Adaptive Extraction Of Geometrical Acoustic Scattering Signal Structure Of Underwater Target

How to accurately estimate the geometrical acoustic scattering signal structure of the target from active sonar echoes is one of the most important problems in the underwater acoustic technology.When using a linear chirp signal as the transmitting signal,the echoes are composed of several signals with the same chirp characteristics of the transmitting signal according to the target echoes highlight model theory.Time-frequency analysis is used to analyze the linear chirp signal,and it has the unique advantage in underwater target echoes analysis and feature extraction.In this paper,the acoustic scattering characteristics of typical underwater targets are studied firstly.Based on the theory,the delay in echoes highlight model of cylindrical shell with a ball coronal head is deduced.Based on this,the WVD and ambiguity functions of single highlight echo and multi-highlight echoes of underwater target are deduced,and the characteristics of highlight echo auto-components and cross-components are described.The causes of cross-components of time-frequency analysis and the deficiency of time-frequency analysis of fixed kernel function are discussed.What's more,a method to design the kernel function-radially Gaussian kernel(RGK)time-frequency distribution is studied in this paper.According to this method,based on the application background of the active sonar to the underwater target detection,the adaptive radial Gaussian kernel time-frequency analysis method using the prior information of the transmitted signal is used to extract the geometric acoustic scattering signal structure.This paper focuses on the application of the adaptive radial Gaussian kernel time-frequency analysis method using the prior information of the transmitted signal to the extraction of geometrical acoustic scattering signal structure of underwater targets.This method not only shows good time-frequency performance in acoustic scattering signal structure extraction,and avoids the cross-components in the WVD,but also having a good anti-noise ability,effectively improve the signal to noise ratio,which is very important in underwater environment.In addition,the target echoes may be disturbed by reverberation.Based on the mechanism of reverberation,the ambiguity function(AF)of the target echoes and reverberation is proposed,and the feasibility of using the anti-reverberation method is analyzed.Finally,the experimental data in the pool and lake are processed,which validates the rationality and validity of the adaptive radial Gaussian kernel time-frequency analysis method for geometrical acoustic scattering signal structure extraction.