ICA在信号分选中的应用
发布时间：2018-12-13 13:14
【摘要】：雷达信号分选是将截获到的多个混合在一起的雷达脉冲信号分选至同一辐射源类别之中。随着电磁环境的不断复杂化,大量的辐射源信号导致的脉冲混叠、参数之间高相关度、调制方式复杂多样,传统的依赖几种常规信号参数进行分选的方法已经逐渐变得不能适应当前情况。本文主要研究一种基于独立分量分析的雷达信号分选方法,其优势在于只需要信源之间是统计独立的,就可以将分属不同信源的信号分选出来。本文所做的主要工作有:研究了传统的信号分选方法,分析了传统算法在现代复杂电磁环境下,存在的局限性以及其存在局限性的原因。分析了盲源分离算法,以及其能够适用于信号分选的原因和原理,并将基于负熵的快速独立分量分析算法引入信号分选。先将截获信号做归一化和白化预处理,然后用独立分量分析进行分离,对分离出的信号估计参数并将不同时间段参数相同的进行归类,达到分选目的。经过仿真实验证实,这种分选方法具有良好性能。同时,将快速独立分量分析分选算法与传统分选算法做对比,从各个方面证实了快速独立分量分析分选算法的优势。多通道接收机并不一定能够满足信号接收通道大于信源数的理想正定问题。本文也提出了在单通道情况下,将经验模式分解与独立分量分析相结合并应用于信号分选的方法。先用经验模式分解算法将多域混合的接收信号分解成多个本征模函数,然后将其筛选后构造多通道进行独立分量分析,对未知混合信号进行分选。经过仿真验证,这种方法拥有较好的分选效果。
[Abstract]:Radar signal sorting is to separate multiple intermingled radar pulse signals into the same emitter category. As the electromagnetic environment becomes more and more complicated, a large number of emitter signals cause pulse aliasing, the parameters are highly correlated, and the modulation methods are complex and diverse. Traditional methods which rely on several conventional signal parameters for sorting have gradually become unable to adapt to the current situation. In this paper, a method of radar signal sorting based on independent component analysis (ICA) is studied. The advantage of this method is that the signals belonging to different sources can be sorted out only if the sources are statistically independent. The main work of this paper is as follows: the traditional signal sorting method is studied, and the limitations of the traditional algorithm in the modern complex electromagnetic environment are analyzed as well as the reasons for its limitations. This paper analyzes the cause and principle of blind source separation algorithm, and introduces the fast independent component analysis algorithm based on negative entropy into signal sorting. Firstly, the intercepted signal is pretreated with normalization and whitening, then separated by independent component analysis (ICA), the parameters of the separated signal are estimated and the parameters of different time periods are classified to achieve the purpose of sorting. The simulation results show that this method has good performance. At the same time, comparing the fast independent component analysis sorting algorithm with the traditional sorting algorithm, the advantages of the fast independent component analysis sorting algorithm are confirmed from various aspects. Multi-channel receiver is not always able to satisfy the ideal positive definite problem that the signal receiving channel is larger than the number of sources. In this paper, we also propose a method of combining empirical mode decomposition with independent component analysis in the case of single channel and applying it to signal sorting. The multi-domain mixed received signal is decomposed into multiple eigenmode functions by empirical mode decomposition algorithm, and then the multi-channel independent component analysis is constructed after its filtering, and the unknown mixed signal is sorted. The simulation results show that this method has better sorting effect.
【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
[Abstract]:Radar signal sorting is to separate multiple intermingled radar pulse signals into the same emitter category. As the electromagnetic environment becomes more and more complicated, a large number of emitter signals cause pulse aliasing, the parameters are highly correlated, and the modulation methods are complex and diverse. Traditional methods which rely on several conventional signal parameters for sorting have gradually become unable to adapt to the current situation. In this paper, a method of radar signal sorting based on independent component analysis (ICA) is studied. The advantage of this method is that the signals belonging to different sources can be sorted out only if the sources are statistically independent. The main work of this paper is as follows: the traditional signal sorting method is studied, and the limitations of the traditional algorithm in the modern complex electromagnetic environment are analyzed as well as the reasons for its limitations. This paper analyzes the cause and principle of blind source separation algorithm, and introduces the fast independent component analysis algorithm based on negative entropy into signal sorting. Firstly, the intercepted signal is pretreated with normalization and whitening, then separated by independent component analysis (ICA), the parameters of the separated signal are estimated and the parameters of different time periods are classified to achieve the purpose of sorting. The simulation results show that this method has good performance. At the same time, comparing the fast independent component analysis sorting algorithm with the traditional sorting algorithm, the advantages of the fast independent component analysis sorting algorithm are confirmed from various aspects. Multi-channel receiver is not always able to satisfy the ideal positive definite problem that the signal receiving channel is larger than the number of sources. In this paper, we also propose a method of combining empirical mode decomposition with independent component analysis in the case of single channel and applying it to signal sorting. The multi-domain mixed received signal is decomposed into multiple eigenmode functions by empirical mode decomposition algorithm, and then the multi-channel independent component analysis is constructed after its filtering, and the unknown mixed signal is sorted. The simulation results show that this method has better sorting effect.
【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
【共引文献】
相关博士学位论文 前7条
1 王法松;盲源分离的扩展模型与算法研究[D];西安电子科技大学;2013年
2 王坤朋;微弱信号检测的盲源分离方法及应用研究[D];重庆大学;2014年
3 周昊;基于盲源分离的风力发电机主轴承振声诊断研究[D];沈阳工业大学;2014年
4 崔立志;HPLC-DAD数据分离模型及其求解算法研究[D];华东理工大学;2015年
5 崔红岩;术中脊髓监护体感诱发电位异常预警动态预测模型研究[D];北京协和医学院;2015年
6 李U，
本文编号：2376592
本文链接：https://www.wllwen.com/kejilunwen/wltx/2376592.html
