Font Size: a A A

Acoustic Scattering Characteristics And Feature Extraction Of Underwater Spherical And Cylindrical Objects

Posted on:2021-02-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y S WuFull Text:PDF
GTID:1368330605979491Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
The detection and identification of underwater“silent”targets is one of the important research directions in the field of underwater acoustics.The scattering echo obtained by active sonar system contains the information of the target,such as the shape,size,thickness,material and so on.Extracting the features associated with these inherent attributes from the scattering echo is the key to realize the target detection,classification,and parameter estimation.According to the target acoustic scattering characteristics,the extraction of acoustic scattering features,the classification of acoustic scattering components and the estimation of target size parameters were studied in this dissertation.For free-field spherical shells,the normal series solutions for the acoustic scattering from targets with different thickness-to-radius ratios were calculated,and the simulated echo signals were obtained.For the acoustic scattering of the cylindrical shell with one hemispherical end-cap and one flat end-cap,which cannot be analyzed with the normal series solution,the free-field acoustic scattering model was built based on the COMSOL Multiphysics.The target scattering characteristics under different incident angles were calculated,and the corresponding simulation acoustic scattering echoes were given.For the buried target,the COMSOL acoustic scattering model of buried spherical shell was also established,and the suppression effect of the burial layer on the subsonic anti-symmetric Lamb?A0??was studied.All above provide the theoretical basis for the subsequent studies.The acoustic scattering characteristics of spherical shells with different thickness-to-radius ratios are different.For the classification of the extremely thin,the thin and the thick shells,the method of constructing feature space to realize target classification by using the auditory perceptual features which can reflect the target acoustic scattering echo waveform structure and spectral distribution characteristics was studied.The separabilities of different shells under different features were analyzed by using the ratio of the inter-and inner-class distance ratio and then proved by simulation.For the thin shell,the radius and thickness can be estimated according to the midfrequency enhancement.The effect of each A0? wave on the midfrequency enhancement of the acoustic scattering spectrum was studied and two estimation methods of the target size were proposed.The first one is to extract the time delay and spectral peak frequency of the first A0? wave with fractional Fourier transform,and estimate the target radius and thickness according to the existing expressions.The second one is to extract the resonant peak interval within the midfrequency enhancement range of the acoustic scattering spectrum to estimate the target radius,the corresponding expression was given.The geometrical acoustic scattering and elastic acoustic scattering have different generating mechanisms,and the separation and classification of them is the premise of the target parameter estimation.For the separation and classification of the scattering components of a free-filed thin shell,a Wigner-Ville distribution?WVD?based target echo processing method was put forward.The WVD of the scattering echo was rotated and the cross-term was removed by filtering and energy weighting in the transform domain.Then the energy and the instantaneous frequency feature were used to detect the scattering components,the time delay and the spectrum of them were also extracted.According to the midfrequency enhancement effect of elastic echo of thin shells,the feature space composed of spectral peak frequency,3dB bandwidth and spectral centroid was established to classify the geometric scattering echo and elastic one.The experimental data processing results verified the proposed method,and the size and thickness of the target were estimated by using the extracted information of the geometrical and elastic echo.Elastic acoustic scattering is the key to buried target detection and recognition.Based on COMSOL model and dispersion curve,the scattering characteristics of buried stainless steel sphere were analyzed.Aiming at the resonance peak feature extracting of elastic acoustic scattering echo under buried condition,a de-chirping based elastic signal extracting method was studied.The elastic resonance was enhanced by extracting the elastic echo,and the experimental data processing results verify the effectiveness of this method.According to the echo processing results of the experimental solid sphere scattering echo and the acoustic scattering theory,it was verified that the resonance peak detected under the free field condition is the Rayleigh wave resonance peak,while under burial condition the Rayleigh wave is suppressed and the detected elastic resonance peak is brought by the Whispering-Gallery wave.For sphere radius estimation under both cases,the relationship between the corresponding resonance characteristics and the radius were studied respectively.The Rayleigh resonance interval was proposed to estimate the radius of the sphere in free-field condition,and the Whispering-Gallery resonance was put forward to the burial condition,the corresponding calculation expressions were given and verified.
Keywords/Search Tags:Underwater target detection, object acoustic backscattering, elastic scattering, feature space, target size estimation
PDF Full Text Request
Related items