WO2004063983A2 - Methode pour modeliser des caracteristiques hydrodynamiques d'ecoulements polyphasiques par reseaux de neurones - Google Patents
Methode pour modeliser des caracteristiques hydrodynamiques d'ecoulements polyphasiques par reseaux de neurones Download PDFInfo
- Publication number
- WO2004063983A2 WO2004063983A2 PCT/FR2003/003583 FR0303583W WO2004063983A2 WO 2004063983 A2 WO2004063983 A2 WO 2004063983A2 FR 0303583 W FR0303583 W FR 0303583W WO 2004063983 A2 WO2004063983 A2 WO 2004063983A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- flow
- regimes
- different
- neural networks
- probabilities
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 230000001537 neural effect Effects 0.000 title claims abstract description 15
- 239000012530 fluid Substances 0.000 claims abstract description 19
- 230000001052 transient effect Effects 0.000 claims abstract description 6
- 238000013528 artificial neural network Methods 0.000 claims description 34
- 238000011156 evaluation Methods 0.000 claims description 9
- 230000006399 behavior Effects 0.000 claims description 7
- 229930195733 hydrocarbon Natural products 0.000 abstract description 3
- 150000002430 hydrocarbons Chemical class 0.000 abstract description 3
- 239000004215 Carbon black (E152) Substances 0.000 abstract 1
- 230000006870 function Effects 0.000 description 13
- 210000002569 neuron Anatomy 0.000 description 9
- 239000012071 phase Substances 0.000 description 9
- 238000004088 simulation Methods 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 4
- 239000007789 gas Substances 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 239000007792 gaseous phase Substances 0.000 description 2
- 239000007788 liquid Substances 0.000 description 2
- 239000007791 liquid phase Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000009491 slugging Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Definitions
- the present invention relates to a method for modeling in real time, by neural networks, hydrodynamic characteristics of multiphase flows in transient phase in conduits.
- the method finds applications in particular for the modeling of the flows of hydrocarbons in oil pipes.
- the transportation of hydrocarbons from production sites to processing units is an important link in the oil chain. It is a delicate link because of the complexity of the interactions between the phases constituting the effluents transported.
- the primary objective of operators is to achieve optimum productivity under the best possible safety conditions. They must therefore manage speed and temperature as well as possible, to avoid unnecessary pressure losses, unwanted deposits and irregularities in flow.
- the generally used method consists in best modeling the transport of complex multiphase flows so as to provide at all times a picture of the flows in the different parts of the production chain, taking into account the precise constitution of the effluent, the flows and pressures and flow regimes.
- the complexity of the simulation tools is like that of the modeled phenomena. Accuracy and performance can only be obtained after a relatively long modeling time, which is hardly compatible with real-time management.
- Another approach allowing, alone or in parallel with the above modeling methods, to manage parameters of a circulation of fluids in real time uses neural networks.
- Neural networks as we recall, define a data processing mode simulating the functioning of biological neuron systems.
- an element performs a relatively simple calculation such as a weighted sum of the signals present at its inputs applied to a non-linear function, which determines the state of its output.
- a large number of such elements are interconnected in series and in parallel.
- a suitable choice of weighting factors allows the network to perform complex functions.
- So-called backpropagation networks for example use multiple layers of elements defined above. The adaptation of such a network to a specific task is done by "training" the network on a certain number of examples and by adjusting the weighting factors for each element to the appropriate values.
- a neural network proceeds according to a non-linear regression method, which is all the more efficient compared to conventional methods.
- MLP Multi Layer Perceptron
- Kohonen networks well known to specialists.
- we perform a modeling of the flow regime by forming a non-linear neural network with an input layer with as many inputs as structure parameters and physical quantities, an output layer with as many outputs as quantities necessary for the estimation of the flow regime and at least one intermediate layer, constituting a learning base with predefined tables connecting different values obtained for the output data to the corresponding values of the input data, and determining by iterations the weighting factors of the activation function making it possible to correctly relate the values in the tables of the input and output data.
- neuron output data is analyzed so as to sort, from the values of the neural network output data, the only relevant data to be taken into account in the iterative determination of the weighting coefficients of the activation function. .
- the module must be integrated into a general hydrodynamic and thermodynamic model of simulation of multiphase flows
- flows are considered globally, without making a distinction according to the different possible flow regimes of fluids in the pipe: stratified flows, dispersed flows, intermittent flows, whose behaviors are different. This can generate modeling errors which are sometimes too large with regard to the quality of estimation required for production monitoring. In addition, they do not take into account the existence of simple models (for example analytical models) translating in mathematical form the characteristics of one (or more) flow regime (s).
