Analysis of Propeller Cavitation-Induced Signal Using Neural Network and Wigner-Ville Distribution

Propeller cavitation noise analysis is necessary since this noise is one of the components of the ship noise signature that would be different for each type of the ship. Therefore, the noise can be used for detection and identification process of a ship. This noise can be simulated by doing the experiment generating one component of the ship noise signature, i.e. the propeller cavitation noise. The experiment is done in the cavitation tunnel of the Indonesian Hydrodynamic Laboratory (IHL), Surabaya, Indonesia to generate this noise. Two techniques, i.e. the Wigner-Ville distribution and the backpropagation neural network, are applied to analyse this noise. The results are compared to examine which technique shows better representation.