复杂环境下雷达信号的分选方法

发布时间：2019-06-02 14:11

【摘要】：雷达信号分选指从多部雷达发射的混叠信号中将各个雷达发出的信号归类的过程。在当前电子对抗战争中,及时准确地侦察敌方信号,捕获敌方信息是取得胜利的关键,而信号分选是电子侦察系统的关键技术。因此,在当前电子环境复杂,信号交叠严重,雷达辐射源未知的条件下,如何有效地将雷达信号分选出来是亟待解决的问题。为此,本文首先对k均值聚类算法和模糊聚类算法进行深入研究:k均值聚类算法是对雷达样本进行的硬划分,聚类准确率不高,模糊聚类则需要事先设定先验信息,对未知雷达辐射源信号不能有效聚类,针对传统聚类算法存在的不足,首先从雷达信号的脉间特征着手,利用传统五参数,提出一种基于入侵性杂草改进的FCM算法,然后研究雷达信号的脉内特征的提取方法,提出一种基于时频原子提取脉内特征的方法。主要研究内容和取得的成果如下:针对雷达信号脉间特征的聚类,入侵性杂草算法具有结构简单,参数少,全局搜索能力强的特点,能够在较少的迭代次数下搜寻最优解,为此,本文提出一种基于杂草改进的FCM算法,该算法主要是对模糊聚类算法对初始聚类中心的依赖性进行改进,首先根据样本数目确定雷达类别数目的解空间,然后根据距离准则,采用杂草算法在整个解空间内搜索最佳的类别数目,作为模糊聚类的初始参数进行聚类,并跟传统的k均值聚类和AP聚类算法进行比较,验证了该算法摆脱了对初始聚类中心的依赖性,具有较高的分选正确率。针对雷达信号脉内特征的聚类,使用雷达信号的脉内特征来研究信号的分选问题是近几年讨论的热点,许多学者验证了基于时频原子提取脉内特征是有效的,但是时频原子的数量巨大,计算复杂度高,针对这一问题本文提出一种改进的时频原子提取脉内特征的方法,首先介绍5种经典的雷达信号的数学模型,然后提出将杂草算法与时频原子相结合的方法,根据距离准则,利用杂草智能算法搜寻能够区分不同调制方式的雷达信号的一组原子,并与待分选的雷达信号做内积运算,作为改进FCM算法的输入矢量进行分类,分别在-3dB到5dB的信噪比下进行仿真实验,验证该算法的有效性。
[Abstract]:Radar signal sorting refers to the process of classifying the signals emitted by each radar from mixed signals transmitted by multiple radars. In the current electronic confrontation war, timely and accurate detection of enemy signals and acquisition of enemy information is the key to victory, and signal sorting is the key technology of electronic reconnaissance system. Therefore, under the condition that the current electronic environment is complex, the signal overlap is serious, and the radar radiation source is unknown, how to effectively sort out the radar signal is an urgent problem to be solved. Therefore, in this paper, the k-means clustering algorithm and fuzzy clustering algorithm are deeply studied: K-means clustering algorithm is a hard division of radar samples, the clustering accuracy is not high, fuzzy clustering needs to set prior information in advance. The unknown radar emitter signal can not be effectively clustering. Aiming at the shortcomings of the traditional clustering algorithm, an improved FCM algorithm based on invasive weeds is proposed by using the traditional five parameters, starting from the inter-pulse characteristics of radar signals. Then the method of extracting intra-pulse features of radar signals is studied, and a method of extracting intra-pulse features based on time-frequency atoms is proposed. The main research contents and achievements are as follows: according to the clustering of radar signal inter-pulse characteristics, the invasive weed algorithm has the characteristics of simple structure, less parameters and strong global search ability, and can search for the optimal solution under less iterations. In this paper, an improved FCM algorithm based on weeds is proposed, which mainly improves the dependence of fuzzy clustering algorithm on the initial clustering center. Firstly, the solution space of radar category number is determined according to the number of samples. Then, according to the distance criterion, the weed algorithm is used to search for the best number of categories in the whole solution space, which is used as the initial parameter of fuzzy clustering, and compared with the traditional k-means clustering and AP clustering algorithm. It is verified that the algorithm gets rid of the dependence on the initial clustering center and has a high sorting accuracy. Aiming at the clustering of intra-pulse features of radar signals, it is a hot topic to study the sorting of radar signals by using the in-pulse features of radar signals in recent years. Many scholars have verified that it is effective to extract intra-pulse features based on time-frequency atoms. However, the number of time-frequency atoms is huge and the computational complexity is high. In order to solve this problem, an improved method of extracting intra-pulse features by time-frequency atoms is proposed in this paper. Firstly, five classical mathematical models of radar signals are introduced. Then, a method of combining weed algorithm with time-frequency atom is proposed. According to the range criterion, the weed intelligent algorithm is used to search for a group of atoms that can distinguish different modulation radar signals, and the internal product operation is performed with the radar signal to be sorted. As the input vector of the improved FCM algorithm, the simulation experiments are carried out under the signal-to-noise ratio (SNR) of-3dB to 5dB to verify the effectiveness of the algorithm.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.51

【参考文献】