APPLICATION OF ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM FOR PREDICTION OF POROSITY FROM SEISMIC ATTRIBUTES; CASE STUDY, FAROUR.A OIL FIELD, PERSIAN GULF, IRAN

摘要

Shiri, Y., Moradzadeh, A., Shiri, A. and Chehrazi, A., 2011. Application of adaptive neuro-fuzzy inference system for prediction of porosity from seismic attributes; case study, Farour.A oil field, Persian Gulf, Iran. Journal of Seismic Exploration, 20: 177-192. Reservoir characterization using seismic attributes has a great impact on quantitative and qualitative interpretation of subsurface property in petroleum industry. Among linear and nonlinear predicting tools like Multi-Regression, polynomial curve fitting and Neural Networks, methods based on Neuro-Fuzzy technique known as the Adaptive Neuro-Fuzzy Inference System (ANFIS) which is a hybrid intelligent system recently has attracted the attention of researchers in many academic, industrial, scientific and engineering areas. In this study, data set was 2D seismic and petrophysical well log data in the Farour.A oil field. First of all, by applying seismic inversion, broad band acoustic impedance as the most relevant seismic attribute to porosity was extracted from these data. Then, optimum numbers of relevant seismic attributes were selected by using stepwise regression and cross validation techniques. At the end, three types of neural network and ANFIS were applied for porosity prediction from seismic attributes. Results were shown that predicting porosity from seismic attributes by ANFIS was performed fast-converged and high accuracy against three types of neural networks.