Research On Anti-multipath Interference And DOA Estimation Algorithm For Underwater Sonar Target

The complex characteristics of underwater environment greatly limit the effective and stable transmission of acoustic signals.Direction of arrival(DOA)estimation is one of the main research directions of underwater sonar detection.The existing DOA estimation methods have some problems,such as low estimation accuracy and poor adaptability to signal environment.On the basis of the existing research results,this dissertation starts from the theory and application of the anti-multipath interference method and DOA estimation method of underwater sonar target,explores the signal processing methods to improve the signal reception performance and accurate estimation of target azimuth,and verifies the feasibility and effectiveness of the proposed methods through simulation and comparison experiments.The main contents of this dissertation are as follows:(1)The common methods of anti-multipath interference and DOA estimation are introduced.Firstly,the LMS algorithm and RLS algorithm are introduced,and the characteristics of the two algorithms are compared and analyzed;secondly,the subspace method and sparse representation method are introduced,and four typical DOA estimation methods are simulated and compared.(2)An underwater multipath interference suppression algorithm based on sparse p-normlike constraint CVSLMS is proposed.Firstly,based on the model of underwater multipath interference signal processing,the LMS algorithm with sparse p-norm-like constraint is improved by using double sigmoid function to change the adaptive iteration step;secondly,the autocorrelation function of error signal is used to improve the iteration step and adjust the zero attraction term,and the compensation term is introduced to ensure that the algorithm has a faster convergence speed when the error signal is uncorrelated;finally,compared with different LMS methods,the algorithm has faster convergence speed,higher convergence accuracy and tracking The performance and stronger interference suppression ability can recover the original transmitted signal well.(3)A DOA estimation method of cuckoo search based on covariance matrix reconstruction is proposed.Firstly,based on the covariance matching criterion,the recovery of array covariance matrix is realized by Vandermonde decomposition theorem,and a trace minimization problem is established by the reconstruction model of low rank matrix;secondly,the isolated niche technology is applied to the cuckoo search algorithm and the path parameters are improved to solve the optimal solution of trace minimization problem in space,so as to realize the two-dimensional angle of the target;finally,compared with different DOA estimation methods,the proposed method does not need to be meshed,has higher estimation accuracy and resolution ability to adjacent targets,and has stronger adaptability to signal environment.