Nearfield noise identification with a passive array in the underwater anechoic basin at KRISO

摘要

Unmanned underwater vehicles (UUVs) are used for various underwater missions. During the UUVs' missions sound waves play an important role because acoustic communication and positioning systems are used for the control of the UUVs. However, they do not only have numerous mechanical components but also use hydraulic and electrical systems, which make loud noise and degrade the performance of acoustic equipment. Therefore, we have to examine the noise characteristics before using the acoustic equipment. The investigation method may include localization of the noise sources by using a line array of hydrophones. In order to apply the conventional beamforming algorithm, experiments should be performed in the anechoic environments because it is derived based on the assumption of no reflection. Although KRISO/KORDI has an underwater anechoic basin, it has a free surface that reflects sound waves. Because the acoustic impedance of water is several thousand times bigger than that of air, it is reasonable to assume that it is pressure-released condition. In this paper a source localization method in such a kind of environment is studied. The effects of phase errors are analyzed for plane and spherical wave models. Simulation results say that the estimated position error is negligible if the aperture size is 2.5 times bigger than the wavelength.