Stable nonlinear predictive operator based on neural network, genetic algorithm and controlled gradient method

Kobrunov, A.I.; Приезжев, И.И.

doi:10.1190/segam2015-5720421.1

Cited by 4 publications

(3 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The study focused on sensitivity analysis of operational parameters and how their interactions affected the co-optimization process. Several studies in the petroleum industry have focused on the use of neural network (NN) concept in optimization processes [34][35][36][37][38][39]. This approach is based on trial and error.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimum design of CO2 storage and oil recovery under geological uncertainty

et al. 2017

View full text Add to dashboard Cite

This paper presents an integrated numerical framework to co-optimize EOR and CO 2 storage performance under uncertainty in the Farnsworth Unit (FWU) oil field in Ochiltree County, Texas. The framework includes a field-scale compositional reservoir multiphase flow model, an uncertainty quantification model and a neural network optimization process. The reservoir flow model has been constructed based on the field geophysical, geological, and engineering data. Equation of state parameters were tuned to achieve field measured fluid properties and subsequently used to predict the minimum miscible pressure (MMP). A history match of primary and secondary recovery processes was conducted to estimate the reservoir and multiphase flow parameters as the base case for analyzing the effect of recycling produced gas, infill drilling and water alternating gas (WAG) cycles on oil recovery and CO 2 storage. A multi-objective optimization model was defined for maximizing both oil recovery and CO 2 storage. The uncertainty quantification model comprising the Latin Hypercube sampling, Monte Carlo simulation, and sensitivity analysis, was used to study the effects of uncertain variables on the defined objective functions. Uncertain variables include bottom hole injection pressure, WAG cycle, injection and production group rates, and gas-oil ratio. The most significant variables were chosen as control variables to be used for the optimization process. A neural network optimization algorithm was utilized to optimize the objective function both with and without geological uncertainty. The vertical permeability anisotropy (Kv/Kh) was selected as one of the uncertain parameters in the optimization process. The simulation results were compared to a scenario baseline case that predicted CO 2 storage of 74%. The results showed an improved approach for optimizing oil recovery and CO 2 storage in the FWU. The optimization model predicted that about 94% of CO 2 would be stored and most importantly, that this increased storage could result in about 25% of incremental oil recovery. The sensitivity analysis reduced the number of control variables to decrease computational time. A risk aversion factor was used to represent results at various confidence levels to assist management in the decision-making process. The defined objective functions were shown to be a robust approach to co-optimize oil recovery and CO 2 storage.

show abstract

Section: Introductionmentioning

confidence: 99%

“…This approach is based on trial and error. Mathematical details and the robustness of NN optimization technology were discussed in the literature [23,37].…”

Section: Introductionmentioning

confidence: 99%

Optimum design of CO2 storage and oil recovery under geological uncertainty

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Для обучения нейронных сетей мы используем гибридные алгоритмы обучения (Kobrunov A., I. Priezzhev, 2015, Kobrunov A., I. Priezzhev, 2016, основные на сочетании методов генерализации первых слоев для глубоких сетей, генетических и градиентных алгоритмов методов оптимизации решений. Основная идея предлагаемой технологии -комбинировать стохастические и детерминированные подходы при построении прогнозирующего оператора на этапе обучения.…”

Section: рисунок 1 многослойный персептронunclassified

Productivity Parameters Prediction on Maps Using Neural Networks

Egorov¹,

Приезжев²,

Schelkunov³

et al. 2017

Proceedings

View full text Add to dashboard Cite

SUMMARYThe presentation shows the technique and examples for predicting the oil and gas productivity parameters on the map on the basis of a deep neural network with hybrid training and Tikhonov regularization. The results of predicting the effective thickness in continental facies of Western Siberia are shown. The results of comparison between prediction maps obtained by the neural network technique and multidimensional regression are also shown. The advantages of a neural network are efficiency, higher resolution and better correlation coefficient with well data.

show abstract