Underwater target localisation and estimation of ocean environmental parameters using a genetic algorithm
article
In the field of underwater acoustics the signal processing technique "Matched Field Inversion" (MFI) is an important research topic. This technique can be used for determining both target location and parameters describing the ocean environment. For example, these parameters are needed as input into models that are used on board Dutch navy vessels and for mine hunting operations. When using MFI, the acoustic field that is measured with a sonar array is compared with the acoustic field that is predicted by an acoustic propagation model for a certain set of the unknown model parameters, such as target location and geo-acoustic ocean bottom parameters. An energy function providing a measure for the similarity between the two acoustic fields is defined. By minimising this energy function, the set of input parameters corresponding to the maximum similarity between both acoustic fields, i.e. the solution, is found. Due to the extremely large number of possible parameter value combinations and the occurrence of local minima, global optimisation methods are required to solve this inverse problem. The developed MFI techniques are applied to shallow water experimental data. A genetic algorithm is used for the global optimisation.
TNO Identifier
529110
Source
Nederlands akoestisch genootschap(november), pp. 37-46.
Pages
37-46