The accuracy of an artificial neural network ann algorithm is a crucial issue in the estimation of an oil fields reservoir properties from the log and seismic data. Artificial neural network and inverse solution method for. Basically both are neural networks but one which recurrently executing its layers is reservoir computing while the one with feed forward approach is simple neural network. Performance evaluation of artificial neural network approaches in forecasting reservoir inflow. In this paper, we presented two approaches for modeling of survival data with different degrees of censoring. A case study from western onshore, india soumi chaki, akhilesh k. In the suggested model, multireservoir operating rules are derived using a neural network from the results of simulation. For multi reservoir operating rules, a simulation based neural network model is developed in this study. Reservoir modeling uses all available information which includes at a minimum logs data, and fluid and rock properties. Predicting permeability from porosity using artificial neural. Reservoir properties from well logs using neural networks. In petroleum applications, the methodology developed can be a substitute for objects basedalgorithms when facies geometry and reservoir continuity are too complex to be modeled by simple object such as channels. The proxy model is generated by training an arti cial neural network ann. Reservoir systems operation model using simulation and.
Applying machine learning algorithms to oil reservoir production optimization mehrdad gharib shirangi stanford university. Each processing node behaves like a biological neuron and performs two. A montecarlo simulation study was performed to compare predictive accuracy of cox and neural network models in simulation data sets. Nsl is an objectoriented language offering objectoriented protocols applicable to all levels of neural simulation. A feedforward network approximates a mathematical function, whereas rnns approximate dynamical systems. Spiking neural networks snns are artificial neural networks that more closely mimic natural neural networks. Reservoir parameter estimation using a hybrid neural network. In the suggested model, multi reservoir operating rules are derived using a neural network from the results of simulation. Predicting permeability from porosity using artificial.
The present study aims at the application of the hybrid model, which consists of artificial neural network and fuzzy logic in the reservoir operating policy during critical periods. Journal of petroleum science and engineering, elsevier, 2014, insu01084932, 123, pp. Rating is available when the video has been rented. Oil reservoir simulation, artificial neural networks. Then, a suitable neural network architecture is selected and trained using input and. A neural net can be learned to collect multiple point statistics from various training images, these statistics are then used to generate stochastic models conditioned to actual data. In this context, we have developed an artificial neural network based model to predict macroporosity of sandstones. Download interactive neural network simulator for free. Machine learning in reservoir production simulation and. Whitacre t and yu x a neural network receiver for emmwd baseband communication systems proceedings of the 2009 international joint conference on neural networks, 18121816 alavi a, cavanagh b, tuxworth g, meedeniya a, mackaysim a and blumenstein m automated classification of dopaminergic neurons in the rodent brain proceedings of the 2009. Stochastic reservoir simulation using neural networks trained. Machine learning in reservoir production simulation and forecast.
One is the pantai pakam timur field, located in northern sumatra, indonesia, where the data from only two wells were available and the other is iwafune oki field, located in the sea of japan, eastern japan, where wells were concentrated in the central part of. The idea is that not all neurons are activated in every iteration of propagation as is the case in a typical multilayer perceptron network, but only when its membrane potential reaches a certain value. Application of artificial neural networks for calibration. Numerous research works among which some related books e. The idea is that neurons in the snn do not fire at each propagation cycle as it happens with typical multilayer perceptron networks, but rather fire only when a membrane. The simulation model itself can be a useful tool in allocating effort and expense in determination of reservoir fluid and rock data.
Shahabs book mohaghegh, datadriven reservoir modeling, 2017, the topdown model. Predicting reservoir water level using artificial neural. Auckland university of technology, auckland, new zealand fields of specialization. Artificial neural network modeling of dissolved oxygen in reservoir. The book describes how to utilize machinelearningbased algorithmic protocols to reduce large quantities of difficulttounderstand data down to actionable, tractable quantities. Neurovis an interactive introduction to neural networks. The deep learning textbook can now be ordered on amazon.
The fundamental building block of a neural network is the neuron. Dynamical systems are essentially functions with an added time component, the same. In essence the neuron is simply a model for a multivariate function whose input variables are weighted by a weight vector. See more ideas about artificial neural network, ai machine learning and deep learning. Current state of reservoir simulation and modeling of shale. Generate field productivity maps by integrating reservoir simulation data with analytics production analysis access all subsurface data pertaining to each field and cluster information. The book inspires geoscientists entrenched in first principles and engineering concepts to think. Neural computations such as artificial neural networks ann have aroused considerable interest over the last decades, and are being successfully applied across a wide range of problem areas, to domains as diverse as medicine, finance. The available porosity and permeability data needed to build a reservoir simulation model are old and sparse. Special issue on neural network applications to reservoirs. Resermine is a decision making platform for surveillance.
