基于改进CEEMDAN的微多普勒特征提取

Micro-Doppler Feature Extraction Based on Improved CEEMDAN

针对基于完备总体经验模态分解(CEEMDAN)的微多普勒特征提取算法存在本征模态函数(IMF)残留噪声多、信号信息出现晚的缺点,提出基于改进的完备总体经验模态分解(ICEEMDAN)的进动目标微多普勒特征提取的方法。该方法以ICEEMDAN将目标信号分解为IMF,对IMF进行希尔伯特谱分析,从得到的希尔伯特谱中提取出目标的进动周期。仿真结果表明,该方法能有效地克服CEEMDAN算法在特征提取中的缺陷,能够更快、更准确地提取出进动目标微多普勒特征。

The micro-Doppler feature analysis based on complete ensemble empirical mode decomposition(CEEMDAN)has the disadvantages of too much residual noise in the intrinsic mode function(IMF)and late appearance of signal information.In this paper,a micro-Doppler feature extraction method for precession target based on improved complete ensemble empirical mode decomposition(ICEEMDAN)was proposed.In this method,the target signal was decomposed into IMFs by ICEEMDAN,then the IMFs were analyzed by Hilbert transform.Finally,the precession period of the target was extracted from the Hilbert spectrum.Simulation results showed that the method used in this paper could effectively overcome the shortcomings of CEEMDAN algorithm and extract micro-Doppler features of precession target faster and more accurately.