Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 1

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
EN
The present research applies six empirical, three statistical, and two soft computing methods to predict water saturation of an oil reservoir. The employed empirical models are ‘Archie (Trans AIME 146(1):54–62, 1942),’ ‘DeWitte (Oil Gas J 49(16):120–134, 1950),’ ‘Poupon et al. (J Petrol Technol 6(6):27–34, 1954),’ ‘Simandoux (Revue deI’Institut Francais du.Petrol, 1963),’ ‘Poupon and Leveaux (1971),’ and ‘Schlumberger (Log interpretation principles/applications, p. 235, 7th printing. Houston, 1998)’; statistical methods are ‘multiple variable regression,’ ‘fine tree, medium tree, coarse tree-based regression tree,’ and ‘bagged tree, boosted tree-based tree ensembles’; and soft computing techniques are ‘support vector machine (SVM)’ and ‘Levenberg–Marquardt (LM), Bayesian regularization (BR), and scaled conjugate gradient (SCG)- based artificial neural network (ANN).’ In addition, log variables are ranked based on their significance in water saturation modeling. To achieve the goals, 521 data points are selected from three wells. Each data point has laboratory-derived core water saturation information and six well log features, such as gamma ray (GR), bulk density (RHOB), sonic travel time (DT), true resistivity (LLD), neutron porosity (φN), and Depth. Statistical indexes, namely regression coefficient, mean squared error, root mean squared error, average absolute percentage error, minimum absolute error percentage, and maximum absolute error percentage, are used to compare the prediction efficiency of study methods. Results show that the empirical models provide exceedingly poor prediction efficiency. Within the study models, fine tree, medium tree-based regression tree; bagged tree, boosted tree-based tree ensembles; fine Gaussian SVM; ANN with LM; and ANN with BR are very efficient predictive strategies. The log ranking reveals that GR and DT are the most important, whereas RHOB and φN are the least vital predictor variables in water saturation prediction.
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.