微多普勒效应在飞机目标分类中的应用研究

发布时间：2019-01-27 09:38

【摘要】：飞机作为现代战争的主要作战工具之一,对战争的胜利起着关键作用。飞机上旋转部件的微运动会对雷达回波产生周期性的频率调制,从而使不同飞机目标表现出不同的微多普勒特性,这对于研究飞机目标分类技术具有重要的指导意义和实用价值。本文在分析了微多普勒效应产生机理的基础上,旨在对飞机目标分类中的特征提取方法进行深入研究。主要完成工作如下:第一,针对直升机、螺旋桨飞机、喷气式飞机三种飞机目标,推导了其旋转部件调制回波理想参数模型,分别研究了飞机参数尾旋翼、发动机个数、桨叶角和雷达参数脉冲重复频率、波束驻留时间以及发射波长对三类飞机旋转部件调制回波的影响。从特征类间差异性和类内聚集性的角度,对比提取了时域二阶矩、时域波形熵、幅度相对量和相位相对量四个特征组成目标分类的特征向量,并利用支撑向量机分类器对三类飞机实现分类。实验结果表明提取的特征向量在信噪比较高时可以有效地对飞机进行分类,并且分类效果随信噪比升高而提高,这四种特征简单易实现,计算复杂度也低,所以有利于工程实现。第二,由于提取喷气引擎调制特征要求回波至少要包含一个完整的目标转动周期,而当雷达工作在扫描状态下时,通常很难满足这一条件,所以针对这种情况,本文提出了一种基于部分周期数据的扫描雷达目标微多普勒参数估计方法。该方法首先通过时频分析技术将信号变换到时频域,从回波信号的时频域中提取出部分周期的微多普勒信号,对转动目标分析可知其微多普勒信号可以通过正弦信号的形式表示,利用正弦函数的性质,可以估计出目标转动频率和转动幅度,从而可以估计出目标转动半径。转动半径作为目标的固有特性,反映目标的结构信息和运动特征,为目标识别和分类提供了有力依据,更为完整地描述了转动目标的运动细节。仿真实验通过与Hough变换方法的对比表明本文方法具有更高的精度和较低的复杂度,并利用实测数据验证了该方法的有效性和鲁棒性。
[Abstract]:Aircraft, as one of the main combat tools in modern warfare, plays a key role in the victory of war. The micromotion of rotating parts on the aircraft produces periodic frequency modulation of radar echo, which makes different aircraft targets exhibit different micro-Doppler characteristics. It has important guiding significance and practical value for the research of aircraft target classification technology. Based on the analysis of the mechanism of micro-Doppler effect, this paper aims to study the feature extraction method in aircraft target classification. The main work is as follows: first, for the helicopter, propeller aircraft and jet aircraft, the ideal parameter model of the modulation echo of the rotating parts is derived, and the tail rotor and the number of engines are studied respectively. The effects of blade angle, pulse repetition frequency of radar parameters, beam-dwell time and emission wavelength on the modulation echo of three kinds of aircraft rotating components. From the point of view of inter-class difference and intra-class aggregation, the second moment, waveform entropy, amplitude relative and phase relative feature of target classification are extracted, and the feature vectors of target classification are compared and extracted, which include the second moment in time domain, the waveform entropy in time domain, the relative amplitude and the relative value in phase. Support vector machine classifier is used to classify three kinds of aircraft. The experimental results show that the extracted feature vectors can effectively classify aircraft when the SNR is high, and the classification effect increases with the increase of SNR. These four features are simple and easy to implement, and the computational complexity is also low. So it is beneficial to project realization. Second, because the extraction of jet engine modulation features requires at least a complete target rotation period to be included in the echo, it is usually difficult to satisfy this condition when the radar is operating in a scanning state, so in this case, In this paper, a method for estimating micro-Doppler parameters of scanning radar targets based on partial periodic data is proposed. Firstly, the time-frequency analysis technique is used to transform the signal into time-frequency domain, and the partial periodic micro-Doppler signal is extracted from the time-frequency domain of the echo signal. The microDoppler signal can be expressed in the form of sinusoidal signal. By using the properties of sinusoidal function, the rotational frequency and amplitude of the target can be estimated, and the rotation radius of the target can be estimated. As the inherent characteristic of the target, the rotational radius reflects the structure information and motion characteristics of the target, and provides a powerful basis for target recognition and classification, and describes the moving details of the rotating target more completely. The simulation results show that the proposed method has higher accuracy and lower complexity than the Hough transform method. The effectiveness and robustness of the proposed method are verified by using the measured data.
【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】：E926.3;TN957.51

【共引文献】