Research On Self-noise Interference Characteristics Analysis And Suppression Method Of Small Underwater Maneuvering Platform

As a typical example of Underwater maneuvering small platforms,Underwater unmanned vehicle(UUV)has the features of concealment,flexibility and intelligence.However,the noise generated by the underwater unmanned vehicle itself,as a strong interference in the near field,has a stronger energy than the attenuated far-field target signal,which has an impact on the detection performance of the sonar array mounted on the UUV.In order to solve this engineering problem,this paper takes the lead to analyze the interference characteristics of UUV self-noise,and on this basis,studies the appropriate method to realize the suppression of UUV self-noise interference.In this paper,a vector array signal model with near and far field sound source has established.Aiming at the problem of self-noise of small underwater maneuvering platform,the time-frequency characteristics and spatial characteristics of UUV self-noise were analyzed by using the test data with only near-field UUV self-noise,and the propagation path of UUV self-noise in the stern was modeled and simulated geometrically.Next,from the aspect of beam domain,the zero notch beamforming algorithm and zero notch broadening beamforming algorithm were studied.By synthesizing many beams with zero gain,the interference source is just in the position of zero gain,so as to effectively suppress the interference signal.Simulation results show that both algorithms can effectively suppress UUV self-noise,and the zero trap broadening beamforming algorithm has better performance than the zero trap beamforming algorithm in the case of fewer array elements.Then,from the perspective of matrix filter,the spatial matrix filter of pseudo-inverse criterion,platform self-noise response suppression criterion and platform self-noise zero response constraint criterion is studied,and the signal in a certain region of space is selectively allowed to pass through without loss,so that the information in the unexpected region is eliminated.Compared with conventional beamforming,the three methods in this study all have better performance in suppressing UUV self-noise interference,and through the comparison of output signal-to-stem ratio of matrix filter,it can be known that the normalized global mean square error of platform self-noise zero-response constraint matrix filter is minimum,and the interference suppression effect is better than the other two matrix filters.Finally,we start with the time domain adaptive filtering algorithm,which can dynamically adjust the filtering model according to the characteristics of UUV self-noise and suppress the influence of UUV self-noise interference.The principle of adaptive filter was analyzed,and then the simulation and performance analysis were carried out for LMS algorithm with fixed step size,G-SVSLMS algorithm with variable step size and G-SVSLMS algorithm with improved correlation characteristics.The improved G-SVSLMS algorithm has smaller steady-state error and faster convergence rate.The result of experimental data processing proves the superiority of the improved algorithm.