Loopy belief propagation based data association for extended target tracking

The data association problem of multiple extended target tracking is very challenging because each target may generate multiple measurements. Recently, the belief propagation based multiple target tracking algorithms with high efficiency have been a research focus. Different from the belief propagation based Extended Target tracking based on Belief Propagation(ET-BP) algorithm proposed in our previous work, a new graphical model formulation of data association for multiple extended target tracki...     

基于数据关联的故障快速检测

 多数情况下,快速实时地进行故障检测是很重要的。将故障看做是通过多传感器观测的动态模型,进行多传感器多模型概率数据关联,以各个模型的关联结果和设定的阈值为依据,可以有效地实现故障检测。联合概率数据关联（JPDA）算法是解决多传感器多目标跟踪的一个有效方法,文中通过分析概率数据关联算法,对联合概率数据关联算法进行了改进：（1）通过正确地选择阈值,移除小概率事件,进而建立一个近似的确认矩阵;（2）根据被跟踪目标故障跟踪门的相交情况,将跟踪空间进行数学划分,形成若干相互独立的区域;（3）对同一区域内公共有效量测的概率密度值进行衰减,计算出关联概率。仿真对比表明,本文的改进算法能显著减少计算时间,有效提高故障检测的快速性和实时性。     

Track-to-object association algorithm based on TLE filtering

This paper focuses on the track-to-object association problem based on the two-line elements (TLE) set. The TLE’s short-term propagation error characteristics are analyzed to capture its uncertainty. Further, a four-step track-to-object association algorithm is designed for the optical observation data. First, for too-short arc tracklets, a circular orbit determination algorithm is proposed to calculate the inclination and the right ascension of the ascending node. Second, the TLEs are filtered based on these results. Nearly 96% of the TLEs can be filtered, which significantly improves the association efficiency. The last two steps consist of two association processes. A first-order association process is implemented first to get candidate objects, with the angles root mean square error as the metric. Then a precise association process checks the candidate objects and gives the final association result. The proposed approach is tested with simulated and observed data, respectively. With simulated data, the true positive rate is 98.7%. With the observed data, the association results were validated using the precise orbit ephemerides.     

An optimization approach for observation association with systemic uncertainty applied to electro-optical systems

The observation to observation measurement association problem for dynamical systems can be addressed by determining if the uncertain admissible regions produced from each observation have one or more points of intersection in state space. An observation association method is developed which uses an optimization based approach to identify local Mahalanobis distance minima in state space between two uncertain admissible regions. A binary hypothesis test with a selected false alarm rate is used to assess the probability that an intersection exists at the point(s) of minimum distance. The systemic uncertainties, such as measurement uncertainties, timing errors, and other parameter errors, define a distribution about a state estimate located at the local Mahalanobis distance minima. If local minima do not exist, then the observations are not associated. The proposed method utilizes an optimization approach defined on a reduced dimension state space to reduce the computational load of the algorithm. The efficacy and efficiency of the proposed method is demonstrated on observation data collected from the Georgia Tech Space Object Research Telescope.     

被动多站多目标的测量数据关联算法研究

将视线距离的概念引入到被动多站多目标的测量数据关联,提出了各观察站对于同一目标的视线之间的距离和最短的关联准则,在此基础上,构造了测量数据的快速关联算法。其主要特点是不需要目标位置的任何先验信息,从而克服了由此产生的各种不利影响,为数据关联提供了一条新的思路。最后的仿真结果验证了算法的有效性。     

一种适于工程应用的近似MSJPDA算法

为提高复杂数据融合系统中的航迹关联正确率，在ZHOU B的DC和AC算法基础上提出一种新的近似多传感器多目标联合概率数据关联算法，它以一个目标为中心的近似聚为构造关联事件的起点，并在计算中将DC和AC结合得到的一种全部的点迹－航迹关联算法，在杂波下目标密集，航迹复杂的数据融合系统中进行实验，对关联正确率，并联时耗等与最近邻法进行了比较，效果较好，它能为有效提高目标点迹－航迹的关联正确率，在计算时耗上较完全联合概率法少得多，能满足工程中实时性的要求。     

基于多媒体数据挖掘的多通道检索技术

首先介绍多媒体数据挖掘的过程,接着分析了多媒体信息的多通道检索的体系结构和多通道检索流程图。针对多媒体信息内容的复杂性,文章的第四部分介绍了多媒体关联规则挖掘算法,对将其应用于多通道检索进行了初步探讨。     

基于仅光学观测的短弧关联分析方法

初始轨道确定是空间目标编目的一个重要部分,尤其在光学仅角度观测下是极具挑战性的.光学观测在短弧的情况下很难进行有效的初轨确定,解决短弧问题的一个重要手段是将不同时刻获取到的短弧数据进行关联匹配,找到属于同一个目标的观测数据.以容许域的方法为基础,通过找到拟合多组观测数据的最优轨道的方式来确定角度预测值和角度测量真实值之...     

一种基于有效修剪的最大频繁项集挖掘算法

对关联挖掘中的最大频繁项集挖掘问题进行了研究,提出了一种基于项集格修剪机制的最大频繁项集挖掘算法.采用项集格生成树的数据结构,将最大频繁项集挖掘过程转化为对项集格生成树进行深度优先搜索获取所有最大频繁节点的过程. 其中提高算法效率的一个重要措施是在遍历项集格生成树的过程中对生成树进行修剪.给出了项集格生成树的三个性质,并在此基础上提出了直接超集修剪、间接超集修剪与事务集等价修剪三种修剪机制,尽可能忽略非频繁节点及其所生成的扩展节点以减少遍历的节点数目.试验结果表明,三种修剪机制都能够有效地减少搜索空间,其中事务集等价修剪机制的效果最好,算法的性能与输入数据集的稠密程度相关.      

灰色关联综合优化法评估超视距空战效能

为克服以往的效能评估中存在效能指标可信度降低、缺乏横向可比性等弱点,提出利用灰色关联综合优化法对战斗机超视距空战效能进行评估.分析影响超视距空战效能的因素,建立超视距空战效能评估层次体系模型,通过评估实例研究利用灰色关联综合优化法评估战斗机超视距空战效能的步骤.通过对评估结果的分析表明,灰色关联综合优化法是评估超视距空战效能的一种有效方法.