Research For Model Construction And Imaging Technology Based On Multi-Beam Synthetic Aperture Sonar

At present,the Multi-Beam Echo Sounder(MBES),the Side-Scan Sonar(SSS)and Side-Scan Synthetic Aperture Sonar(SSSAS)are recognized as the most effective tools in imaging sonar field.Among of them the MBES has the advantage in getting the high-resolution bathymetry data and achieving the full-scan detection,but the along-track resolution of MBES is determined by the size of beam footprints,due to this,the accuracy of underwater target detection and the quality of sonar imaging are limited by the low along-track resolution.The SSS and SSS AS can provide the relief map of the seafloor,especially the SSSAS it can use a small physical aperture to obtain high along-track resolution by signal processing,but because of the sonar model they can not get the below gap data.In order to improve the quality of acoustic imaging,get the high-resolution bathymetry data and achieve the full-scan detection,this thesis combining with the MBES and SSSAS technologies propose the Multi-Beam Synthetic Aperture Sonar(MBSAS)technology which based on the China-Made MBES.The specific study includes:First,this thesis propoes the theory simulation sonar mode of MBSAS.The MBSAS needs to combine the MBES technology and SSSAS technology,so it needs a new sonar model which can be used by these two technologies.But it will have a large scale if the sonar model is combined directly.In order to solve this problem,this thesis uses the multi-aperture,frequency diversity,single phase center multiple beams and other technologies to improve the MBSAS Sonar model design and get a smaller sonar model.Second,this thesis studies the 3D data simulation of MBSAS and proposes the 3D shape recovery method.Due to the high cost of the real experiments and the high complexity of entironment setting,the 3D data simulation of MBSAS is more and more important.Firstly,the theiss studies the conventional MBSAS 3D data simulation.Then,the results show that the conventional MBSAS 3D data simulation can be divided into four parts which are composition,shadow calculation,delay echo simulation and imaging.Among of them the composition relies very much to the design,and the shadow calculation needs a lot of computer resources that make the conventional Multi-Beam Synthetic Aperture Sonar 3D data simulation very tedious.Therefore,this thesis propoes a new method that uses the SSSAS imaging picture superficial gray data to compute the target curvature,and call this method as Shape From Shading algorithm(SFS).The SFS algorithm replaces the conventional MBSAS 3D data simulation composition and shadow calculation steps into one step and solves the problems of them.At last computer simulation results prove the effectiveness of the new method.Third,this thesis studies the space domain imaging algorithm of MBSAS.Firstly,the problem of MBSAS grating lobes is studied.In order to solve the along-track grating lobes problem this thesis has proposed the multi-aperture,frequency diversity,single phase center multiple beams and other technologies methods and proves them through the computer simulation.Then,the thesis focuses the attention to the MBSAS imaging algorithm.The thesis proposes the space domain dot-by-dot algorithm which is the most precise MBSAS imaging algorithm.The thesis uses the simulation results of Muli-beam Synthetic Aperture Sonar to campair with MBES and SSSAS and finds the advantages of MBSAS.Finally,the thesis aims to large computation of dot-by-dot algorithm problem and proposes the line-by-line imaging algorithm.Truthly,the line-by-line imaging algorithm can solve the large computation problem but it is limited by the uniform along-track sampling and the accurate carrier motion error estimation.Therefor,the thesis proposes the MBSAS space domain Fast Factorized Back Projection(FFBP)imaging algorithm which based on the dot-by-dot imaging algorithm.It uses the sub-aperture technology to reduce the computation time,the simulation proves the effectiveness of this method.Fourth,Basing on the reclassification of the carrier motion error,the thesis proposes the pretreated before imaging processing and the pretreated after imaging processing which is called n-order Auto Regressive(AR)Kalman filter Beam Displaced Phase Center emulation algorithm and sub-aperture imaging registration emulation algorithm.The MBSAS has strict requirement about the along-track path.Firstly,the theses focuses on the yaw,lateral displacement and roll motion error effecting to the SSSAS and analyses the feathers of them.Then two new methods which uses the n-order AR model,kalman filter,beamforming algorithm,DPC and sub-aperture imaging registration algorithm are proposed to solve the estimation accuracy of yaw,lateral displacement and roll problems.At last the computer simulation has proved the correctness of the methods.Fifth,the anechoic tank experiments of MBSAS is introduced.Two kinds of China-Made MBES systems are used to prove the feathers of MBSAS,and the data processing results of anechoic tank experiments show the performance of MBS AS underwater targets detection and effective exraction.Then on the basis of the anechoic tank experiments the key technology related to MBSAS is discussed.Campairing with the MBES and SSSAS the anechoic tank experiments of MBSAS verify the validity of algorithm and prove the feasibility of engineering application.