Forward modeling of seabed logging with controlled source electromagnetic method using radial basis function networks

Abstract

Forward modeling is an important step in processing data of seabed logging (SBL) with controlled source electromagnetic (CSEM) method to determine the location and dimension of a hydrocarbon layer under the seafloor. In this research, forward modeling was conducted using a radial basis function (RBF) network, which is an important type of artificial neural networks. To train this RBF network, a data set was generated using a simulation software: COMSOL Multiphysics. The network designed has 3 layers with 3 neurons in the input layer and 1 neuron in the output layer. The single hidden layer contained neurons whose number had been varied between 1 and 20 neurons. The performance comparison showed that the RBF network with 10 neurons in its hidden layer was the best to model SBL with CSEM method.