Studies On Underwater Source Localization With Deficient Environmental Parameters

Passive localization techniques of underwater source has been a hot and difficult issue in the field of underwater acoustic.Approaches to source depth estimation are especially limited.A chassic way to solve this problem is by means of Matched Field Processing(MFP),whose practical performance strongly depends on the accurate establishment of the underwater acoustic environmental model.However,the marine environment is spatially and temporally fluctuating,which makes three-dimensional observation of the marine environment difficult to achieve.Therefore,the underwater source localization is usually launching based on the historical data observation and single point measurementon on the spot of the marine environment,which makes the localization performance could not be guaranteed.In response to this issue,this thesis studies the impact of environmental parameter mismatches on the performance of MFP localization to provide a reference for the acquisition of important environmental parameters during sea trials;on the other hand,based on the studies of underwater sound propagation,this thesis alse explores the estimation techniques of underwater source speed,range and depth with deficient environmental parameters.The main study contents and innovations are as the follow:1.Basic theory of Normol Modes and Wavenumber Integration Techniques are introduced;one of the numerical solution method of Wavenumber Integration Techniques,Fast Field Program,is discussed;modal expension of the Green’s Function is studied;theoretical and numerical studies about the relationship between sound field and Green’s Function are produced.2.The impacts of environmental mismatches on the performance and results of coherent and incoherent broadband MFP localization are evaluated by numerical simulations.The results show that the water depth mismatch and the sound velocity profile mismatch have the greatest impact on MFP performance,and incoherent broadband MFP localization error is proportional to the water depth mismatch.The broadband MFP is conducted by using the 21-element vertical linear array data of the open-access SWellEx-96 trial.The results verify the conclusion of the simulation experiment.3.Based on the studies of underwater sound field interference,making use of the proportional relationship between sound field cross-correlation in horizontal direction and the radial velocity of source,one approach of source velocity estimation with no need of a priori environmental information is studied.The simulation experiment with a received SNR not less than 3dB and only one single receiver and single-frequency signal processing achieves good estimation of source speed.Using the experimental data,the estimated source radial velocities are in good agreement with the GPS records,and estimation performance to within 10% of the real range is predicted.4.The Green function retrieval technique with known source distance is studied.Make use of the source depth information in Green function,a source depth estimation approach is studied.Furthermore,the joint estimation of source velocity,range and depth is proposed.Under the condition of single receiver and single-frequency signal processing,numerical simulation obtained accurate estimate of source speed,range and depth.Experimental data processing also obtains a good depth estimation whose relative error does not exceed 10% of the water column depth.