چكيده به لاتين
Relative permeability is one of the essential properties of the reservoir rock, which is required in almost all calculations of multiphase flows in porous environments. The exact amount of the relative permeability, makes possible to reservoir engineers to evaluate the performance of the reservoir,the amount of final utilization and the improvement of EOR techniques. Therefore, obtaining accurate information about relative permeability is very important and has always been considered in the oil industry. Several laboratory methods have been designed to determine relative permeabilities. Although these methods are most accurate to determine relative permeability, but doing thier is very costly and time consuming. Consequently, computational methods have always been considered and various experimental relationships have been suggested to estimate this parameter, but most of the which are not sufficiently correct. In recent years due to the development of computational software and intelligent tools, this tool is used to model complex engineering systems. In this thesis, using intelligent tools such as artificial neural networks, support vector regression and fuzzy systems to predict the two phase relative permeability in the oil-water and oil-gas systems. The data used to develop the models collected from the articles and after preprocessing such as integration of data and outlier detection, randomly, 80 percent of the data was selected as training data and 20 percent of the data was selected as testing data and used. For the development of all models, several parameters have been investigated and some of the parameters in the model have been optimized with classical and intelligent methods such as genetic algorithm. Finally, a hybrid model of the three intelligent methods is presented to estimate the two-phase relative permeability. The error rate in this model is lower than that of the experimental and each other intelligent methods,and the regression coefficient obtained 0.94 for the oil-water system and 0.97 for the oil-gas system in the test data
