节点文献

无源单站定位技术研究

Study on Passive Localization Technology on Single Obserer

【作者】 李秀英

【导师】 彭代渊;

【作者基本信息】 西南交通大学 , 通信与信息系统, 2008, 硕士

【摘要】 无源定位与跟踪技术由于自身不辐射电磁波而且探测距离远,所以它在电子侦察中扮演着越来越重要的角色。而单站定位与跟踪系统由于避免了复杂的时间同步和多个观测站之间的数据融合,因此其受到人们很大的重视,并成为一个研究重点。无源定位跟踪技术的实质是定位跟踪方法和定位跟踪算法的融合。定位跟踪方法和定位跟踪算法是无源定位技术的核心,它们决定着系统的定位精度和实时性。因此本论文从这两方面着手,提出了利用角度及其变化率的定位法,并重点研究了多种滤波算法。文章首先介绍了无源定位跟踪技术的概况。在无源定位领域中,可观测性分析和系统建模是定位跟踪最基本的问题,因此本文对定位系统进行了可观测性分析,并就系统的可观测性建立定位跟踪模型。然后重点介绍了一些定位跟踪滤波算法。卡尔曼滤波(EKF)算法是最经典的非线性估计方法,在无源定位中有不少成功的应用。结合对采用EKF算法的无源定位跟踪系统进行计算机仿真实验,分析了EKF的性能。接着,介绍了无迹卡尔曼滤波(UKF)算法,UKF算法是用确定的粒子点来近似非线性函数的概率分布,因此克服了EKF算法的线性化误差,具有更高的滤波精度,并进行了计算机仿真,分析其性能。然后又介绍了粒子滤波(PF)算法,该算法是用大量的随机粒子来近似非线性函数的概率分布,对后验分布无条件限制,在无源定位跟踪中有较好的精度和稳定性。标准的粒子滤波(PF)算法为了解决粒子退化现象,往往进行重采样。但是传统的粒子滤波算法中重采样往往会导致采样枯竭。因此为了解决这个问题,本文中引用了遗传算法中杂交和突变的概念,来改变传统的粒子滤波所采用的再采样策略,提出新的再采样方法,即遗传粒子滤波(GPFA)算法,解决了粒子枯竭的问题。

【Abstract】 Passive localization and tracking system plays an important role in the electronic reconnaissance, as it works silently without electromagnetic radiation and covering larger region.Single observer passive localzaition and tracking system avoid time synchronization and data fusion,so it attracts more and more research focus.Essentially, passive localization and tracking technology consists of localization and tracking methods and algorithms. They are the two key points of passive localization technology, which decide the precision and rapidity. Thus, the dissertation does deeper researches on the two aspects.The dissertation introduces the model of two-dimensional single-observer passive location using bearing and its changing rate information .Specially, the dissertation does deeper researches on some filtering estimation algorithm.Firstly, background of single observer passive localization is introduce. As the observability and model development are the fundamental problem, they are discussed in dissertation.Followly, some considerable estimation algorithms of localization and tracking are presented. EKF algorithms is the most classical nonlinear method, successfully applying in many passive localization problems. The computer simulations are carried out and the performance in practical application is analyzed. Followly, in order to avoid the weakness of EKF, UKF algorithm in passive localization is studied deeply. UKF use a set of assured particles to estimate the posteriors probability, have a good property in the passive localization and tracking problems. whereafter,particle filtering is studied and applied in passive localization system. It use a set of random samples (also called particles) to estimate the posteriors probability, have a good property in the passive localization and tracking problems.Particle filtering algorithm in order to resolve degeneracy of particle frequently carry out resampling.But resampling of classical particle filtering algorithm lead sample impoverishment of particle. Therefore describe genetic particle filter algorithm (GPFA) in dissertation, indicate advanced resampling method, resolve problem of sample impoverishment

节点文献中: