Research On Reverb Suppression Algorithm In Active Sonar Detection

With the development of underwater detection technology,sonar is an important means of underwater target detection.According to the working principle,sonar can be divided into active sonar and passive sonar.Active sonar is better than passive sonar in ranging and direction finding accuracy,so it has been the focus of attention in the past decades,and has been used in a variety of fields.Reverberation is a special interference that affects the detection performance of active sonar,but it is always a difficult problem to suppress it effectively because of its characteristics.Traditional sonar generally adopts artificial method for judgment,but it has poor effect and high requirements on sonar personnel.Automatic detection and processing methods have attracted people’s attention,and constant false alarm detection is one of them.At present,most of the research on constant false alarm processing is carried out for radar application,so how to apply constant false alarm processing to sonar becomes particularly important.The main work of this thesis is to study the application of constant false alarm in high frequency sonar.The research contents are as follows:(1)Before the processing of constant false alarm,beamforming technology is introduced,including conventional beamforming method,minimum variance without distortion algorithm and multiple signal classification algorithm.STMU algorithm is proposed to improve the problem that the multi-signal classification algorithm can not distinguish the target under the condition of low signal-to-noise ratio and the fast change of the target source azimuth.STMU algorithm can better resolve the target under the condition of low signal-to-noise ratio.It can also achieve better resolution of the target under the condition that the rapid change of the target source azimuth makes the width of the main lobe larger and thus leads to the reduction of signal-to-noise ratio.(2)Two types of traditional constant imaginary police methods are highlighted,among which the mean-value constant imaginary police method,the cell-averaged constant imaginary police method,the cell-averaged selected large constant imaginary police method and the cell-averaged selected small constant imaginary police method are discussed,and the selected-order statistical method,the selected-order selected large constant imaginary police method and the selected-order selected small constant imaginary police method are discussed.The basic principle and implementation process of each method are analyzed,and the performance,advantages and disadvantages of each method are analyzed in the uniform Gaussian background,multi-target environment and at the reverberation edge.(3)The constant false alarm algorithm is applied to the actual high resolution sonar data processing,and the performance of the unit average constant false alarm,the selection sequence statistics constant false alarm,the selection sequence selection large constant false alarm and the selection sequence selection small constant false alarm is studied under K-distribution environment,and the simulation analysis is carried out in the K-distribution uniform environment and the multi-target environment.The above methods are applied to lake test data processing to verify the relevant conclusions.