密集信号环境下的雷达信号分选算法研究

发布时间：2018-10-31 15:33

【摘要】：雷达信号分选是电子侦察系统中的一项关键技术，不断优化信号分选方法有助于提高侦察系统的信号处理能力。目前，，用于信号分选的一些可测特征参数主要有脉冲到达时间（TOA）、到达角（AOA）、脉宽（PW）、载频（RF）、脉幅（PA）等，传统雷达信号分选方法是对TOA进行处理，得到脉冲重复间隔（PRI），再根据该参数的特性进行信号分选。为了提高信号处理的速度和分选正确率，常选用AOA、PW和RF作为预分选参数，然后使用PRI进行主分选。本文重点研究密集信号环境下的雷达信号分选算法，包括对上述可用的雷达信号分选参数进行简要介绍，并描述如何进行雷达信号环境的数字模拟。在仿真建模的基础上，对几种传统的基于PRI的雷达信号分选方法进行了对比分析。为适应现代战场环境的需求，许多学者在不断地探索和研究实时高效的雷达信号分选方法，包括对传统算法进行不断改进，使用多参数进行信号分选，采用盲分离技术和脉冲特征提取等。在此，本文研究基于多参数的分选方法，因此，对几种常见的多参数信号分选算法进行了介绍。针对这些算法在进行分选时是处于一种被动状态，即算法需要通过设定一系列的原则来判定数据对象是否属于同一类，提出了基于集对分析聚类的雷达信号分选方法，集对分析方法是直接从数据对象间的相似性出发进行研究的，具有计算速度快的优点。针对原基于集对分析聚类的雷达信号分选方法的聚类结果具有不稳定性，本文对该方法进行了改进，并通过仿真实验进行了验证，为了进一步研究改进算法的性能，对改进算法在含不同噪声比例的情况下进行了仿真分析。此外，在使用支持向量聚类算法进行信号分选时，为了避免该算法的一些影响，提出了一种基于支持向量聚类改进的雷达信号分选算法，该算法采用将支持向量聚类与上述改进的集对分析方法相结合的方法，不仅有效避免了使用支持向量聚类算法所带来的一些不利因素，同时也改善了改进的集对分析方法的抗噪性能，通过实验对该算法进行了验证。
[Abstract]:Radar signal sorting is a key technology in electronic reconnaissance system. Continuous optimization of signal sorting method is helpful to improve the signal processing ability of reconnaissance system. At present, some measurable characteristic parameters used in signal sorting are mainly pulse arrival time (TOA),) arrival angle (TOA), pulse width (AOA), (PW), carrier frequency (RF), pulse amplitude (PA), etc. The traditional method of radar signal sorting is to process TOA. The pulse repeat interval (PRI), is obtained and the signal is sorted according to the characteristic of the parameter. In order to improve the speed and accuracy of signal processing, AOA,PW and RF are often selected as presorting parameters, and then PRI is used for primary sorting. This paper focuses on the radar signal sorting algorithm under the dense signal environment, including a brief introduction of the available radar signal sorting parameters, and describes how to carry out the digital simulation of the radar signal environment. On the basis of simulation modeling, several traditional radar signal sorting methods based on PRI are compared and analyzed. In order to meet the needs of the modern battlefield environment, many scholars are constantly exploring and studying real-time and efficient radar signal sorting methods, including the continuous improvement of traditional algorithms and the use of multi-parameter signal sorting. Blind separation technique and pulse feature extraction were used. In this paper, the multi-parameter sorting method is studied, so several common multi-parameter signal sorting algorithms are introduced. Because these algorithms are in a passive state when sorting, that is, the algorithm needs to set a series of principles to determine whether the data object belongs to the same class, a method of radar signal sorting based on set pair analysis and clustering is proposed. The method of set pair analysis is studied directly from the similarity of data objects, which has the advantage of fast computing speed. In view of the instability of the clustering results of the original clustering method based on set pair clustering, this paper improves the method and verifies it through simulation experiments, in order to further study the performance of the improved algorithm. The improved algorithm is simulated and analyzed with different noise ratio. In addition, in order to avoid the influence of the support vector clustering algorithm, an improved radar signal sorting algorithm based on support vector clustering is proposed. This algorithm combines support vector clustering with the improved set pair analysis method, which not only effectively avoids some unfavorable factors brought by using support vector clustering algorithm. At the same time, the improved set pair analysis method is improved, and the algorithm is verified by experiments.
【学位授予单位】：江南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.51

【参考文献】