【摘要】 经验模态分解(EMD)是目前信号去噪中应用较多的一种方法,但处理与噪声时频特征相近的信号时,该算法存在内蕴模态函数(IMF)混叠现象.本文从信号降噪的角度出发,提出基于经验模态分解与小波分析的超声信号降噪方法,首先利用EMD将信号分解为多个IMF分量,通过计算各分量与信号间的互相关系数判断存在模态混叠现象的过渡IMF,从多个IMF分量辨识出噪声与信号的分界,对过渡IMF进行小波去噪,去除过渡分量中的噪声;然后将去噪后的过渡分量IMF与其后续分量进行信号重构,得到去噪后的信号.为了验证所提方法的有效性,本文分别以含噪bumps信号和实际超声信号为例,将该方法与其它4种去噪方法进行了对比.实验结果表明:EMD结合小波法优于单独小波法,而本文方法进一步提高了EMD方法的去噪能力,为EMD去噪方法的改进提供了新思路.更多还原

【Abstract】 Empirical mode decomposition(EMD)is one of the most widely used methods in signal denoising.However,there is a intrinsic modal function(Intrinsic Mode function,IMF)aliasing in the algorithm(the part of the modal function is still a combination of signal and noise)when the signal and noise frequency characteristics are similar.A denoising method based on EMD and wavelet analysis was proposed from the perspective of signal denoising in this paper.EMD was used to decompose the signal into multiple IMF components.By calculating the cross correlation coefficient between each component and the signal,the transition of IMF was used to identify the noise and signal boundaries and the transition of IMF was denoised to remove the noise in the transition component by wavelet.The denoised transition component IMF and its subsequent components were reconstructed to obtain denoised signal.In order to demonstrate the effectiveness of the proposed method,the noisy bumps signal and the actual ultrasonic signal were took as examples in this paper.The method was applied to contrast to the other four method.The numerical simulation and experimental results showed that EMD combined with wavelet method can better single wavelet,and the method in this paper further improved the denoising ability of EMD method and provided a new idea for the improvement of EMD denoising method.更多还原