摘要
对于带未知噪声统计的单输出系统,本文提出了一种新的自适应Kalman滤波器,应用现代时间序列分析方法,基于ARMA新息模型的滑动平均(MA)参数的在线辨识,提出了稳态最优Kalman滤波器增益估计的一种新算法,比Mehra的算法简单。同时还提出了辨识滑动平均(MA)模型参数的一种新的自适应Kalman滤波算法。此外,给出了在雷达跟踪系统中的应用,且仿真结果说明了本文算法的有效性。
A new adaptive Kalman filter is presented for the single output system with unknown noise statistics. By a time series analysis, a new and simpler estimation algorithm for the gain of the steady-state optimal Kalman filter is given. A new adaptive Kalman filtering algorithm is also given for identifying the parameters of moving average (MA) model. An application to a radar tracking system is given to show the usefulness of the proposed new algorithms.
出处
《自动化学报》
EI
CSCD
北大核心
1992年第4期408-413,共6页
Acta Automatica Sinica
基金
黑龙江省自然科学基金
关键词
自适应
KALMAN滤波
ARMA
Adaptive Kalman filtering
steady-state Kalman filter gain estimation
AR- MA innovation model
identification
radar tracking system.