- the object of the method according to the invention is to construct a model for simulating in real time the hydrodynamic behavior of a flow of multiphase fluids in a transient phase in a pipe, taking account of fixed operating conditions relating to a certain number of parameters. defined structural relating to driving, and a set of defined physical quantities, with ranges of variation fixed for said parameters and said physical quantities.
- Specialized neural networks are used, each dedicated to a defined task, the outputs of which are combined with weightings under the control of an evaluation neural network to produce results necessary for the estimation of hydrodynamic behavior, these networks receiving on their inputs of the parameters of structure and physical quantities, the neural networks being formed iteratively to adjust to the values of basic learning.
- the method essentially includes the following steps:
- a modeling tool is used to form specific learning bases corresponding respectively to different flow regimes of fluids in the pipe, each of them grouping together parameter values specifically characterizing a flow regime and a base d specific learning gathering probability values from the different regimes, by analyzing data from the learning bases;
- the neural evaluation network is formed to make it capable of evaluating at all times the probabilities that the flow in the pipe corresponds respectively to the different flow regimes by reference to the probability values obtained for each of the flow regimes;
- the method comprises the construction of at least three neural networks dedicated respectively to the stratified flow regime, to the dispersed flow regime and to the intermittent flow regime, the probabilities are evaluated that the flow of fluids in the pipe corresponds respectively to the three flow regimes and the results are linearly combined with the outputs of the three dedicated neural networks by weighting them by said probabilities.
- the method retains the capacity of the above-mentioned methods to perform the simulation of flows in real time, and the results obtained take advantage of the regularity of the estimation function obtained.
- the input data come from:
- Each model produces, for example at the output, the hydrodynamic behavior of the effluents, and, in particular, the flow regime.
- D evaluates and delivers on two main outputs, hydrodynamic data in the pipe part for which it is desired to determine the flow regime, the difference dN of speed between gas and liquid for example, or the fraction ⁇ ( ⁇ e [ ⁇ ; l ]) flow of the regime treated by him.
- Other quantities qualifying the flow regime can be calculated from these two outputs.
- the outputs provided by the experts are essentially the speed differences between the phases, under the assumption of a certain flow regime (for example, the Laminated expert delivers the estimate of the speed difference between the phases in the 'hypothesis of a stratified flow).
- the outputs provided by the probability network is the probability of belonging to each of the flow regimes processed by expert networks, knowing the inputs.
- the different neural networks or experts dedicated to the different flow regimes are preferably Multi Layer Perceptron (MLP) networks.
- MLP Multi Layer Perceptron
- the number of hidden layers and the number of neurons that compose them are determined from the learning and validation results of the networks.
- the network is fully connected.
- the non-linearity of this network is obtained either by a sigmoid activation function governing the behavior of neurons in the hidden layer, or the identity function or the softmax functions for the output layer.
- Neural networks have an input layer, one or two hidden layers, and an output layer.
- the activation functions of the various neurons are either the sigmoid function (for the hidden layers), the identity function or the softmax function (for the output layers).
- the weights of each of the networks or experts are determined at the end of a learning phase; during this phase, we feed them with a set of data constituting their learning base, and we optimize the configuration and the weights of the network by minimizing the errors observed for all the samples in the database, between the data output from the network calculation and the data expected at the output, given by the base.
- the errors can be the absolute errors between the input and output quantities or the relative errors, depending on the desired performance for the network.
- the generalization capabilities of the network are then tested on its ability to properly calculate the two outputs for inputs that are unknown to it.
- the databases used are of different types:
- each base contains pairs of input / output values, each output value being the desired value of the quantity estimated in the case of the flow regime processed by the dedicated network;
- the desired output is a vector of magnitude equal to the number N ⁇ om of flow regimes considered (in the example of FIG. 1, the vector is of dimension 3); this vector contains (N flmvs -1) zero values, and a value equal to 1, which corresponds to the probability that the flow regime of fluids in the pipe corresponds to that of which the dedicated neural network deals.
- the initial learning base includes the data defined above: geometric data, data describing the fluid, the mixture, etc. From this basis, by forcing the simulation tool to interpret the results it gives respectively in terms of specific flow, for example in terms of stratified flow, then dispersed flow then intermittent flow, we constitute as many dedicated bases as there are flow regimes in the pipe, each of them grouping together parameter values specifically characterizing a flow regime.
- the simulation tool is also used to form a specific learning base gathering probability values from the different regimes, by analyzing data from the initial learning base.
- the different dedicated neural networks E Stm , E Dlip , E lni for example
- the neural evaluation network R ProJfl
- the neural evaluation network is also formed to make it capable of evaluating at all times the probabilities that the flow in the pipe corresponds respectively to the different regimes d 'flow by reference to the probability values obtained for each of the flow regimes.
- the results provided by the different neural networks weighted by said probabilities are combined, in accordance with equation (1).
