Study On The Key Techniques Of Underwater Target Parameters Estimation

The modern marine defense system can discover,identify,estimate,and track underwater targets in the early stages by comprehensively using many approaches such as surveillance,detection,and antisubmarine,which plays a crucial role in maintaining maritime security and improving marine defense capabilities of our country.As an indispensable part of the modern marine defense system,underwater target parameters estimation technology has become one of the research hotspots and difficulties in the field of marine science and technology.The underwater target direction-of-arrival(DOA)estimation and underwater target tracking are two fundamental problems in the field of target parameters estimation.Since target DOA estimation provides data for target tracking,it is the basis of achieving high-precision target tracking.In this dissertation,the underwater target DOA estimation technology are firstly studied,and then underwater target tracking is studied on this basis.In general,the main work and contributions of this dissertation are given as follows:(i)Considering the off-grid error caused by the grid mismatch problem,an off-grid DOA estimation algorithm of narrowband signals via the underwater acoustic vector-sensor array is proposed.Firstly,the grid point where the true DOA is located is determined by the coarse estimation process based on the joint sparse reconstruction according to the sparseness of the target in space.Then,the first-order expansion of the Taylor series is utilized to accurately approximate the direction vector of the true DOA to eliminate the off-grid bias and achieve an accurate off-grid error estimation.This algorithm can effectively eliminate the off-grid bias and improve the accurate DOA estimation.(ii)Considering the impact of underwater multipath environment on DOA estimation,a passive DOA estimation algorithm of underwater multipath signals via spatial time-frequency dsitribution is proposed.This method combines the time-frequency analysis with the traditional array signal processing to yield the spatial time-frequency distribution matrix.The multi-ridge detection algorithm is introduced to separate different paths of the multipath signals according to their different time-frequency characteristics and then deal with them separately.This algorithm realizes the accurate estimation of the frequency and DOA of the closely-spaced multipath signals(the DOAs of different paths are overlapped or very close)under passive scenarios and underdetermined cases.(iii)Considering the DOA estimation of wideband signals under the sparse sub-Nyquist sampling,a wideband signal DOA estimation algorithm based on spatio-temporal joint sparse sampling is proposed.This method uses multicoset sampling to perform sparse sampling in the time domain and achieves the estimation of the spectrum utilizing the collected information.An extended coprime array is used for sparse sampling in the spatial domain.By solving an optimization problem based on the joint sparse reconstruction,DOA estimation of wideband signals is obtained under the conditions of spatio-temporal joint sparse sampling.(iv)Considering the limited energy of the tracking system,a bearing-only underwater target tracking algorithm is proposed.This method uses the adaptive node selection method to dynamically select a new cluster head and cluster members according to the information utility,remaining energy,and energy consumption of the nodes at each sampling time and then forms a new tracking cluster to participate in the passive tracking process of the target.Under the framework of the distributed fusion process,the linear minimum variance fusion criterion is used to obtain the optimal fusion state estimation.The algorithm effectively reduces the energy consumption of the system on the premise of the convergence of the tracking error.