Acoustic Localization Of Underwater Mobile Nodes Adapted To Environmental And Platform Conditions

The underwater localization technology has always been a challenge in exploring the ocean and a bottleneck in ocean engineering,and thus has been studied extensively in academia in recent years.Since the propagation ability of electromagnetic waves in the underwater environment is limited,the traditional Global Positioning System(GPS)technology is not applicable underwater.Acoustic waves are the main means of the underwater localization currently.Acoustic propagation is closely related to the characteristics of the ocean environment.The inhomogeneity,time-variance of water and the moving targets make the acoustic target localization problem extremely challenging.Therefore,from the moving target perspective,this thesis combines the characteristics of both the moving nodes and the environment the nodes are in,and proposes theories and technologies for localizing underwater moving nodes.The overall aim is to provide new ways for the practical engineering applications of the underwater localization technology and the further development of the underwater sensing networks.The scenario is to localize the Autonomous Underwater Vehicles(AUVs)by deploying several moving or fixed acoustic buoys over the ocean surface.One of the main focuses is the influence of the Sound Speed Profiles(SSP)on the localization of the moving nodes.The couplings between the acoustic localization and the motion control of AUVs are also considered.Specifically,the main research contents and results of this thesis are as follows:(1)Dolphin Inspired Ambiguity Function for Enhanced Delay-Doppler ResolutionCetacean marine mammals possess excellent ability in discriminating and localizing objects in complex shallow water.Their unique acoustic signals and biological structure may play an important role.In Hangzhou Polar Ocean World,a broadband signal recording system is deployed to record the acoustic signals of bottlenose dolphins.According to the signal classification of the bottlenose dolphins,the click signal is found to be the key in discrimination and localization.In this thesis,the time-frequency characteristics of the click signals are analyzed with the short-time Fourier transform,Wigner–Weyl distribution and reduced interference distribution.Based on the analysis,a dolphin-inspired acoustic signal model is developed,and two nonlinear methods are proposed to enhance the delay-Doppler resolution.Moreover,the detection performance between the two nonlinear-based receivers and the conventional matched filter are compared.The mathematical expressions for the detection probability and false alarm probability are derived,and Monte Carlo simulations are carried out to verify the correctness of the mathematical derivations.Based on the recorded dolphin signals,a real linear frequency modulated(LFM)signal is used to establish a bionic signal model.Then,two ambiguity function images can be generated from the single LFM signals by extracting the complex parts with the Hilbert transform.The product method and the minimum method are applied to further process the two images.Simulation results show that the proposed single LFM based nonlinear processing methods achieve significant improvement of delay-Doppler resolution,with only slightly sacrificing the detection performance.The proposed method can provide high precision measurements for the following Time Difference Of Arrival(TDOA)and Frequency Difference Of Arrival(FDOA)based localization methods.(2)Acoustic Localization with Multilayer Isogradient Sound Speed ProfilesThe speed of sound in the water varies with pressure,temperature and salinity.This inhomogeneity leads to the bending in the propagation of sound rays in water,which makes the existing localization algorithms based on straight-line propagation imprecise.In order to improve the localization accuracy,the influence of SSPs needs to be considered.A multilayer isogradient SSP model is proposed using the linear segmentation approach.Then,the mathematical description of the sound ray tracking problem is presented.Based on the multilayer isogradient SSP model,the TDOA localization algorithm is mentioned.For moving targets,the node localization algorithm based on FDOA is proposed.Furthermore,a joint estimation algorithm with TDOA and FDOA is also proposed.The performance of the proposed localization algorithms is verified and analyzed through a series of simulations.The impact of different SSP modeling errors on the localization accuracy is also discussed.Results show that the algorithms based on the multilayer isogradient SSP model can effectively track the sound rays,and mitigate the errors created by the assumption of the straight-line propagation.Overall,accurate and reliable node localization of moving targets can be achieved.(3)Integrated Acoustic Localization and Tracking Control of AUVsThe influence of ocean currents is one of the main challenges facing the autonomy of AUVs.An integrated acoustic localization and tracking control method for AUVs is proposed in this thesis.The uncertainties involved in the localization and ocean currents are handled together in the framework of the extended Kalman filter(EKF).Firstly,the AUV dynamics considering the uncertain ocean currents are established.Three buoys with known global positions that emit acoustic signals periodically are deployed at the surface.Time of arrivals of these acoustic signals at the AUV are then obtained and used to calculate an estimated position of the AUV.Considering the uncertainties in the ocean currents and the acoustic localization,EKF is applied to correct the AUV states from the dead-reckoning technique.The corrected AUV states are then used in the model predictive control(MPC)based path tracking controller design.MPC is a model based control technique by utilizing the predicted system trajectory over a future time horizon.Simulation results demonstrate that the proposed integrated acoustic localization and tracking control approach outperforms the dead-reckoning based tracking control.The proposed method provides new favors for an integrated localization and control design for underwater sensing networks.It also contributes to the theoretical development and practical applications of both the underwater sensing networks and AUVs.