EKF和UKF算法在无线传感器网络目标跟踪中的应用  

EKF and UKF algorithm and their applications to WSN-based target tracking

作　　者：史岩[1] 朱涛[1] 傅军[1] 张亚宁[1] 

机构地区：[1]海军工程大学导航工程系,武汉430033

出　　处：《自动化与仪器仪表》2013年第4期131-133,226,共3页Automation & Instrumentation

基　　金：总装装备预研基金(9140A09050112JB11112)

摘　　要：目标跟踪技术随着无线传感器网络的发展,已经成为倍受重视的研究课题,而目标跟踪技术中的一个核心部分就是滤波算法。扩展卡尔曼滤波(EKF)和无迹卡尔曼滤波(UKF)是重要的非线性滤波方法。本文介绍了EKF和UKF的基本思想和算法实现步骤,并对无线传感器网络下的目标跟踪进行了仿真。结果表明,UKF算法的跟踪性能优于EKF算法,但UKF算法计算复杂,跟踪时间长,实时性差。With the development of wireless sensor network, target tracking has become a very active topic. Filtering is one of the core parts of target tracking. Both extended Kalman filtering （EKF） and unscented Kalman filtering （UKF） are important nonlin- ear filtering methods. The basic idea and algorithm description of EKF and UKF are introduced in this paper. Simulations about them in wireless sensor network （WSN） target tracking are done. Results showed that the UKF algorithm outperforms the EKF algo- rithm, but the EKF algorithm suffers from high computational complexity.

关 键 词：传感器网络 扩展卡尔曼滤波 无迹卡尔曼滤波 目标跟踪 

分 类 号：TN957[电子电信—信号与信息处理]