多普勒计程仪的数据平滑方法

Data smoothing methods for DVL

多普勒计程仪输出数据中有局外数据,可能会导致组合导航系统中的卡尔曼滤波器发散。文章介绍了4种数据平滑方法:动窗平均、动窗加权平均、动窗中值滤波以及动窗稳健平均。仿真结果表明:动窗稳健平均的输出不会被多个局外数据所影响,并且其处理增益接近理论极限。

The outliers contained in the DVL output data might cause the divergence of the Kalman filter of the integrated navigation system. Four data smoothing methods （moving window averaging, moving window weighted averaging, moving window median filtering and moving window robust averaging） are described in this paper. Simulation results indicate that the output of the moving window robust averaging is not influenced by a small fraction of outliers and the processing gain is approximately equal to the theoretical limit.