基于地磁定轨和扩维卡尔曼滤波的导航算法

Navigation Algorithms Using Magnetometers and Augmented Kalman Filtering

为了克服扩展卡尔曼 (Kalman)滤波算法对噪声统计特性的约束 ,针对磁强计量测噪声为有色噪声伴常值干扰的特性 ,提出了一种基于扩维Kalman滤波算法的地磁导航算法 .该算法为了使量测噪声白噪化 ,引入了 2个新的状态变量 ,此时测量噪声是均值为 0的高斯白噪声 ,并满足扩展Kalman滤波算法的约束条件 .对由此构成的系统使用扩展Kalman滤波算法 ,可以得到扩维Kalman定轨导航算法 ,利用该算法可以获得国产磁强计的导航精度 ,其地心距模的估计误差为 2 0km ,速度模的估计误差为 10m/s .卫星的实测数据仿真结果表明 ,该导航算法具有较好的稳定性和收敛性 。

In order to overcome the noise restriction of the extend Kalman filtering, an augmented navigation algorithm is established. As the magnetometer's measurement disturbance includes color noises with constant disturbance, based on satellite orbit motion equation with earth-fixed coordinate, two new state variables for white noise shaping are introduced in this algorithm, and the extended Kalman filtering is used to form the navigation model for the magnetometer-based satellite navigation system. The results of testing augmented Kalman filtering with actual magnetometers are presented. It is validated that this algorithm owns better reliability and convergence. The position estimation errors can be converged within 20 km, and the velocity estimation errors can be converged within 10 m/s. It can be concluded that the algorithm solves the divergence problem of extend Kalman filtering.