| Detection and location for multiple sources are important division of underwater acoustic signal processing. The traditional methods based on plane wave are unable to be used to locate far underwater targets due to the complicated the manner in which sound propagate in the ocean, while time-reversing and matched-field processing can be used to estimate both the distance and depth of the targets, because the manner of the propagation of the sound wave in the ocean has been taken into consideration. It is hard for conventional methods to locate the weak sources when weak sources covered up by the interference caused by strong sources, so projection time-reversing matched-field processing and matrix filter method are used to deal with this situation. The main contents of the dissertation are as follows.Firstly, the sound field is calculated, and the principle of the time-reversing matched-field processing is analyzed. The wave motion equation of sound propagation and the program for sound field calculation such as Kraken, WKBZ and BDRM are introduced. The rationale of matched-field processing and time-reversing processing and the relationship between them are elaborated. The method of locating underwater target using a horizontal array is proposed.Secondly, the bearing angle and the number of the targets are estimated. Conventional methods of bearing angle estimating based on plane wave assumption and the method of subspace intersection (SI) are described. As the performance of conventional method declines when applied to locating underwater targets, SI shows good characteristic because all the horizontal wave numbers have been taken into consideration. The method based on information criterion for estimating the number of the signal sources is described, and simulation experiment testifies that this method can make the correct estimation when applied to underwater situation.Thirdly, several high resolution methods are introduced, and the influence of the parameter of the array on the performance of the methods is studied by simulation. The projection method and matrix filter method are used to detect and locate the weak sources covered by inference caused by strong ones. Inter-point method is used to calculate the matrix which allows signals from the pass region to pass and rejects the signals transmitted from the stop region. |
