SEISMIC ATTRIBUTES SELECTION AND POROSITY PREDICTION USING MODIFIED ARTIFICIAL IMMUNE NETWORK ALGORITHM

Abstract

In petroleum field, well logs such as porosity, permeability, resistivity, etc. play a major role in hydrocarbon reservoir studying, but these logs require expensive and slow measurement methods so that seismic attributes can be used as predictors to predict the logs depending on the already measured logs. This research presents a proposed modification for Artificial Immune Network (aiNet) algorithm to achieve better performance. The modified algorithm was used to select the best combination of attributes that leads to the best possible prediction by maximizing Correlation Feature Selection (CFS) objective function that leads to low correlated attributes between each other and high correlated with the porosity. The prediction is accomplished by using Artificial Neural Network learned by the modified aiNet.

Keywords

• Artificial immune system
• Correlation feature selection
• Feed-forward artificial neural network
• Immune network theory
• Optimization