激光陀螺漂移数据的自适应Kalman滤波

Application of adaptive Kalman filter in process of laser gyro drift data

针对Sage-Husa自适应Kalman滤波算法存在不能同时估计系统过程噪声方差和量测噪声方差的问题,结合激光陀螺漂移数据的特点,设计了2种改进的Sage-Husa自适应Kalman滤波算法:系统过程噪声时变的自适应滤波算法和系统过程噪声与量测噪声统计特性分开估计的并行Sage-Husa自适应滤波算法。仿真结果表明:所述方法改进算法能够有效提高数据精度,且对系统的初值不敏感。

According to the characteristics of laser gyro drift data, two innovations are presented to avoid the deficiencies of Sage-Husa adaptive Kalman filter. The first innovation simplifies Sage-Husa adaptive Kalman filter to estimate system process noise only. The second is parallel Sage-Husa adaptive filter which estimates the statistical features of the unknown time-varying system process noise and measurement noise separately. The result of simulation shows that two algorithms are useful and effective with high precision.