Research On Simulation Method Of Target Signal Uncertainty In Passive Sonar

This thesis focuses on the uncertainty of passive sonar target signal and its simulation method.It mainly completes the uncertainty simulation of ship radiated noise and ocean channel,and proposes a GPU parallel computing based uncertainty simulation system optimization,which improves the authenticity of passive sonar target signal simulation and improves the calculation efficiency of the uncertainty simulation system.The uncertainly simulation method of ship radiated noise is studied from three aspects: stationary continuous spectrum simulation,line spectrum simulation and modulation spectrum simulation.Aiming at the uncertainty of the broadband continuous spectrum,a simulation method of continuous spectrum based on the three-parameter model and empirical formula of ship radiated noise level under the conditions of ship speed and tonnage change is designed;Aiming at the uncertainty of the line spectrum,a simulation method of undulation line spectrum based on the random degree model is proposed;Aiming at the uncertainty of the modulation spectrum,a simulation method of non-uniform modulation characteristics based on the sub-band modulation depth estimation is proposed.According to the variation law of the sound velocity profile,seabed topography,seabed bottom quality and other factors,the research on the uncertainty simulation method of ocean channel propagation is carried out.Aiming at the uncertainty of sound propagation caused by the change of sound velocity profile,a sound velocity profile disturbance model based on uniform distribution and normal distribution is designed by analyzing multiple sets of measured sound velocity profile data.The simulation of the acoustic propagation uncertainty caused by sound velocity profile disturbance is realized based on the Monte Carlo method.This thesis also analyzes the uncertainty caused by the change of seabed topographical factors such as sloping seabed and seabed hills,and simulates the ocean channel propagation loss based on the ray model.This thesis designs a uniformly distributed seabed sediment change model,including sedimentary layer sound velocity,density and absorption coefficient.The simulation of the acoustic propagation uncertainty caused by the change of seabed sediment is realized based on the Monte Carlo method,and the sensitivity of various seabed parameters to the sound field disturbance is analyzed.A GPU-based passive sonar target signal uncertainty simulation system framework is proposed,and a GPU-based parallel simulation module for ship radiated noise and ocean channel uncertainty is designed and implemented.Compared with the traditional CPU-based simulation framework,the speedup is comparable up to 4?10 times.