Signal Denoising Of MEMS Vector Hydrophone Based On Joint Denoising Method

signal is the carrier of information,in the process of receiving signals,often mixed with some noise,in order to improve the signal-to-noise ratio,differences between people according to the various characteristics of signal and noise,design a lot of denoising method,and obtained good results in application of the traditional denoising method such as Fourier transformation method,and has some effect in signal denoising,but not be able to process non-stationary nonlinear signal experience modal method and the wavelet analysis method can deal with non-stationary nonlinear signal so this article will research experience modal denoising method of wavelet denoising methodWavelet packet denoising method combined with empirical mode denoising method and wavelet denoising method.Firstly,we introduces the basic principle of Fourier transform method basic principle of wavelet denoising,and then introduces the basic principle of several kinds of empirical mode decomposition,finally introduced the combination of several denoising methods in order to compare several kinds of noise control methods of noise reduction effect,the noise reduction effect evaluation index is established for 2,respectively,the minimum mean square error(mse)signal-to-noise ratio will be a single denoising method were applied to the simulation signal,it is concluded that empirical mode in the denoising MEEMD good denoising effect and then use the chosen denoising signal wavelet denoising and wavelet packet denoising denoising effect evaluation index of the two methods.The results show that the denoising effect of wavelet soft threshold method is better than that of wavelet packet method.Given that the previous conclusions,choose MEEMD decomposition method and the improvement of the wavelet denoising method combined denoising.MEEMD method in this method will be used to dye the noise signal is decomposed into several intrinsic mode functions and innovative to the intrinsic mode function and the coloured noise signal linear correlation analysis,to select the correlation between larger several intrinsic mode functions,characterize choose intrinsic mode function spectrum diagram,the burr is serious intrinsic mode function of improved wavelet soft threshold processing.finally,get new intrinsic mode function reconstruction new intrinsic mode functions,and choose the burr phenomena of intrinsic mode function get signal denoising.This denoising method is compared with the wavelet soft threshold denoising method based on MEEMD decomposition method and the wavelet soft threshold denoising method based on CEEMD decomposition method by simulation,and their two performance indexes and denoising effect pictures are obtained respectively.The results show that the effect of MEEMD decomposition and the improved wavelet denoising method is better.Finally,the denoising method was applied to the measured data of fenhe machine in the laboratory of zhongbei university in 2014.Three signals with frequency of 315 Hz and 630 Hz were selected from the measured data for experiments,and 1000 points of each signal were intercepted.The comparison between the de-noising signal and the original measured signal shows that the time-frequency diagram of the de-noising signal is smoother than that of the measured signal with fewer burrs,so the improved de-noising method has a good effect on the de-noising of the measured signal.