【作者】 周琳娜；

【导师】 王黎明；

【作者基本信息】 中北大学 ， 信号与信息处理， 2011， 硕士

【摘要】 目标跟踪和轨迹预测技术在国防科研以及雷达、声呐信号处理及其他相关领域中是一个非常重要的研究课题,尤其是在当卡尔曼滤波理论提出并且广泛应用以后,目标跟踪技术更是发生了很大的进步,其中很多成果已经应用于侦察与预警、弹道导弹防御、交通管制等军事和民用领域当中,本论文着重研究了非线性卡尔曼滤波在目标跟踪预测中的应用。文章首先对目标跟踪的基本原理进行了系统介绍,阐述了几种常用的目标跟踪模型、量测模型的选择和建立,然后介绍了常应用于线性系统中的卡尔曼滤波方法,根据实际的飞行体目标的要求,重点研究了扩展卡尔曼滤波算法(EKF)和粒子滤波算法(PF),针对扩展卡尔曼滤波算法在滤波过程中非线性系统的模型线性化处理过程中引入,因此不可避免地引入了线性化误差,采用粒子滤波进行处理取得较好的效果,其次本文采用非线性滤波建立了相应的目标跟踪预测模型,同时应用扩展卡尔曼滤波和粒子滤波对目标飞行体分别进行了预测滤波的仿真实验,通过仿真实验结果分析,表明了采用粒子滤波算法在目标飞行体跟踪预测方面有着更好的滤波跟踪性能。最后文章应用了VC和Matlab混合编程的方式搭建了简易的预测系统的软件平台,也为今后目标跟踪预测提出了新的发展方向。更多还原

【Abstract】 Target tracking and trajectory predicion technology is a very important research subject in defense research and radar, sonar signal processing and other related fields. A lot of progress has occurred in the area of target tracking technology and it has been used with wide especially after the development of the Kalman filter theory. Some examples of it have been used in reconnaissance and early warning systems for ballistic missile defense, Atraffic control, and other military and civilian areas. This thesis focuses on the use of the nonlinear Kalman filters in high-speed flight body tracking prediction.Firstly, this paper introduces a detailed outline of the principle of target tracking as well as expounding on measurement model selection and development, and several common target tracking models.Then it introduces Kalman filter methods which are often applied to linear systems. According to the actual needs of high-speed flying bodies, this thesis focuses on the extended Kalman filter (EKF) and particle filter algorithm(PF). But the extended Kalman filter algorithm should first be linearized in the filter process of the nonlinear system, so it inevitably introduces linearization errors. However, the particle filter has advantages such as good results. This thesis uses nonlinear filtering to establish the relevant target tracking model. At the same time, it carries out a simulation experiment on high-speed flight bodies using both the extended Kalman filter and particle filter alternatively. Simulation analysis experiments showed the particle filter algorithm has better filter tracing performance in the tracking of high-speed flight bodies.Finally, this paper builds a prediction system software platform using a mixture of VC and Matlab programming, revealing a new direction for target tracking prediction.更多还原