摘要
本文研究了自适应卡尔曼滤波技术在惯导系统中的应用。在噪声统计特性未知或近似已知的情况下，采用常规卡尔曼滤波会导致较大的状态估计误差，甚至使滤波发散；而自适应卡尔曼滤波在估计状态的同时，利用观测数据带来的信息，可在线估计噪声的统计特性，从而不断地改进滤波器的设计，由此得到的滤波估计比常规卡尔曼估计精度更高。本文采用Ｓａｇｅ 和Ｈｕｓａ 自适应滤波算法，结合惯导初始对准，给出了计算机仿真。仿真结果进一步证实在噪声统计特性不确切知道的情况下，自适应卡尔曼滤波的估计精度高于常规卡尔曼滤波的估计精度。
The application of adaptive Kalman filter technique in inertial navigtion system is studied in this paper.In case of the noise statistic characteristic being unknown or not completely known,there would exist a large error in estimating state using conventional Kalman filter,even made the filter diverge. However,adaptive Kalman filter can,at the same time of estimating state,estimate the statistic characteristic of noise online using observation information.Therefore the design of the filter can be continuously improved and the obtained estimation using adaptive Kalman filter possessed higher accuracy compared with that of conventional Kalman filter.The paper gives the simulation of initial alignment,which adopted Sage & Husa adaptive Kalman filter algorithm.The validity of the proposed means is proved by the simulation result.
出处
《中国惯性技术学报》
EI
CSCD
1999年第3期15-17,共3页
Journal of Chinese Inertial Technology
基金
国防"九五"预研项目
关键词
惯性导航
自适应卡尔曼滤波
初始对准
inertial navigation
adaptive Kalman filter
initial alignment
