改进EKF算法在固定单站被动目标跟踪中的应用  

Application of Improved EKF to Passive Target Tracking in Single Fixed Station

作　　者：王晴晴[1] 鲁汉榕[1] 郭锐[1] 张斌[1] 

机构地区：[1]空军预警学院,武汉430019

出　　处：《空军预警学院学报》2013年第5期359-362,共4页Journal of Air Force Early Warning Academy

摘　　要：针对传统EKF算法线性化误差较大的问题,基于固定单站被动目标跟踪模型,研究了非线性函数线性化展开点及雅可比矩阵取值点对线性化逼近误差的影响,然后对传统EKF算法的线性化展开方式进行优化,提出一种基于中值定理的后向平滑BS-EKF算法.仿真结果表明,与传统EKF相比,该算法在时间复杂度上略有增加,但稳定性、滤波精度和收敛速度有所提高,适用于对定位精度要求较高的情况.Aimed at the problem of greater linearization error of conventional extended Kalman filter （EKF）, this paper studies the impact of linearization expansion points of non-linear function and valued point in Jacobin matrix on the linearization approximation error, based on the model of passive target tracking in single fixed sta-tion, and then brings forward a backward smooth BS-EKF algorithm based on the mean value theorem via opti-mizing the linearization expansion of the conventional EKF algorithm. Simulation results show that the new algo-rithm, in spite of having slightly higher time complexity, outperforms the conventional EKF in stability, filtering accuracy and rate of convergence, which could be applicable to the case of higher positioning accuracy.

关 键 词：中值定理 后向平滑 扩展卡尔曼滤波 固定单站被动目标跟踪 

分 类 号：TN957[电子电信—信号与信息处理]