基于协方差矩阵的自适应目标跟踪研究

发布时间：2018-04-05 15:35

本文选题：多目标跟踪　切入点：能量控制　出处：《南京航空航天大学》2014年硕士论文

【摘要】：目标跟踪技术是当今雷达数据处理的研究热点之一。目前常用的跟踪滤波器是卡尔曼滤波器。本文围绕雷达跟踪时的能量控制方法展开研究,主要工作是在研究几种滤波算法的基础上用更优化的求积分卡尔曼滤波和容积卡尔曼滤波,结合交互式多模型,设计出自适应目标跟踪算法,控制雷达辐射能量。即在不改变雷达跟踪精度的前提下,增大采样间隔,从而降低雷达辐射能量,有利于雷达跟踪更多的目标,进一步提高雷达的工作效率。本文主要研究内容如下:1、分析了目标跟踪技术的发展,以及在目标跟踪过程中自适应采样时能量控制的研究意义。分析了非线性滤波跟踪算法的研究现状和自适应目标跟踪时的能量控制的研究现状。2、阐述了目标跟踪的基础理论,包括常见的目标运动模型,重点研究了几种滤波算法,仿真比较了几种滤波算法的性能,同时分析了交互式多模型算法、灰色关联度理论、粒子群优化算法等,为后续章节的研究奠定了基础。3、给出了一种交互式求积分卡尔曼滤波的自适应采样间隔目标跟踪算法。首先介绍了目标协方差矩阵估计,然后构造协方差控制资源管理模型,再基于灰色关联度理论和粒子群优化算法,结合求积分卡尔曼滤波算法,设计了交互式求积分卡尔曼滤波的自适应采样间隔目标跟踪算法,并与基于协方差控制的两类资源管理算法进行了性能比较,仿真结果表明,该方法能够在保证跟踪精度的情况下,增大采样间隔,可以节省更多的雷达时间资源。4、在交互式求积分卡尔曼滤波自适应采样间隔目标跟踪算法的基础上,设计了交互式容积卡尔曼滤波的自适应采样间隔目标跟踪算法。首先分析了容积卡尔曼滤波算法,然后在三阶球面-径向准则的基础上,实现高阶容积卡尔曼滤波算法,最后设计出基于交互式容积卡尔曼滤波的自适应采样间隔算法,与交互式求积分卡尔曼滤波算法进行了性能比较,仿真结果表明,该方法较之前的滤波算法,能够进一步优化采样间隔,减少雷达照射次数。
[Abstract]:Target tracking technology is one of the hotspots in radar data processing.At present, the commonly used tracking filter is Kalman filter.An adaptive target tracking algorithm is designed to control radar radiation energy.That is, without changing the tracking accuracy of radar, the sampling interval is increased, thus reducing the radar radiation energy, which is conducive to radar tracking more targets, and further improve the working efficiency of radar.The main contents of this paper are as follows: 1. The development of target tracking technology and the significance of energy control in adaptive sampling are analyzed.The research status of nonlinear filter tracking algorithm and energy control in adaptive target tracking are analyzed. The basic theory of target tracking is expounded, including the common target motion model, and several filtering algorithms are emphatically studied.The performance of several filtering algorithms is compared, and the interactive multi-model algorithm, grey correlation degree theory, particle swarm optimization algorithm and so on are analyzed.It lays a foundation for the following chapters. 3. An adaptive sampling interval target tracking algorithm based on interactive integral Kalman filter is proposed.Firstly, the objective covariance matrix estimation is introduced, then the covariance control resource management model is constructed, and then based on the grey correlation degree theory and particle swarm optimization algorithm, the integral Kalman filtering algorithm is combined.An adaptive sampling interval target tracking algorithm based on interactive integral Kalman filter is designed and compared with two kinds of resource management algorithms based on covariance control. The simulation results show that,This method can increase the sampling interval and save more radar time resource. It is based on the interactive integrated Kalman filter adaptive sampling interval target tracking algorithm.An adaptive sampling interval target tracking algorithm based on interactive volumetric Kalman filter is designed.Firstly, the volumetric Kalman filter algorithm is analyzed, then the high-order volumetric Kalman filter algorithm is realized on the basis of the third-order spherical and radial criteria. Finally, an adaptive sampling interval algorithm based on interactive volumetric Kalman filter is designed.Compared with the interactive integral Kalman filtering algorithm, the simulation results show that the proposed method can further optimize the sampling interval and reduce the radar irradiating times.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN953

【参考文献】