首页|期刊导航|空天预警研究学报|一种自适应鲁棒的Sage-Husa无迹Kalman滤波算法

一种自适应鲁棒的Sage-Husa无迹Kalman滤波算法OA

An adaptive robust Sage-Husa unscented Kalman filter algorithm

中文摘要英文摘要

针对目标跟踪中无迹Kalman滤波(UKF)在初值不准、模型失配下精度受限,以及Sage-Husa 无迹Kalman滤波(SH-UKF)联合估计状态与量测噪声时易发散的问题,提出自适应鲁棒的SH-UKF(ARSH-UKF)算法.该算法包含四项改进:自适应遗忘因子、有界多重次因子、主辅协同的分级发散判据和基于实际残差协方差Huber 鲁棒权.仿真结果表明,当初值严重失配时,ARSH-UKF算法的合成位置和速度的均方根误差(RMSE)相对UKF 和SH-UKF算法均有降低,初值接近真值时与UKF 和SH-UKF算法精度相当,兼具高精度与强鲁棒性.

Aimed at the problem of the limited accuracy of unscented Kalman filter(UKF)under inaccurate initial values and model mismatch,and the divergence of the Sage-Husa unscented Kalman filter(SH-UKF)when the state and measurement noises are estimated jointly,an adaptive robust SH-UKF(ARSH-UKF)algorithm is proposed,which incorporates four improvements:adaptive forgetting factor,bounded multiple sub-factors,a hier-archical divergence criterion with primary and auxiliary coordination,and Huber robust weight based on the actu-al residual covariance.Simulations results show that under severely mismatched initial values,both the root mean square errors(RMSE)of the synthetic position and velocity of the ARSH-UKF algorithm have been reduced com-pared with the UKF and SH-UKF algorithms,and that when the initial value is close to the true one,the accuracy is comparable to that of the UKF and SH-UKF algorithms and the ARSH-UKF algorithm has both high accuracy and strong robustness.

宋亚伟;叶泽浩;韩伟;颜建

空军预警学院,武汉 430019空军预警学院,武汉 430019空军预警学院,武汉 430019空军预警学院,武汉 430019

信息技术与安全科学

目标跟踪Sage-Husa无迹Kalman滤波自适应鲁棒

target trackingSage-Husa unscented Kalman filter(Sage-Husa UKF)adaptive robustness

《空天预警研究学报》 2026 (4)

247-252,6

10.3969/j.issn.2097-180X.2026.04.003

评论