摘要
针对常规卡尔曼滤波由于噪声的统计特性与实际情况不相符而引起滤波误差增大的问题,提出了一种新的在线估计系统噪声和量测噪声的自适应滤波算法。新算法通过新息序列自适应量测噪声,对Sage-Husa滤波算法进行改进以估计系统噪声,该算法在噪声统计特性未知的情况下能进行滤波计算。最后对改进的新算法与常规卡尔曼滤波算法作了对比试验分析,结果表明了改进算法的有效性。
A new adaptive Kalman filter algorithm is presented aiming at figure out accretion on the error of Kalman filter when process noise covariance matrix（Q） and measurement noise covariance matrix（R） are incorrectly with actual noise.The algorithm estimates R based on innovation sequence of measure and Q through an improved Sage-Husa adaptive filter,which can keep the filter work steadily and correctly when Q and R are unknown.Practical test on the new approach and the conventional Kalman filter algorithm comparably is given,the results demonstrate that the new algorithm is valid when applied in SINS/GPS integrated navigation system.
出处
《测控技术》
CSCD
北大核心
2011年第3期114-116,共3页
Measurement & Control Technology
关键词
卡尔曼滤波
自适应滤波
新息序列
Sage-Husa滤波
组合导航
Kalman filtering
adaptive filtering
innovation sequence
sage-husa filtering
integrated navigation Building the Model of IMU in Flight Simulation
