组合导航系统多重衰减因子自适应估计算法比较研究  

Comparative Research on Multiple Fading Kalman Filter in Integrate Navigation System

作　　者：耿延睿[1] 郭伟[2] 崔中兴[1] 

机构地区：[1]北京航空航天大学,北京100083 [2]空军第二航空学院,长春130022

出　　处：《中国惯性技术学报》2004年第2期18-22,共5页Journal of Chinese Inertial Technology

基　　金：国防"十五"预研项目资助(413220403)

摘　　要：提出了多重衰减因子自适应估计卡尔曼滤波方法,用该方法对系统每个误差状态估计进行控制,提高滤波器的估计性能。仿真结果表明,新算法在系统噪声特性不准确的情况下,能够抑制卡尔曼滤波估计的发散,GPS/SINS组合导航精度比强跟踪滤波估计的精度高。这种算法推导形式简单,计算量小,适合在线运算。A new adaptive estimation of multiple fading Kalman filter is proposed, which can control each error status estimation of the system and improve the effect of filtering. The simulation result shows that the proposed method can restrain the filtering divergence when system noise attributes are not accurate, and has better effect than that of strong tracking filter. The new method derivation is simple, and the computation burden is low. This is adapted for the online calculation.

关 键 词：组合导航系统 卡尔曼滤波 白适应滤波 强跟踪滤波器 自适应估计算法 衰减记忆滤波算法 

分 类 号：TN967.2[电子电信—信号与信息处理]