基于微多普勒的空间多目标特征提取与参数估计

发布时间：2018-01-22 06:04

本文关键词： 微多普勒多目标特征提取参数估计仿真系统　出处：《电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：当前,基于微动特征的目标识别技术成为了雷达目标识别领域的一个新课题,并被认为是雷达目标识别技术中最有发展潜力的技术之一。本文以空间多目标为研究对象,在建立窄带雷达空间多目标多分量微动模型的基础上,对多目标微动特征提取及参数估计进行了研究,并设计实现了窄带雷达微多普勒仿真系统。论文的主要工作和研究成果有:1.在分析微多普勒效应的基础上,建立了空间目标的微动模型、多目标的位置模型和回波模型。在研究散射中心遮挡效应的基础上,推导了锥体目标微动姿态角的数学表达式。通过微波暗室实验对遮挡效应进行了验证。2.针对时频图中存在噪声的问题,提出了门限阀值降噪处理方法,该方法能够有效的滤除噪声。对viterbi算法进行了深入的研究,该算法能提取多目标多分量时频图中各分量的瞬时频率,并分离对应分量的时频特征,但对存在间断点的时频图的瞬时频率提取失效。针对viterbi算法的不足,提出了瞬时频率极大值估计法,该方法能够有效的提取存在间断点时频图的瞬时频率。3.深入分析了空间锥体目标的微多普勒,得出以下结论:进动的微多普勒为谐波函数形式;章动的微多普勒不是谐波函数,但为周期函数。4.基于极大值法提取的瞬时频率,提出了基于分量累加的周期估计方法,该方法能准确的估计单目标的微动周期,对多目标能进行周期粗估计。针对微多普勒为谐波函数的微动,基于viterbi算法提取的瞬时频率,提出了谐波分析方法,该方法能估计进动锥体目标的自旋和锥旋周期。针对谐波分析方法估计误差较大的问题,给出了一种误差补偿的方法;当回波积累时间大于微动周期时,该方法能有效的补偿谐波分析方法的估计误差。5.针对微多普勒为正弦函数(谐波函数的特例)的情况,研究了正弦函数拟合和多项式函数拟合的参数化估计方法,这两种方法能有效的估计微多普勒的周期、幅度和相位。针对微多普勒为谐波函数的情况,提出了谐波函数拟合的参数化估计方法,该方法能有效的估计各次谐波的幅度和相位,但需要通过谐波分析方法提前估计各次谐波的角频率。6.采用Matlab GUI设计实现了窄带雷达微多普勒仿真系统。该系统能够在线设置目标场景及微动参数,并能对微动特征提取及参数估计算法进行验证。
[Abstract]:At present, the technology of target recognition based on fretting feature has become a new topic in the field of radar target recognition. It is considered to be one of the most promising technologies in radar target recognition. In this paper, a multi-target fretting model of narrow-band radar is established. The multi-target fretting feature extraction and parameter estimation are studied, and the narrowband radar micro-Doppler simulation system is designed and implemented. The main work and research results of this paper are: 1. Based on the analysis of micro-Doppler effect. The fretting model, multi-target position model and echo model of spatial target are established, and the occlusion effect of scattering center is studied. The mathematical expression of fretting attitude angle of pyramidal target is derived. The occlusion effect is verified by microwave darkroom experiment. Aiming at the problem of noise in time-frequency image, a threshold threshold noise reduction method is proposed. This method can effectively filter the noise. The viterbi algorithm is studied deeply. The algorithm can extract the instantaneous frequency of each component in the multi-target multi-component time-frequency map and separate the time-frequency characteristics of the corresponding components. However, the instantaneous frequency extraction of time-frequency graph with discontinuous points is invalid. Aiming at the deficiency of viterbi algorithm, the method of estimating the maximum instantaneous frequency is proposed. This method can effectively extract the instantaneous frequency of the time-frequency map with discontinuous points. The micro-Doppler of the space conical object is deeply analyzed and the following conclusions are drawn: the precession micro-Doppler is a harmonic function; The micro-Doppler of nutation is not a harmonic function, but a periodic function. 4. Based on the instantaneous frequency extracted by the maximum method, a period estimation method based on the accumulation of components is proposed. This method can accurately estimate the fretting period of a single target, and estimate the period coarser for the multi-target. The instantaneous frequency extracted from the micro-Doppler is based on the viterbi algorithm for the micro-Doppler fretting with harmonic function. A harmonic analysis method is proposed, which can estimate the spin and cone rotation periods of the precession cone target. A method of error compensation is presented to solve the problem that the estimation error of the harmonic analysis method is large. When the echo accumulation time is longer than the fretting period, the method can effectively compensate the estimation error of the harmonic analysis method. 5. For the case that the micro-Doppler is a sinusoidal function (the special case of harmonic function). The parameterized estimation methods of sinusoidal function fitting and polynomial function fitting are studied. These two methods can effectively estimate the period, amplitude and phase of micro-Doppler. A parameterized estimation method of harmonic function fitting is proposed, which can effectively estimate the amplitude and phase of each harmonic. However, it is necessary to estimate the angular frequency of each harmonic in advance by harmonic analysis. Matlab is adopted. GUI designs and implements a narrowband radar micro-Doppler simulation system, which can set the target scene and micro-motion parameters online. The algorithm of fretting feature extraction and parameter estimation can be verified.
【学位授予单位】：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】