改进的EEMD算法在时域航空电磁信号降噪中的研究

常规降噪方法在应用于时域航空电磁信号降噪时需根据噪声情况人为进行参数调整，自适应性较差。总体经验模态分解（EEMD）算法对非线性、非平稳信号处理具有良好的自适应特性，传统的EEMD算法进行噪声抑制是将高频本征模态分量滤除，将低频分量重构得到降噪信号，这种方法易失掉高频分量中的有效信号。本文提出一种改进的EEMD降噪算法，应用于时域航空电磁信号的处理。该方法结合时域航空电磁信号的衰减特性，将信号EEMD分解后得到本征模态分量，其中包含信号和噪声，经Savitzky Golay平滑滤波，再将高频部分进行阈值去噪，最后得到干净的本征模态分量进行重构。实验结果表明在输入信号信噪比小于等于15 dB的情况下,输出信噪比能够提高12 dB左右,在抑制噪声的同时保留了更多有效信息。 

Research of Improved EEMD Algorithm for Time domain Airborne Electromagnetic Signal De noising

Common noise reduction methods need to artificially adjust the parameters according to the noise situation when applied to time domain airborne electromagnetic noise reduction,adaptability was poor. The ensemble empirical mode decomposition (EEMD) has good adaptive characteristic for nonlinear, non stationary signal processing,the traditional signal denoising method based on EEMD reconstructed the denoised singal by combining the low frequency components and filtering out the high frequency components, this approach was easily losing effective signal in high frequency components. This paper proposed an improved EEMD noise reduction method, applied to the time domain airborne electromagnetic signal processing. This method combined with the attenuation characteristics of the time domain airborne electromagnetic signal,decomposed the signal into intrinsic mode function which contains signal and noise by EEMD, then through the Savitzky Golay smoothing filtered and threshold denoised the high frequency part, finally reconstructed the denoised signal by the clear Imf components. Experimental results show that in the case of the input signal noise ratio less than or equal 15 dB,the output signal noise ratio can be increased by about 12 dB, this approach not only reduces noise but also retains information more effectively.