窄带雷达车辆目标分类方法及实现

发布时间：2018-03-07 13:10

本文选题：窄带雷达　切入点：轮式和履带式车辆　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着雷达技术的发展,雷达自动目标识别已经成为了未来雷达发展的方向。窄带雷达的探测距离远,并且在我国装备的数量较多,所以对窄带雷达目标分类具有重要意义。另一方面,窄带雷达的分辨率较低,目标回波包含的信息较少,对目标进行识别的难度大。而在地面运动的轮式车辆和履带车辆目标,由于它们的驱动方式不同,车身相对旋转部件不一样,导致二者的微多普勒回波有较大差异,可以用此特征进行分类。本文即围绕窄带雷达条件下的运动车辆目标分类问题进行了研究,主要内容可以概括为以下三方面:1.对运动车辆目标的微多普勒效应进行介绍,给出了车辆目标的微多普勒模型。首先对于旋转单散射点的微运动进行了介绍,给出了其微多普勒信号模型的数学表达式。在此基础上,给出了旋转体的微运动模型,并且依据模型的数学表达式指出了影响目标的微多普勒信号的相关变量。针对车辆目标的运动特点,进一步给出了车轮与履带的微多普勒模型,从二者模型的数学表达式上推导出微多普勒信号的差异之处。通过对实测的车辆目标雷达回波信号进行分析,指出了轮式车辆和履带式车辆回波的差异。2.针对运动车辆目标的分类识别流程,对一些通用的信号处理方法以及分类方法进行了研究。在目标回波的杂波抑制方面,分别介绍了基于脉冲对消MTI、基于CLEAN算法和基于广义匹配滤波器的杂波抑制方法,通过对比各种方法的杂波抑制效果,介绍了各方法的优缺点。在目标识别特征的提取方面,介绍了基于多普勒分布特征和能量分布特征的提取方法,通过对真实目标进行分类,给出了各种特征的分类效果。在分类算法的选择方面,介绍了LDC、KNN、SVM三种分类算法,通过对各算法的分析比较,给出了各种不同分类算法的适用条件。3.对于窄带雷达运动车辆目标的分类识别,在DSP上进行了硬件工程的实现。在上述分析的目标分类识别的各种实现方法的基础上,综合考虑目标分类效果以及工程的实时性需求,选用合适的方法进行目标分类系统的设计,然后分析了DSP上硬件工程各模块的设计。最后,通过分析硬件分类系统的运算精度、运算时间和分类效果,验证了该窄带雷达运动车辆目标分类系统的可行性与可靠性。
[Abstract]:With the development of radar technology, radar automatic target recognition has become the direction of radar development in the future. On the other hand, the resolution of narrowband radar is low, the target echo contains less information, and it is difficult to recognize the target. On the other hand, the target of wheeled vehicle and tracked vehicle moving on the ground, Because of their different driving modes, the relative rotating parts of the body are different, which leads to a great difference in the micro-Doppler echo between them. This paper focuses on the classification of moving vehicle targets under the condition of narrowband radar. The main contents can be summarized as follows: 1. The micro-Doppler effect of moving vehicle targets is introduced. In this paper, the micro-Doppler model of vehicle target is given. Firstly, the micro-motion of rotating single scattering point is introduced, and the mathematical expression of the micro-Doppler signal model is given. On this basis, the micro-motion model of rotating object is given. According to the mathematical expression of the model, the related variables of the micro-Doppler signal affecting the target are pointed out. According to the motion characteristics of the vehicle target, the micro-Doppler model of the wheel and track is further given. From the mathematical expressions of the two models, the differences of the micro-Doppler signals are deduced. The difference of echo between wheeled vehicle and tracked vehicle is pointed out. Aiming at the classification and recognition flow of moving vehicle target, some general signal processing methods and classification methods are studied. In the aspect of clutter suppression of target echo, The methods of clutter suppression based on pulse cancellation MTI, CLEAN algorithm and generalized matched filter are introduced, and the advantages and disadvantages of each method are compared. The extraction method based on Doppler distribution feature and energy distribution feature is introduced. By classifying real target, the classification effect of various features is given. In the selection of classification algorithm, three classification algorithms of LDC/ KNN- SVM are introduced. Through the analysis and comparison of each algorithm, the suitable conditions of different classification algorithms are given. 3. For the classification and recognition of moving vehicle target of narrowband radar, The realization of hardware engineering on DSP is carried out. On the basis of all kinds of realization methods of target classification and recognition mentioned above, the effect of target classification and the real time requirement of engineering are considered synthetically. A suitable method is selected to design the target classification system, and then the design of each module of hardware engineering on DSP is analyzed. Finally, the operation precision, operation time and classification effect of the hardware classification system are analyzed. The feasibility and reliability of the moving vehicle target classification system for narrowband radar are verified.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.51

【参考文献】