无源定位与跟踪相关技术研究

发布时间：2018-01-06 15:36

本文关键词：无源定位与跟踪相关技术研究　出处：《西安电子科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：无源定位与跟踪是多传感器数据融合的一个重要研究内容,无源定位跟踪系统因具有抗干扰、抗低空突防、抗反辐射导弹、反隐身等优点,所以在现代电子战争中扮演着至关重要的角色,同时,其在航海、航空、等民用领域也发挥着重要作用。本文针对无源定位跟踪系统中的相关问题进行了仔细地研究。首先,对于多站定位跟踪系统中,针对传感器观测噪声未知会影响定位跟踪精度的问题,研究了最小二乘自适应卡尔曼滤波算法;其次,针对野值会影响滤波精度问题,研究了基于新息似然的野值的检测与剔除方法。以上研究为无源定位跟踪的工程应用提供了理论和方法支撑。取得的主要研究成果如下:1.研究了多站无源定位跟踪系统的基本理论。首先简单介绍了多站无源定位的建模。其次介绍了几种常见的目标运动模型,包括针对直线运动的匀速运动模型以及匀加速运动模型,以及针对机动目标的Singer模型、当前统计模型和转弯模型;最后详细研究了经典的滤波算法,包括针对线性的卡尔曼滤波、以及非线性的扩展卡尔曼滤波、无迹卡尔曼滤波和粒子滤波。2.针对多站无源定位跟踪系统中传感器观测噪声协方差未知的问题,研究了最小二乘自适应卡尔曼滤波。该算法结合测向多站交叉定位与变分贝叶斯卡尔曼滤波,最终联合估计目标的状态以及观测噪声协方差。首先利用最小二乘法对目标进行测向多站交叉定位,并把其作为卡尔曼滤波的伪量测;同时针对系统中传感器观测噪声协方差未知的问题,将变分贝叶斯引入卡尔曼滤波中,通过变分贝叶斯卡尔曼滤波联合估计目标的状态以及观测噪声协方差。很好地解决了无源定位跟踪系统中传感器观测噪声未知将会影响定位跟踪性能的问题,有效地提高了定位跟踪精度。3.针对无源定位跟踪中野值的出现会降低滤波的可靠性和稳定性问题,研究了一种基于似然的野值检测与剔除方法。首先通过理论分析的角度研究了野值的出现对于定位跟踪精度的影响;接着重点介绍了三种基于新息的野值检测与剔除的方法,几种方法对于也值得检测与剔除都具有很好的效果;最后针对野值问题,结合新息似然的概念,研究了针对野值检测与剔除的新息似然法。该算法基于卡尔曼滤波框架,首先通过计算卡尔曼滤波更新中得到的似然值,并设定判定门限,然后根据此门限来对量测进行判定,进而达到野值的检测与剔除的目的。该算法能够有效地对野值进行检测与剔除,很好地解决了无源定位跟踪中野值对定位跟踪精度的影响,使得目标定位跟踪精度有了很大的提高。
[Abstract]:Passive location and tracking is an important research content of multi-sensor data fusion. Passive positioning and tracking system has the advantages of anti-jamming, anti-low-altitude penetration, anti-radiation missile, anti-stealth and so on. So it plays a vital role in modern electronic warfare, at the same time, in navigation, aviation. And other civil areas also play an important role. In this paper, the passive positioning and tracking system related issues are carefully studied. First, for the multi-station positioning and tracking system. The least square adaptive Kalman filter algorithm is studied to solve the problem that the sensor observation noise does not know the location and tracking accuracy. Secondly, the outliers will affect the filtering accuracy. The methods of outlier detection and elimination based on innovation likelihood are studied. The above research provides theoretical and methodological support for the engineering application of passive location and tracking. The main research results are as follows:. 1. The basic theory of multi-station passive location and tracking system is studied. Firstly, the modeling of multi-station passive location is briefly introduced. Secondly, several common target motion models are introduced. It includes uniform motion model for linear motion and uniform acceleration motion model, Singer model for maneuvering target, current statistical model and turning model. Finally, the classical filtering algorithms are studied in detail, including linear Kalman filtering and nonlinear extended Kalman filtering. Unscented Kalman filter and particle filter. 2. To solve the problem of unknown covariance of sensor noise in multi-station passive positioning and tracking system. The least square adaptive Kalman filter is studied, which combines direction finding multi-station cross location and variational Bayesian Kalman filtering. Finally, the state of the target and the covariance of the observed noise are estimated jointly. Firstly, the least square method is used to locate the target in multi-station crossover, and it is regarded as the pseudo-measurement of Kalman filter. At the same time, the variational Bayes is introduced into Kalman filter to solve the problem that the covariance of sensor observation noise is unknown. By using variational Bayesian Kalman filter to estimate the state of the target and the covariance of the observation noise, the problem that the unknown sensor observation noise will affect the tracking performance of the passive positioning and tracking system is solved. The outliers in passive location tracking can reduce the reliability and stability of filtering. In this paper, a method of outlier detection and elimination based on likelihood is studied. Firstly, the influence of outliers on location and tracking accuracy is studied from the point of view of theoretical analysis. Then three methods of outlier detection and elimination based on innovation are introduced. Finally, aiming at the outliers problem and combining the concept of innovation likelihood, this paper studies the innovation likelihood method for the detection and elimination of outliers. The algorithm is based on the Kalman filter framework. At first, the likelihood value of Kalman filter update is calculated, and the decision threshold is set, and then the measurement is judged according to this threshold. The algorithm can detect and eliminate outliers effectively and solve the influence of outliers in passive location tracking. The accuracy of target positioning and tracking has been greatly improved.
【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】：2015
【分类号】：TN713

【相似文献】