自适应恒虚警算法研究

发布时间：2019-05-27 19:54

【摘要】：在雷达系统中,恒虚警率问题是每个雷达设计者必须要面对的重要问题之一。恒虚警检测技术是雷达目标检测系统中控制虚警率的最重要手段。本文在深入理解目标检测理论与恒虚警率处理方法的基础上,主要研究了在高斯杂波背景下的自适应恒虚警检测技术。这其中包括均匀高斯背景、存在杂波边缘的非均匀环境以及多目标情况下的检测问题。本文重点研究的恒虚警率处理方法有均值类恒虚警率、有序统计类恒虚警率以及杂波图恒虚警率处理方法。其中,对于均值类恒虚警率处理方法,主要研究了单元平均恒虚警率方法、单元平均选大恒虚警率方法和单元平均选小恒虚警率方法。本文简单介绍了每种方法的基本原理以及实现过程,并分析了这些方法在均匀高斯环境、杂波边缘环境以及多目标环境下的特点,之后对这些方法在不同环境下的目标检测性能进行计算机数值仿真分析,最后讨论了这些方法的恒虚警率损失。杂波图处理方法为空域杂波变化剧烈而时域上变化平稳的环境下的目标检测提供了一种很好的方法。本文采用杂波图点技术和杂波图面技术对慢速移动目标和快速移动目标的检测性能进行仿真分析,并且对比分析这两种方法的优缺点。在理论仿真之后,本文对实测数据进行处理。根据处理结果,进一步验证了每种检测方法的特点。最后,本文提出了一种自适应恒虚警算法选择的方法,该方法是在均值类CFAR和杂波图方法的基础上提出来的。首先建立待检测区域的杂波图,然后根据杂波图判断出杂波的分布情况,即判断杂波平稳区、杂波边缘的弱杂波区与杂波边缘的强杂波区,最后根据杂波的分布情况选用最合适的CFAR处理方法。从实验结果中看出,本文提出的方法能有效地减少雷达检测中的虚警和漏警,显著地提高系统的性能。
[Abstract]:In radar system, constant false alarm rate is one of the important problems that every radar designer must face. CFAR detection technology is the most important means to control false alarm rate in radar target detection system. Based on the deep understanding of target detection theory and CFAR processing method, this paper mainly studies the adaptive CFAR detection technology in the background of Gao Si Clutter. This includes uniform Gao Si background, non-uniform environment with clutter edge and detection problem in the case of multi-target. In this paper, the methods of constant false alarm rate are studied, such as mean constant false alarm rate, ordered statistical constant false alarm rate and clutter graph constant false alarm rate processing method. Among them, for the mean constant false alarm rate treatment method, the unit average constant false alarm rate method, the unit average selection large constant false alarm rate method and the unit average selection small constant false alarm rate method are mainly studied. In this paper, the basic principle and implementation process of each method are briefly introduced, and the characteristics of these methods in uniform Gao Si environment, cluttered edge environment and multi-target environment are analyzed. Then the target detection performance of these methods in different environments is analyzed by computer numerical simulation. Finally, the constant false alarm rate loss of these methods is discussed. The wavelet pattern processing method provides a good method for target detection in the environment where the spatial clutters change violently and the time domain changes smoothly. In this paper, the detection performance of slow moving target and fast moving target is simulated and analyzed by using clutter map point technology and clutter surface technology, and the advantages and disadvantages of the two methods are compared and analyzed. After the theoretical simulation, the measured data are processed in this paper. According to the processing results, the characteristics of each detection method are further verified. Finally, an adaptive CFAR algorithm selection method is proposed, which is based on the mean CFAR and hash graph methods. Firstly, the hash map of the area to be detected is established, and then the distribution of the clutter is judged according to the hash map, that is, the stable region of the clutter, the weak hash region of the edge of the clutter and the strong clutter region of the edge of the clutter. Finally, the most suitable CFAR processing method is selected according to the distribution of clutters. The experimental results show that the proposed method can effectively reduce the false alarm and missed alarm in radar detection, and significantly improve the performance of the system.
【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】：TN957.52

【参考文献】