机动目标跟踪技术研究

《机动目标跟踪技术研究》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、机动目标跟踪技术研究机动目标跟踪技术研究摘要：机动目标跟踪理论在国防和民用等领域具有重要的应用价值。本文重点研究机动目标的建模和非线性系统的滤波的问题。滤波算法是机动目标跟踪过程中一个重要的组成部分。在对机动目标建模后，通过滤波算法对模型中的状态向量进行预测和估计。本文首先在估计理论和方法的基础上引入了在线性系统中最常用和最基础的卡尔曼滤波算法。然后针对本文所研究的问题，介绍了传统的非线性系统滤波算法扩展卡尔曼滤波算(EKF)，重点研究了无迹卡尔曼滤波算法(UKF)。由于扩展卡尔曼滤波在滤波过程中首先要对非线性系统的模型进行线性化处理，因此就需要引入线性化误差，而无迹卡尔曼滤波是一种新的专门针

2、对非线性系统的滤波算法，具有实现简单、通用性强，性能稳定等特点。最后，在机动目标选定运动模型和滤波算法的基础上，对机动目标的运动作了仿真实验。从仿真分析中可以看出，与传统的扩展卡尔曼滤波算法相比较，无迹卡尔曼滤波算法有较小的跟踪误差，较高的跟踪精度。关键词：非线性系统滤波；机动目标；扩展卡尔曼滤波；无迹卡尔曼滤波ResearchonTrackingofManeuveringTargetsAbstract：Theproblemsofbuildingthemodelofthemaneuveringtargettrackingand33机动目标跟踪技术研究thefilteringofthenonli

3、nearsystemsarestudiedmainly.Meanwhile，therelatedsimulationexperimentsaredoneonthebaseofthetheoryintroduced.Thefilteringalgorithmisaimportantpartintheprocessoftrackingthemaneuveringtarget.Afterthemodelofthemaneuveringtargetisconfirmed，thevectorofthestatewillbepredictedandestimatedthroughthefilteringa

4、lgorithm.TheusualandbasicKalmanfilteringalgorithmisintroducedonthebaseofthetheoriesandthemethodsoftheestimation.Aimingatthestudiedproblem,thetraditionalfilteringalgorithmsofthenonlinearsystemnamedtheExtendedKalmanFilteringareintroduced.theUnscentedKalmanFilteringisintroducedmostly.Becausethemodeloft

5、henonlinearsystemfirstlymustbelinearizedintheprogressoftheExtendedKalmanFiltering，theerrorintroducedintheprogressoflinearizationisunavoidable.However，theUnscentedKalmanFilteringisanewalgorithmwhichstudyspeciallythenonlinearsystemandhavesometraitssuchastherealizationeasily，comprehensiveapplication，st

6、ableperformanceandsoon.Withaviewtotheaccuracyoftracking，theapplicationoftheUnscentedKalmanFilteringintheTrackingtothemaneuveringtargetisstudiedmainly.Atlast，accordingtothemodelandthefilteralgorithm，thesimulationexperimentsaboutthemovementofthemaneuveringtargetisdone.Totheconcludefromtheanalysesofthe

7、simulation，theUnscentedKalmanFilteringhashighaccuracyintracking.WiththecomparisontotheExtendedKalmanFiltering，theUnscentedKalmanFilteringhasthelesserrorofthetracking.Keywords：nonlinearsystemfiltering；