Modelling of Sound Transmission Loss in the Baltic with a Genetic Algorithm

摘要

The sound transmission loss in shallow water is influenced by the geoacousticparameters of the seabed. One way of determining these parameters might be by using a theoretical forward model and match the results to actual transmission loss measurements with known water conditions (sound velocity profile). With the accurate parameters as input, the model can then be used to predict the sound transmission for given water conditions. This report deals with seabed parameter matching with the aid of a Genetic Algorithm, an iterative process which imitates selection. A range-dependent model (only pressure waves) as well as a range-independent model (both pressure and shear waves) have been used as forward models. The measurements were made in Kalmarsund in the Baltic 1988 and 1992. The investigation emphasizes the need of a computational forward model which can handle range-dependent geometries even with the seabed modeled as a solid.