摘要
天波超视距雷达跟踪目标时电磁波是通过电离层的折射传播的,因而导致在地理坐标系下的量测方程中存在强非线性,而采用传统的EKF(Extended Kalman Filter)实现的跟踪算法,在非线性方程的线性化中舍去了含强非线性的二阶以上的高阶项,导致目标的跟踪精度较低;提出采用UKF(Unscented Kalman Filter)方法处理超视距雷达系统在跟踪算法中的强非线性问题。UKF算法有效降低了非线性方程中的舍入误差,可确保三阶以上的精度。仿真结果表明UKF滤波算法较EKF算法提高估计精度。
The measurement equation is nonlinear in the Over-the-Horizon-Radars （OTHR） target tracking system due to the wave is propagated by the inospheric layers. Especially tracking the maneuvering target based on the traditional EKF （Extended Kalman Filter） , the output precision is still poor for discarding the over 2rd order terms in the linearizing process. A new application of the unscented Kalman filter to the OTHR tracking system is represented. The state distribution is approximated by Gussian variable random （GVR） by a set of carefully chosen sample points, and they are propagated through the state and measurement equation and ensure accuracy estimation to the 3rd order term of Taylor series expansion. The simulation results show that the UKF exceeds the standard EKF （Extended Kalman Filter） in tracking the maneuvering target.
出处
《计算机测量与控制》
CSCD
2005年第10期1094-1095,1111,共3页
Computer Measurement &Control
基金
国防科技"十五"预研
国家自然科学基金项目(60404011)西北工业大学引进高层人才科研启动资助项目和青年创新基金和校英才计划项目。
关键词
超视距雷达
UKF滤波算法
目标跟踪
over-the-horizon radar
unscented Kalman filter
target tracking