How to tune this variable and what is the impact of this one on the reservoir. For multireservoir operating rules, a simulationbased neural network model is developed in this study. Voltages recorded from the output of the two layer reservoir network. Artificial neural network modeling of dissolved oxygen in.
In petroleum applications, the methodology developed can be a substitute for objects basedalgorithms when facies geometry and reservoir continuity are too. To this end, new automated procedures are established. Machine learning applied to 3d reservoir simulation. Here, each circular node represents an artificial neuron and an arrow represents a connection from the output of one artificial neuron to the input of another. T1 geochemical equilibrium determination using an artificial neural network in compositional reservoir flow simulation.
Optimal operation of multi reservoir system using dynamic programming and neural network h. Physicsbased models and data models tahar aifa to cite this version. Simulating reservoir operation using a recurrent neural. Additionally, a reasonable and effective reservoir operating plan is essential for realizing reservoir function. Im exploring reservoir computing more precisely, echo states network. Levenbergmarquardt training algorithm was used for training a neural network architecture with one hidden layer and thirty hidden neurons. The water quality of reservoirs is one of the key factors in the operation and water quality management of reservoirs. One is the pantai pakam timur field, located in northern sumatra, indonesia, where the data from only two wells were available and the other is iwafune oki field, located in the sea of japan, eastern japan, where wells were concentrated in the central part of the field. Development and application of reservoir models and. The proposed hybrid model fuzzy neural network fnn combines the learning ability of artificial neural networks and the transparent nature of fuzzy logic. Reservoir characterization from 3d seismic data using.
Optimal operation of multireservoir system using dynamic programming and neural network h. This paper demonstrates the use of the k fold cross validation technique to obtain confidence bounds. Accelerating physicsbased simulations using neural network. Historically, the most common type of neural network software was intended for researching neural network structures and algorithms. The online version of the book is now complete and will remain available online for free. Neural computations such as artificial neural networks ann have. To overcome this problem, in this study, a backpropagation neural network is trained to approximate the simulation model developed for the chennai city water supply problem. Abstract a combined approach of a dynamic programming algorithm and artificial. The rate of censorship in each of these models was considered from 20% up 80%.
Neural computations such as artificial neural networks ann have aroused considerable interest over the last decades, and are being successfully applied across a wide range of problem areas, to domains as diverse as medicine, finance, engineering, geology and physics, to problems of complex dynamics and complex behaviour prediction, classification or control. This paper presents a study aimed at forecasting water level of reservoir using neural network approaches. Fundamentals of higher order neural networks for modeling and simulation. Arti cial neural network as a proxy arti cial neural network ann is a. This allowed direct simulation of the trained neural network to obtain an updated reservoir parameters. A new approach to reservoir characterization using deep. To explore the application of a deep learning algorithm on the field of reservoir operations, a recurrent neural network rnn, long shortterm memory lstm, and. Applying machine learning algorithms to oil reservoir. Accelerating physicsbased simulations using endtoend neural. Recurrent networks o er more biological plausibility and theoretical computing power, but exacerbate the aws of feedforward nets. Optimal operation of multireservoir system using dynamic. Available well logs and cores were used as inputs to the hybrid model.
Machine learning in reservoir production simulation and forecast serge a. We develop a proxy model based on deep learning methods to accel erate the simulations. Novel connectionist learning methods, evolving connectionist systems, neurofuzzy systems, computational neurogenetic modeling, eeg data analysis, bioinformatics, gene data analysis, quantum neurocomputation, spiking neural networks, multimodal information processing in the brain, multimodal neural network. Monte carlo simulation and artificial neural network are applied to two areas for predicting the distribution of reservoirs. Dissolved oxygen do in water column is essential for microorganisms and a significant indicator of the state of aquatic ecosystems. Reservoir systems operation model using simulation and neural. What is the realitionship between deep learning methods and. Lens the light, efficient neural network simulator 2. We develop a proxy model based on deep learning methods to accelerate the simulations of oil reservoirsby three orders of. Fundamentals of higher order neural networks for modeling and. A spiking neural network considers temporal information. In many cases, complex simulation models are available, but direct incorporation of them into an optimization framework is computationally prohibitive. These rnns are useful because they have superior theoretical computational power.
