#Colored inversion hampson russell series#

Each individual intelligent system, which is used in the first step of the method, has a power factor and a weight coefficient in the PLCM which is derived by ICA. The PLCM is constructed by imperial competitive algorithm (ICA) to enhance the accuracy of prediction. In the next step, a power law committee machine (PLCM) is exploited to obtain final porosity prediction. At the first step of this study, the porosity is predicted from seismic attributes using MLFN, PNN, RBFN, SVR and adaptive neuro-fuzzy inference system (ANFIS).

In this study, in addition to the use of a set of individual systems, a novel methodology is used for enhancing the accuracy of prediction. It combines the results of a group of intelligent systems and acquires the final output.

Committee machine (CM) is a valuable tool when there are multiple ways to solve a problem (Asoodeh and Bagheripour, 2013, Asoodeh et al., 2014, Kadkhodaie-Ilkhchi et al., 2009). Their results revealed that the support vector regression (SVR) technique has a higher accuracy regarding the neural networks.ДЄccording to aforementioned studies, there are various methods to obtain porosity. Al-Anazi and Gates (2010) predicted porosity in a heterogeneous sandstone reservoir. The results of the CFIS method show better performance in comparison with neural network and individual fuzzy systems. The CFIS is constructed based on genetic algorithm pattern search. (2009) predicted porosity and water saturation using committee fussy inference system (CFIS) on a sandstone oilfield of Iran. In their studies, RBFN and PNN improved the prediction log properties.В Kadkhodaie-Ilkhchi et al. Radial basis function network (RBFN) is used in the same work by Russell et al. (2001) predicted well log properties using multi-layer linear regression and two variations of neural networks: probabilistic neural network (PNN), and multi-layer feed forward network (MLFN). Hampson et al., 2001, Kadkhodaie-Ilkhchi et al., 2009, Khoshdel and Riahi, 2011, Nikravesh and Aminzadeh, 2001, Russell et al., 2002, Russell et al., 2003, Soubotcheva and Stewart, 2006, Trappe and Hellmich, 2000, Wang, 2012, Wong and Nikravesh, 2001). Several studies have been performed on this subject (e.g. In the nonlinear method, the neural networks are the most widely used tools for target prediction. Coefficients of this equation can be derived by a minimizing technique such as the least squares errors. In the linear method, a linear equation between inputs (seismic attributes) and target (rock properties) is assumed. For this purpose, there are two types of methodology: linear and nonlinear transformation between seismic attributes and rock properties. Seismic attribute transform is an indirect study to characterize petroleum reservoirs. Therefore, proper attributes must be extracted from seismic data for attribute transform. This is challenging in carbonates because of strong heterogeneity (Al-BinHassan and Wang, 2011, Al-Moqbel and Wang, 2011). The seismic data analysis for estimating petrophysical properties such as porosity is among the conventional procedures in reservoir evaluation. The results show that PLCM-ICA model improved the results of neural networks, support vector machine and neuro-fuzzy system.

This technique is called PLCM-ICA in this paper. Finally, a power law committee machine (PLCM) is constructed based on imperial competitive algorithm (ICA) to combine the results of all previous predictions in a single solution.

Second, porosity log is predicted from seismic attributes using common intelligent computation systems including: probabilistic neural network (PNN), radial basis function network (RBFN), multi-layer feed forward network (MLFN), Оµ-support vector regression (Оµ-SVR) and adaptive neuro-fuzzy inference system (ANFIS). Acoustic impedance is an important attribute that is obtained by the SCI method in this study.

#Colored inversion hampson russell series#

First, a series of sample-based attributes is calculated from 3D seismic volume. For this purpose, seismic colored inversion (SCI) and a new approach of committee machine are used in order to improve porosity estimation. Because there is strong heterogeneity in carbonate formations, estimation of rock properties experiences more challenge than sandstone. The method focuses on one of the Iranian carbonate fields in the Persian Gulf. In this paper we propose a new method for predicting rock porosity based on a combination of several artificial intelligence systems.