自适应两级UKF算法及其在时变偏差估计中的应用  

An Adaptive Two-Stage Unscented Kalman Filter and Its Application in Unknown Random Bias

作　　者：程会艳[1,2] 郝云彩[1,2] 熊凯[1,2] 

机构地区：[1]北京控制工程研究所,北京100190 [2]空间智能控制技术国家级重点实验室,北京100190

出　　处：《空间控制技术与应用》2010年第3期33-37,54,共6页Aerospace Control and Application

基　　金：863课题(2007AA12Z325);国家自然科学基金(60702019)资助项目

摘　　要：针对存在时变偏差的非线性系统状态估计问题,提出了一种新型的自适应两级UKF算法(ATUKF,adaptive two-stage unscented Kalman filter).该算法利用衰减因子补偿模型的不确定性,对状态和偏差实施分离估计,并利用估计偏差对状态进行更新.将该算法用于存在随机时变偏差的近地卫星自主导航系统的状态估计问题中,仿真结果验证了其有效性。An adaptive two-stage unscented Kalman filter （ATUKF） is proposed to consider the nonlinear system in the presence of random time-varying bias in a number of practical situations. The ATUKF compensates model uncertainties by using the decay factor, estimating both the state and random bias, and using estimated bias to update state. A near-earth satellite autonomous navigation system with the random bias is simulated by this algorithm to estimate position and velocity of the satellite, the results validate the effectiveness of the ATUKF, and finally autonomous navigation system＇ s observability is analyzed.

关 键 词：非线性滤波 随机时变偏差 可观测性 

分 类 号：V448.2[航空宇航科学与技术—飞行器设计]