Ship Hydrodynamic Pressure Signal Detection Based on Neural Network Prediction

Abstract

The ship hydrodynamic pressure signal is generally hard to detect from the rough ocean wave hydrodynamic pressure background signal. An algorithm based on neural network prediction is provided to detect the ship pressure signal from the ocean wave pressure signal. The predictions of hydrodynamic pressure signal are compared with measurements of the same variables to form prediction errors that are used to test for the presence of the ship pressure signal. If the prediction errors are relatively high, it will mean the appearance of a ship hydrodynamic pressure signal. Through the simulation results on the simulated and the measurement data, the algorithm based on feed forward neural network prediction proved its validity.