密集信号环境下的分选技术研究

【作者】 魏娟；

【导师】 杨万麟；

【作者基本信息】 电子科技大学 ， 模式识别与智能系统， 2005， 硕士

【摘要】 在电子对抗领域,信号分选与跟踪算法的研究是一个非常重要的问题。基于平面变换技术的信号分选和基于神经网络模型的人工智能系统是近年来涌现出的比较有创新性的思路,但是这两种方法需要的前提条件在实时信号分选中较难满足,所以其在实际中的应用受到限制。目前,在工程中使用较广的是累积差值直方图法(CDIF)和序列差值直方图法(SDIF)。本文对这几种方法进行了介绍和仿真对比分析。电子战的信号环境日趋复杂,高占空比,高密度是我们所面临的信号环境的特殊性。在这种信号环境下,脉冲丢失概率会比通常的信号环境大大增加。在脉冲丢失概率较大的情况下,由于脉冲信息的丢失和改变,致使传统分选算法的正确率显著下降。因此,针对这样的信号环境,我们应该重新寻求解决问题的方法。本文在这方面作了一些研究,其工作内容和成果概述如下:1. 构建了一个实时预分选系统,改进优化了基于概率统计思想的分选算法。2. 针对脉冲丢失概率较大的情况,提出了还原TOA 差值的两步校验修正算法和CDIF 直方图修正算法。仿真结果表明,两步校验修正算法和CDIF 直方图修正算法能把错误的前后脉冲时间间隔还原到正确值上,有效的补充出由于脉冲重叠而造成的丢失的脉冲信息。同时,本分选系统能取得较好的分选效果,并且计算简单,物理意义明确,这对于实时性要求高的场合是有利的。3. 讨论了系统实现的具体方案,对系统功能进行了分模块设计,并用VHDL语言和C语言对该分选系统的各模块进行了编程实现,并对系统实时性进行了仿真和分析。更多还原

【Abstract】 Signal sorting is a very important problem in electroniccountermeasures. Plane transformation and artificial intelligencesystem based on neural network are novel ways appeared in severalyears. But it is difficult to satisfy the conditions of these two methodsin practical application. Nowadays, cumulative difference histogram(CDIF) and sequential difference histogram (SDIF) are used widely inengineering-oriented application. In this thesis, these methods areintroduced, simulated and compared.The special radar signal environment we face in electronic warfareis high pulse-density and high pulse-losing probability. Under this kindof environment, due to losing and changing of pulse information, thevalidity of the conventional sorting algorithms falls a lot.This thesis is well researched for signal sorting under this kind ofsignal environment. A real-time signal sorting system is designed andthe sorting algorithm based on the probability statistics is improved.Besides, the two-step TOA difference correcting algorithm and theCDIF correcting algorithm are presented. The feasibility of this sortingsystem and the algorithms is verified by lots of simulated experiments.It is easy to implement the system and the algorithm because of theirdirect architecture, which is highly advantageous to real-time sorting.At last, the system is implemented by programming with VHDLlanguage and C language.更多还原