粒子滤波应用于光电跟踪定位的研究

发布时间：2018-07-09 16:48

本文选题：粒子滤波 + 目标跟踪　；参考：《西安电子科技大学》2013年硕士论文

【摘要】：随着现代科技水平的发展，军事装备竞赛日趋激烈，武器打击的精确度更准、进攻速度更快，从而要求我们及时对目标位置和运动状态做出准确的探测和估计。同时，现代电子战技术强调高隐蔽性，所以具有隐蔽性高、抗干扰性强的无源被动定位技术越来越受到重视。本文开展了仅利用角度信息目标无源跟踪定位研究，其中，，对目标状态的预测滤波算法采用了粒子滤波算法。在粒子滤波算法的理论框架下，作者论述了目标的运动模型、目标的观测模型的建立、重采样算法的具体实现。在基本的粒子滤波算法基础上，提出了改进的粒子滤波算法，即在原粒子滤波的重采样算法中加入代间约束。仿真实验结果表明，改进后的粒子滤波算法具有更高的跟踪精度，可以更好的对目标进行跟踪定位。考虑到修正的球面坐标系（MSC），相对于传统的笛卡尔坐标系，具有非线性特性和含有丰富的状态变量等优势，可以有效的提高跟踪精度，本文主体算法采用混合坐标系：在修正球面坐标系中进行滤波跟踪，在笛卡尔坐标系中进行目标运动状态的外推。仿真实验结果表明，这种方法可以得到更好的跟踪效果。
[Abstract]:With the development of modern science and technology, the military equipment competition is becoming more and more intense, the precision of weapon strike is more accurate, and the attack speed is faster, which requires us to make accurate detection and estimation of target position and moving state in time. At the same time, passive positioning technology with high concealment and strong anti-interference is paid more and more attention to because of the high concealment of modern electronic warfare technology. In this paper, only angle information is used for passive tracking and localization of target, in which particle filter algorithm is used to predict the target state. In the theoretical framework of particle filter algorithm, the author discusses the moving model of the target, the establishment of the observation model of the target, and the realization of the resampling algorithm. Based on the basic particle filter algorithm, an improved particle filter algorithm is proposed, in which intergenerational constraints are added to the original particle filter resampling algorithm. The simulation results show that the improved particle filter algorithm has higher tracking accuracy and can better track and locate the target. Considering the modified spherical coordinate system (MSC), compared with the traditional Cartesian coordinate system, it has the advantages of nonlinear characteristics and rich state variables, which can effectively improve the tracking accuracy. The main body of this paper adopts mixed coordinate system: filter tracking in modified spherical coordinate system and extrapolation of moving state of object in Cartesian coordinate system. Simulation results show that this method can achieve better tracking effect.
【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】：2013
【分类号】：P225.2

【参考文献】