Long-Range Source Localization Using The Flank Array Of An AUV In Shallow Water

Currently, along with the intense coastal local conflicts of countries with coastline, navies’ strategy emphasis begins to shift from deep water to shallow water. Autonomous underwater vehicles’(AUVs’) flexibility and imperceptibility make them more fit for carrying out underwater military missions in the complicated environmental water areas and the forbidden areas. The flank array is the important passive detecting equipment for AUVs. However, due to the limited length of AUVs, the flank array is usually difficult to get sufficient spatial gain. Additionally, the performance of array signal processing methods based on plane-wave hypotheses and other simplified sound propagation models extremely degrades due to the much more complicated sound propagation environment in shallow water, especially when the signal travels under the low frequency and long distance situation. Therefore, passive source localization in shallow water has always been a difficult and challenging subject for researchers. Thus, based on the summary and analysis of existing researches, aiming at the problems mentioned above, the research on how to utilize the flank array of an AUV to passively localize the long-range sources in shallow water is carried out.In recent years, matched field processing(MFP) has been widely studied and used, which involves the sound propagation model much closer to the real ocean environment situation. It actually utilizes the large aperture sonar to collect and deal with the unique information of sound field made from the target of interest in the complicated ocean environment, and then discriminates the target of interest from the interferences and noise background. If the sufficient ocean environment information is given, MFP could rearch a very good estimation performance. It has already gone from theoretical researches and scientific experiments to practical applications. The dissertation firstly gives rational analysis and hypotheses in the number of the hydrophones, element spacing, and deployed depth of the short horizontal linear array, sound propagation model, model parameters, modeling method and characteristics of the source of interest. And then, matched field processors with different resolution and robustness are applied to the array. Finally, a very detailed analysis of the estimation performance based on the array in the two-dimensional ocean environment is provided.According to the theory of MFP and the analysis of matched field localization performance based on the stationary short horizontal linear array, MFP methods usually needs large aperture receiving arrays to satisfy the requirement of spatial gain. For the vertical linear array, when the length of array is long enough to penetrate the whole water layer, the good localization performance could be acquired. But for the horizontal linear array, a longer array is needed to get the same result of the vertical linear array. In order to address this problem, two different methods(coherent or incoherent processing) are proposed in the dissertation. Both of methods try to make up the shortage of spatial gain of the short horizontal linear array by processing the array data collected by the mobile array at different sampling positions. Through the theoretical analysis and simulation experiments, the working conditions, resolution ratio and the ability against to the environmental mismatches and the array position errors of the two proposed methods are studied and provided.Detecting performance of the sonar is seriously affected by the installed platform. According to research results of matched field localization based on the mobile short horizontal linear array, in order to serve the research goal of passive source localization using the flank array of AUVs, the autonomous localization accuracy of maneuvering AUVs should be improved. Thus, the errors of flank array sampling positions can be reduced. The dissertation studies the motion states of AUVs in a detailed way, and then proposes a wavelet-based grey particle filtering algorithm. Through the contrastive analysis using the experimental data, the autonomous localization accuracy maneuvering AUVs based on proposed algorithm is evaluated. The results verify the effectiveness and availability of proposed algorithm.Summarizing the above research results in the dissertation, a three-dimensional passive long-range source localization method in shallow water based on the AUV’s flank array is finally proposed. This method requires setting up several sampling positions in pre-established AUVs’ travelling route. It consists of two steps, processing in spatial domain and estimation in time domain. In order to achieve the preliminary passive localization ability in shallow water using the flank array of AUVs, a high resolution matched field processor is firstly employed to process the flank array data collected at different sampling positions. And then, the matched field localization outputs at different sampling positions are incoherently processed to exploit the characteristics of sound field aroused by the target of interest as soon as possible. This way can also enhance the robustness of the proposed method against to the ocean environmental mismatches and flank array position errors. Thus, the three-dimensional coordinates of the target of interest can be estimated over time in a relatively high probability and accuracy.