用时频峰值滤波去除多普勒频率估计野值点的方法

Method of Removing the Estimating Outliers of Doppler Frequency Using Time-Frequency Peak Filtering

研究去除多普勒频率估计野值点的方法.在分析多普勒频率估计野值点的成因和特点的基础上,对含有野值点的多普勒频率估计按比例调整到适当的频率范围内,用时频峰值滤波方法对其进行处理,恢复原多普勒频率比例尺度,从而达到去除野值点的目的.实验结果表明,选择较小窗长的伪Wigner-Ville分布,一次迭代就可有效地去除多普勒频率估计中的野值点,对后续的脱靶量参数估计没有任何不利影响,方法简单可靠.

A method of removing the estimating outliers of Doppler frequency is studied. Based on the analysis of reasons of bringing out the outliers and characteristics of the outliers, Doppler frequency is scaled within appropriate limits. On using time-frequency peak filtering in processing the scaled Doppler frequency and recovering the original scale, the outliers are successfully removed, based on the pseudo WVD with the small data window length. Experimental results showed that the outliers of Doppler frequency may be removed effectively with only one iteration, and also that this is helpful to the successive estimation of miss distance parameters. The given method is simple and feasible.