Research On Signal Denoising Method Of MEMS Vector Hydrophone Based On Optimization Algorithm And VMD

The natural environment and marine environment are constantly changing and developing,so that the signals received by the micro-elecortomechanical system（MEMS）vector hydrophone are usually mixed with complex noise,which seriously affects the target positioning and recognition.Aiming at this kind of phenomenon,this paper proposes a research on denoising method of MEMS vector hydrophone signal based on optimization algorithm and Variational Mode Decomposition（VMD）.The main research contents are as follows:（1）A joint denoising method based on optimized VMD and wavelet packet denoising is proposed.This method combines the dynamic differential simulated annealing algorithm and the VMD algorithm to adaptively obtain the optimal parameters （the number of intrinsic mode functions）and （quadratic penalty factor）,and then use the wavelet packet to reduce Denoising the signal by denoising.The simulation experiments and the measured data experiments collected by the National Defense Key Laboratory of North China University in2014 show that the method can effectively remove the noise in the signal,and has a certain effectiveness.（2）（2）A denoising method of MEMS vector hydrophone based on optimized VMD,compressive sensing and wavelet threshold is proposed.In this part,the improved locust algorithm is used to optimize the parameters of VMD,and then the pure component and the noise component are distinguished according to the cross-correlation coefficient and center frequency of each natural mode component.Thresholding with wavelet,and finally reconstruction.The simulation comparison experiments and the measured data experiments of the National Defense Key Laboratory of North Central University in 2011 and 2014 both show that the method can remove noise while retaining the basic characteristics of the signal,which lays the foundation for further azimuth estimation.This paper proposes two methods for denoising MEMS vector hydrophones based on the combination of optimization algorithm and VMD.Both methods can effectively remove noise,have certain advantages and practicability,and are beneficial to target localization and identification.