Shear wave velocity is a critical physical property of rock, which provides significant data for geomechanical and geophysical studies. This study proposes a multi-step strategy to construct a model estimating shear w...
详细信息
Shear wave velocity is a critical physical property of rock, which provides significant data for geomechanical and geophysical studies. This study proposes a multi-step strategy to construct a model estimating shear wave velocity from conventional well log data. During the first stage, three correlation structures, including power law, exponential, and trigonometric were designed to formulate conventional well log data into shear wave velocity. Then, a geneticalgorithm-patternsearch tool was used to find the optimal coefficients of these correlations. Due to the different natures of these correlations, they might overestimate/underestimate in some regions relative to each other. Therefore, a neuro-fuzzy algorithm is employed to combine results of intelligently derived formulas. Neuro-fuzzy technique can compensate the effect of overestimation/underestimation to some extent, through the use of fuzzy rules. One set of data points was used for constructing the model and another set of unseen data points was employed to assess the reliability of the propounded model. Results have shown that the hybrid genetic algorithm-pattern search technique is a robust tool for finding the most appropriate form of correlations, which are meant to estimate shear wave velocity. Furthermore, neuro-fuzzy combination of derived correlations was capable of improving the accuracy of the final prediction significantly.
Photoelectric factor, formation true resistivity, and formation water saturation are three functional parameters of a hydrocarbon reservoir that could provide invaluable data for reservoir characterization and formati...
详细信息
Photoelectric factor, formation true resistivity, and formation water saturation are three functional parameters of a hydrocarbon reservoir that could provide invaluable data for reservoir characterization and formation evaluation. The present study proposes an improved strategy for making a quantitative formulation between conventional well log data and the mentionewd parameters. At the first stage of this study, three architectures of artificial neural networks, including generalized regression neural network, radial basis neural network, and Bayesian regulation backpropagation neural network, were employed to predict the aforementioned parameters from conventional well log data. Consequently, a committee neural network was constructed by virtue of hybrid genetic algorithm-pattern search technique. The propounded committee neural network combines the results of generalized regression neural network, radial basis neural network, and Bayesian regulation backpropagation neural network to improve the accuracy of final prediction. It assigns a weight factor to each of the individual artificial neural networks indicating its contribution in overall prediction. A set of data points was used for model construction and another set was employed to assess the model performance. The results showed that integration of artificial neural networks using the concept of committee machine could improve the precision of target prediction, although each of the artificial neural networks has performed adequately for prediction of photoelectric factor and formation true resistivity. The values obtained for formation water saturation are not as accurate as results obtained for photoelectric factor and formation true resistivity, although the correlation coefficient between measured and predicted values for formation water saturation is higher.
The scaling equation is the most popular mathematical modeling of asphaltene precipitation as a problematic issue in petroleum industry. There are eight adjustable coefficients in the scaling equation that govern the ...
详细信息
The scaling equation is the most popular mathematical modeling of asphaltene precipitation as a problematic issue in petroleum industry. There are eight adjustable coefficients in the scaling equation that govern the quality of the fit between titration data and the scaling equation model. In this study, a hybrid geneticalgorithm-patternsearch (GA-PS) tool was employed to extract optimal values of the involved coefficients in the scaling equation through the stochastic search. For better performance of the GA-PS tool, dimensionality of the problem was broken into two simpler parts using the divide-and-conquer principle by introducing two fitness functions. The renovated scaling equation was compared with previous works;it was shown that the proposed method outperforms previous works.
Quantitative formulation between conventional well logs and Poisson's ratio, the most critical geomechanical property of reservoir rocks, could be a potent tool for planning and post analysis of wellbore operation...
详细信息
Quantitative formulation between conventional well logs and Poisson's ratio, the most critical geomechanical property of reservoir rocks, could be a potent tool for planning and post analysis of wellbore operations. Direct estimation of Poisson's ratio from conventional well logs makes the problem too complicated. Therefore, the present study proposes an improved multi-step strategy for making a quantitative formulation between conventional well logs and Poisson's ratio. In the first stage, shear wave slowness was predicted from conventional well logs using a radial basis neural network, Sugeno fuzzy inference system, neuro-fuzzy algorithm, and simple averaging method. Consequently, the Poisson's ratio was computed from the results of each expert, independently. Eventually, a committee machine with intelligent systems was constructed by virtue of a hybrid genetic algorithm-pattern search technique. The values of Poisson's ratio, derived from the results of a radial basis neural network, Sugeno fuzzy inference system, neuro-fuzzy algorithm, and simple averaging method, were used as inputs of the committee machine with intelligent systems. The proposed committee machine with intelligent systems combines the results of aforementioned experts for overall estimation of Poisson's ratio from conventional well log data. It assigns a weight factor to each expert, indicating its contribution in overall prediction. The proposed methodology was applied in Asmari formation, which is the major carbonate reservoir rock of Iran. A group of 1,582 data points were used to establish the intelligent model, and a group of 600 data points were employed to assess the reliability of the proposed model. The results show that the committee machine with intelligent systems method performs better than individual intelligent systems, which perform alone.
暂无评论