Research On Target Detection Technology Based On Circular Vector Array

Underwater target detection is an important part of underwater acoustics.It is of great significance to the detection sonar based on the underwater preset platform.A circular array is generally used instead of a linear array due to the size limitation of the platform.And vector hydrophones are favored because they can receive more information and have directivity independent of frequency.Therefore,this paper studies the target detection technology based on circular vector array.First,according to the actual background of the underwater preset platform,the signal and array are modeled.Meanwhile,the basic parameters of the array are analyzed.And the basic performance of CBF,MVDR and MUSIC on vector circular arrays is studied through theoretical and simulation analysis.Then,according to the application environment of the underwater preset platform,the rotation problem of the underwater platform is analyzed.And on this basis,the position compensation algorithm based on the convex optimization space rotation matrix is studied,and it is compared with the position compensation algorithm based on the two-step space rotation matrix.The basic idea of these two algorithms is to assume that the target's position in the geodetic coordinates remains unchanged in a short time.Then,they use the rotation matrices that are off-line designed to compensate the influence of the steering vector caused by the rotation of the array.The array position compensation algorithm based on the two-step spatial rotation matrix only keeps the output within the desired range unchanged,while it cannot suppress the interference in other directions.But the matrix occupies a small memory.The array position compensation algorithm based on convex optimization space rotation matrix uses convex optimization to calculate the rotation matrices.Due to different interference intervals,the rotation matrix is also different,which resulting in a larger memory.However,it can suppress the interference in other directions.Simulation experiments and lake test data processing verify the effectiveness of the two algorithms.Finally,aiming at the problem of missing target track and many false alarm points in other directions caused by low-SNR weak target and clutter,two types of weak target detection methods based on spatial spectrum are studied.One is the target detection based on image processing.The track enhancement algorithm based on adaptive growth is studied.It establishes a growth criterion for track enhancement for the problem of missing target track.And Hough transform are used to track detection.Then the robustness of target detection is improved.The other is target detection based on tracking algorithms.The tracking algorithm studies the Rao-Blackwellized particle filter and the probability hypothesis density filter based on Gaussian mixture.These two algorithms use appropriate mathematical models and filtering techniques to estimate the target state,to fill in discontinuous points and remove false alarms to complete the target track detection.