Research On Moving Target Tracking Based On Joint Array Shape And DOA Estimation

With the development of ocean engineering,autonomous underwater vehicles(AUV),unmanned surface vehicles,and other vehicles with strong mobility and high autonomy are rapidly evolved and widely utilized.The towed array sonar carried by vehicles can independently detect the designated area,greatly expanding the detection range of the ocean.However,the shape of the towed array is prone to distortion due to the influence of the ocean current and its own motion,resulting in a serious decline in the performance of orientation estimation.To resolve the above problems,this thesis mainly carries out the research on target tracking of short towed array,focusing on using acoustic information to estimate array shape and Direction of Arrival(DOA).And for the uncertainty of target orientation estimation,the Bernoulli filtering algorithm is considered to improve the DOA tracking performance of towed array sonar.Aiming at the DOA estimation problem of short towed array sonar,this thesis firstly investigates the traditional DOA estimation algorithm and utilizes spatial filtering and wavenumber filtering to suppress radiated noise.In order to improve the DOA estimation performance,a joint estimation algorithm of array shape and DOA based on Sparse Bayesian Learning(SBL)is used.Furthermore,considering the array shape mismatch problem of the joint estimation algorithm,four array models are proposed to effectively expand the scope of application of the algorithm and improve the generality.Aiming at the low efficiency of the joint estimation algorithm,an acceleration strategy combined with an improved SBL algorithm is proposed.Finally,adopting the idea of spatial spectrum search,a joint estimation algorithm based on the graph Fourier transform is proposed to improve the robustness of DOA estimation in small snapshots.In order to improve the performance of the DOA estimation in the uncertain ocean environment,this thesis then investigates the Bernoulli filtering algorithm,which can effectively filter out the spatial spectrum clutter caused by the ocean noise and improve the accuracy of the target DOA of the short towed array.To solve the model mismatch of target motion,an improved Bernoulli filter combined with the Interacting Multiple Model(IMM)algorithm is proposed.Also,on the basis of the improved Bernoulli filter algorithm,a weighting factor is introduced to make better use of the DO A information to further improve the performance of DO A tracking.Finally,the simulation result shows that compared with the traditional DOA estimation methods,the improved towed array shape and direction of arrival joint estimation algorithm,the joint estimation method based on graph Fourier transform and the weighted Bernoulli filtering algorithm are more robust and efficient,and the practicability of the algorithm is verified by the MAPEX2000 sea trial experiment.