Full text loading...
-
Artificial Neural Networks Approach for Well Loging Data Interpretation
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 2nd EAGE St Petersburg International Conference and Exhibition on Geosciences, Oct 2006, cp-20-00107
- ISBN: 978-90-73781-64-1
Abstract
Geoscientists started use artificial neural networks since early 90’s. But the most popular net configuration is preceptron error back propagation network. This paper we discus hierarchical Kohonen network and hybrid opposite direction propagation networks approach for same petrophysical problem decision, like lithology detection and porosity, permeability calculations. For porosity calculations, in case of representative core collection, we use core samples for network training, but in case of none-representative core collection we use model data for network training, the core data used only to control model reliability.