Research On Cooperative Robust Detection Technology Of Multiple Sound Sources

The multi-static active detection system,which is composed of active sonar mounted on several fixed observation platforms,has the advantages of comprehensive and rapid information processing,and provides a new sonar application mode for improving regional underwater vigilance ability.However,the multi-static active system will not only face the traditional problems such as complex multipath channel interference,Gaussian and impulse noise interference,but also face physical problems such as multi-source echo aliasing and strong direct acoustic signal interference in the multi-transmit-one-receiver detection mode.In order to solve the above problems,this paper studies multi-static sonar biological transmitted signal modeling technology,strong direct acoustic interference suppression technology and robust target echo time delay estimation technology.Firstly,by modeling the dolphin whistle signal,the active sonar transmitted signal library is constructed,and the quasi-biological transmitted signal is used to realize the multi-source echo identification of the multi-static active system.Firstly,the modeling criterion of quasi-biological signal is established by analyzing the characteristics of dolphin whistle signal.In view of the fact that the dolphin whistle signal has the characteristics of nonlinear frequency modulation and multi-harmonics,which makes it impossible for the traditional time-frequency analysis technology to accurately estimate its frequency,the quasi-biological signal modeling method based on second-order synchrosqueezing transform is studied to realize the accurate reconstruction of dolphin signal.Through the analysis of the correlation coefficient and resolution of the reconstructed dolphin whistle signal,the echo identification ability and good target detection ability of quasi-biological multi-source transmitted signal are further explained.Then the DOA estimation method and the suppression method of strong direct acoustic interference based on small-scale array are studied.Firstly,in order to solve the problem that the resolution of the traditional DOA estimation method is low and it is unable to distinguish multiple interference sources under the condition of small array,a gridless DOA estimation method based on atomic l₀ norm minimization is studied,which can realize high-resolution DOA estimation under small-scale array and complex operating conditions.On this basis,in order to solve the problem of the expansion of the spatial spectrum of the signal,the strong direct acoustic interference suppression technology based on null broadening beamforming is studied.At the same time,in order to solve the problem of beamforming performance degradation caused by inaccurate prior information of array structure in practical work,an interference covariance matrix reconstruction method based on spherical uncertain set is studied to improve the output signal-to-interference-to-noise ratio and realize weak target extraction under strong interference background.Finally,the target echo time delay estimation technology which has the ability to overcome multipath channel interference and noise suppression is studied.First of all,in order to overcome the interference of complex multipath channel and wideband Gaussian noise,a target echo time delay estimation technique based on frequency-sliding generalized cross-correlation is proposed,which highlights the arrival time of target echo and suppresses the interference of multipath channel and Gaussian noise.Then,in order to suppress the impulse noise interference in the process of target detection,a target echo time delay estimation technique called l_p-norm based matched filter is proposed,which can realize the robust estimation of target time delay under the background of impulse noise.the effectiveness of the algorithm is verified by simulation analysis and experimental data processing,which can provide signal processing technology reserve for echo identification,undisturbed detection and robust detection in the process of target detection in multi-static active system.