摘要
针对初始对准的精度直接影响系统工作性能问题,在初始对准中引入扩展卡尔曼滤波算法(EKF)和Unscented卡尔曼滤波算法(UKF)。EKF算法将非线性系统线性化,是在非线性系统中常用的一种滤波方法;UKF算法没有把非线性系统近似为线性系统,而是使用确定性样本的方法处理非线性问题,使采样点的均值和方差完全符合实际的非线性系统的均值和方差。仿真结果显示,使用Unscented卡尔曼滤波时惯性导航系统初始对准的精度和稳定性都好于普遍使用的扩展卡尔曼滤波算法。
The performance of inertial navigation systems is directly influenced by the accuracy of initial alignment, so the extended Kalman filtering（EKF） algorithm and the unscented Kalman filtering（UKF） algorithm were drawn into initial alignment, EKF, which linearizes the nonlinear systems, is the primary filtering method, while UKF does not regard nonlinear systems as linearization, but it takes a deterministic sampling approach to deal with the nonlinear systems. Thus the sample points completely capture the true mean and covariance of the nonlinear systems. Simulation results show that UKF can yield better performance of stability and higher accuracy than EKF.
出处
《长江工程职业技术学院学报》
CAS
2009年第2期40-43,共4页
Journal of Changjiang Institute of Technology
关键词
初始对准
EKF
UKF
非线性滤波方法
initial alignment
extended Kalman filtering
unscented Kalman filtering
nonlinear filtering method
