提取等方法实现成像识别。最后,学习、研究分析了模糊理论解决融合识别中不确定性问题,探讨了临近空间高超声速目标运用多传感器融合目标识别方法。研究分析了融合识别信息各层机理,提出在可利用异质传感器融合识别的决策层进行临近空间高超声速目标的融合识别。针对模糊积分中模糊密度研究分析了混淆矩阵、遗传算法、自适应模糊密度赋值方法的应用与各自的优势,使用服从高斯分布和均匀分布的训练样本仿真计算、比较三类方法确定的模糊密度对融合识别平均识别率的优劣。
[Abstract]:The near space hypersonic vehicle is a hot research topic in recent years. The near space is usually considered as the airspace between the upper bound and the lower bound of the general aeronautical vehicle. Aircraft at speeds above Mach 5 are generally considered hypersonic vehicles. The hypersonic vehicle is designed by stealth. Compared with the traditional aerial vehicle, its radar scattering cross section area is smaller, the hypersonic flight causes the radar echo signal range to move, and the phase ambiguity problem is acute. The performance of radar detection and recognition is seriously affected. It is necessary to study the aerodynamic shape, motion characteristics and radar electromagnetic wave scattering characteristics of the near space ultra-high sound target. In this paper, the theory, model and fusion method of hypersonic target radar identification in adjacent space are studied. The main contents are as follows: firstly, the principle and mechanism of radar target recognition are introduced and analyzed. This paper studies the application range and availability of different radar technologies to target recognition, and combines with the design ideas of the existing hypersonic vehicle planning testing machine to identify the heavy difficulties of this kind of target recognition and the direction of feature extraction and identification. Secondly, according to the characteristics of hypersonic target motion and radar scattering in adjacent space, a stepped combination radar recognition method is proposed. First, the target range Doppler feature obtained by long range narrow band radar is used to extract the target height, velocity and "coarse" recognition of the whole target, and then the target bispectral feature in one dimensional range image is studied in wideband radar. Parameterized model features and ISAR(Inverse Synthetic Aperture Radar, inverse synthetic aperture radar (ISAR(Inverse Synthetic Aperture Radar,) range-transient Doppler information extraction are used to realize imaging recognition. Finally, the fuzzy theory is studied to solve the uncertainty problem in fusion recognition, and the multi-sensor fusion target recognition method is discussed. The mechanism of each layer of fusion recognition information is analyzed and the fusion recognition of hypersonic targets in adjacent space is proposed in the decision layer which can be identified by heterogeneous sensors. Based on the research of fuzzy density in fuzzy integration, the application of confusion matrix, genetic algorithm, adaptive fuzzy density assignment method and their respective advantages are analyzed. The training samples from Gao Si distribution and uniform distribution are used to simulate and calculate. The average recognition rate of fusion recognition is compared with the fuzzy density determined by the three methods.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：E926.4;TN95

【相似文献】