In this study, a feedforward neural network with backpropagation learning algorithm was used. Reservoir simulation is an area of reservoir engineering that, combining physics, mathematics, and computer programming to a reservoir model allows the analysis and the prediction of the fluid behavior in the reservoir over time it can be simply considered as the process of mimicking the behavior of fluid flow in a. Geochemical equilibrium determination using an artificial. The development of spiking neural network simulation software is a critical component enabling the modeling of neural systems and the development of biologically inspired algorithms. There is a range of artificial neural network architectures designed and used in various fields. What is the realitionship between deep learning methods. Performance evaluation of artificial neural network. Performing reservoir simulation with neural network.
It contains stateoftheart techniques to be applied in reservoir geophysics, well logging, reservoir geology, and reservoir engineering. Pdf artificial neural networks for predicting petroleum quality. The training of the neural network is done using a supervised learning approach with the back propagation algorithm. The primary purpose of this type of software is, through simulation, to gain a better understanding of the behavior and the properties of neural networks. The simulation model, shown in figure 1, is a portion of a very large geological model developed by castro 3.
Prediction of reservoir properties by monte carlo simulation. Stochastic reservoir simulation using neural networks. In addition to neuronal and synaptic state, snns incorporate the concept of time into their operating model. I cant find good explication for number of drop of transient states. We employed matlabs neural network fitting toolbox to train a proxy neural network model. The performance is analyzed using a simulation model for the.
The arti cial neural network paradigm is a major area of research within a. Terekhov neurok techsoft, llc, moscow, russia email. Reservoir computing is a framework for computation derived from recurrent neural network theory that maps input signals into higher dimensional computational spaces through the dynamics of a fixed, nonlinear system called a reservoir. In this study, a deep learning neural network was developed to estimate the petrophysical characteristics required building a full field earth model for a large reservoir. Genetic algorithms combined to ann applied to reservoir simulation and. The development of artificial neural networks began approximately 50 years ago, inspired by a desire to understand the human brain and emulate its functioning. This paper demonstrates the use of the k fold cross validation technique to obtain confidence bounds on an anns accuracy statistic from a finite sample set. Reservoir parameter estimation using a hybrid neural.
Many spiking neural network frameworks exist, each with a unique set of use cases. The reservoir is an important hydraulic engineering measure for human utilization and management of water resources. Some focus on the biologically realistic simulation of neurons, while others on highlevel spiking network functionality. Neural networkbased simulationoptimization model for. N2 the application of chemical method for hydrocarbons extraction has attracted increasing interest in the reservoir simulation community.
Modeling and simulation, computational systems biology, bioinformatics. Mar 22, 20 download interactive neural network simulator for free. Fuzzy neural network modeling of reservoir operation. The program is intended to be used in lessons of neural networks. An artificial neural network is an interconnected group of nodes, inspired by a simplification of neurons in a brain. Soft computing for reservoir characterization and modeling. A machine learningoriented spiking neural networks. Reservoir computing, recurrent neural network learning architectures, agent architectures, machine learning applications. In the algorithm, a few simulation runs of different reservoir realizations are first made using 3level fractional factorial design. In the simulation study, four different models were considered. Nonlinear survival regression using artificial neural network. Application of artificial neural networks for calibration of. The application of rom to a realistic reservoir simulation model is illustrated and the ability of the rom to provide accurate predictions for cases that di. The basic element of a backpropagation neural network is the processing node.
Also, it may consist of a single layer of neurons with each neuron feeding its output signal back to the inputs of all the other. A system and method for modeling technology to predict accurately wateroil relative permeability uses a type of artificial neural network ann known as a generalized regression neural network grnn the ann models of relative permeability are developed using experimental data from waterflood core test samples collected from carbonate reservoirs of arabian oil fields three groups of data sets. The network used in this study employs an architecture called backpropagation that is good at. The volume is the first comprehensive book in the area of intelligent reservoir characterization written by leading experts in academia and industry. Neurovis is an interactive neural network visualizer and tutorial. The deep learning textbook is a resource intended to help students and practitioners enter the field of machine learning in general and deep learning in particular. Artificial neural networks ann or connectionist systems are. Computer runs may be performed at an early stage of the reservoir study to estimate sensitivity of calculated reservoir performance to variations in the various required input data. Well tops guided prediction of reservoir properties using modular neural network concept. Datadriven reservoir modeling introduces new technology and protocols intelligent systems that teach the reader how to apply data analytics to solve realworld, reservoir engineering problems. In this chapter, the authors provide fundamental principles of higher order neural units honus and higher order neural networks honns for modeling and. Pdf artificial intelligence application in reservoir characterization.
1597 1058 1224 496 745 1098 1245 679 1593 669 1237 483 1331 1265 376 1281 132 1511 228 1501 1410 1376 1429 508 338 531 1405 478 1362