陀螺随机漂移的抗差Kalman滤波

Application of Robust Kalman Filtering to Gyro Random Drifts

在捷联寻北仪中，陀螺漂移是导致寻北精度下降的主要原因。为了削弱陀螺信号中异常噪声对寻北精度的影响，在小波阈值消噪的基础上，将建立的随机漂移AR（3）模型引入Kalman滤波状态方程，基于抗差估计原理，构造了能够抵制观测异常和动力学模型异常干扰的抗差Kalman滤波算法。测试结果显示，Kalman滤波有效地对陀螺随机漂移进行了最优估计和补偿，进而可以提高寻北仪的精度。

In the strap -down gyro north seeker, gyro drift is a primary factor to degrade orientation accuracy. In order to decrease the influence of abnormal noises existed in gyro signal on the accuracy of north seeker, on the basis of wavelet threshold de - noising, the AR （ 3 ） model of gyro random drift is introduced in the state equation of Kalman filtering, and a robust Kalman filtering is constructed based on robust estimation to resist the influence of measurement outliers and dynamics model error. The test results show that robust Kalman filtering can effectively estimate and compensate gyro random drift and in turn, improve the accuracy of north seeker.