基于变分模态分解的进动目标微多普勒特征提取方法

A Variational Mode Decomposition Based Micro-Doppler Characteristics Extraction Method of Precession-Cone Target

针对在低信噪比条件下微动特征不能有效提取的问题,提出基于变分模态分解(VMD)的微多普勒特征分析方法。该方法对Gabor时频分布结果进行高斯平滑滤波处理并结合VMD将微动曲线分解成若干独立曲线,由分解得到的曲线结合其微多普勒曲线特征来估计目标的自旋频率和锥旋频率。仿真结果表明,该方法能够在低信噪比条件下将目标的微动曲线分解成若干独立曲线,并获得其微动频率。

To address the matter of inadequate anti-noise performance in extracting micro-motion feature,a novel method was proposed based on Variational Mode Decomposition(VMD).The Gabor time-frequency distribution results were processed by Gabor smoothing filter and decomposed into several independent curves by VMD.The spin frequency and cone-spin frequency of the target were estimated by the decomposed curves combined with their micro-Doppler curve characteristics.Simulation results proved that the method could decompose the fretting curve of the target into several independent curves under the condition of low signal-to-noise ratio and obtain fretting frequency.