基于M估计的抗野值卡尔曼滤波方法

M-estimate Based Kalman Filter with Immunity to Outliers

摘要:

针对标准卡尔曼滤波方法中野值对其滤波精度的影响,提出了基于M估计的野值处理方法。这种方法不仅可以去除单个野值的影响,而且在野值成片出现时还能很好地保持滤波稳定性。通过仿真验证了该方法的有效性。

Abstract:

The influence of the outliers on the Kalman filter is analyzed.A method based on M-estimate dealing with outliers is proposed.The method has the immunity to not only single outlier but also a mass of outliers.Simulation test has validated this method.