- the probability estimation function makes it possible to create a global hydrodynamic law from the different flow laws modeled by the different dedicated neural models.
- the transition between two flow laws is more or less steep (more or less strong derivative) according to the precision given to the estimation of the probabilities, but it is continuous, which eliminates the possible uncertainties in the results of the model related to l existence of discontinuities.
- the global model is suitable either for use independent of any other module, or for integration into a complete model.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Automation & Control Theory (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Measuring Volume Flow (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Feedback Control In General (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
GB0513744A GB2412466B (en) | 2002-12-10 | 2003-12-03 | Method of modelling the hydrodynamic characteristics of multiphase flows using neuronal networks |
US10/538,089 US7177787B2 (en) | 2002-12-10 | 2003-12-03 | Method for modelling hydrodynamic characteristics of multiphase flows using neuronal networks |
BR0317152-3A BR0317152A (pt) | 2002-12-10 | 2003-12-03 | Método para modelar as caracterìsticas hidrodinâmicas de fluxos multifásicos utilizando redes neurais |
NO20052731A NO20052731L (no) | 2002-12-10 | 2005-06-07 | Fremgangsmate for modellering av hydrodynamiske karakteristikker til flerfasestrommer ved bruk av neurale nettverk |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR0215570A FR2848320B1 (fr) | 2002-12-10 | 2002-12-10 | Methode pour modeliser des caracteristiques hydrodynamiques d'ecoulements polyphasiques par reseaux de neurones |
FR02/15570 | 2002-12-10 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2004063983A2 true WO2004063983A2 (fr) | 2004-07-29 |
WO2004063983A3 WO2004063983A3 (fr) | 2005-05-12 |
Family
ID=32320135
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/FR2003/003583 WO2004063983A2 (fr) | 2002-12-10 | 2003-12-03 | Methode pour modeliser des caracteristiques hydrodynamiques d'ecoulements polyphasiques par reseaux de neurones |
Country Status (6)
Country | Link |
---|---|
US (1) | US7177787B2 (fr) |
BR (1) | BR0317152A (fr) |
FR (1) | FR2848320B1 (fr) |
GB (1) | GB2412466B (fr) |
NO (1) | NO20052731L (fr) |
WO (1) | WO2004063983A2 (fr) |
Families Citing this family (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8197700B2 (en) * | 2005-12-30 | 2012-06-12 | Saudi Arabian Oil Company | Computational method for sizing three-phase separators |
US20080270328A1 (en) * | 2006-10-18 | 2008-10-30 | Chad Lafferty | Building and Using Intelligent Software Agents For Optimizing Oil And Gas Wells |
US20080202763A1 (en) * | 2007-02-23 | 2008-08-28 | Intelligent Agent Corporation | Method to Optimize Production from a Gas-lifted Oil Well |
US8386221B2 (en) * | 2009-12-07 | 2013-02-26 | Nuovo Pignone S.P.A. | Method for subsea equipment subject to hydrogen induced stress cracking |
US9134454B2 (en) | 2010-04-30 | 2015-09-15 | Exxonmobil Upstream Research Company | Method and system for finite volume simulation of flow |
CA2803066A1 (fr) | 2010-07-29 | 2012-02-02 | Exxonmobil Upstream Research Company | Procedes et systemes pour une simulation de flux par apprentissage automatique |
EP2599032A4 (fr) | 2010-07-29 | 2018-01-17 | Exxonmobil Upstream Research Company | Procédé et système de modélisation d'un réservoir |
WO2012015518A2 (fr) | 2010-07-29 | 2012-02-02 | Exxonmobil Upstream Research Company | Procédés et systèmes de simulation d'écoulement basée sur un apprentissage machine |
CA2807300C (fr) | 2010-09-20 | 2017-01-03 | Exxonmobil Upstream Research Company | Formulations souples et adaptatives pour des simulations de gisements complexes |
US9015093B1 (en) | 2010-10-26 | 2015-04-21 | Michael Lamport Commons | Intelligent control with hierarchical stacked neural networks |
US8775341B1 (en) | 2010-10-26 | 2014-07-08 | Michael Lamport Commons | Intelligent control with hierarchical stacked neural networks |
WO2013039606A1 (fr) | 2011-09-15 | 2013-03-21 | Exxonmobil Upstream Research Company | Opérations matricielles et vectorielles optimisées dans des algorithmes à instructions limitées qui effectuent des calculs eos |
EP2901363A4 (fr) | 2012-09-28 | 2016-06-01 | Exxonmobil Upstream Res Co | Suppression des failles dans des modèles géologiques |
WO2014200669A2 (fr) | 2013-06-10 | 2014-12-18 | Exxonmobil Upstream Research Company | Détermination de paramètres de puits pour une optimisation de rendement de puits |
EP3175265A1 (fr) | 2014-07-30 | 2017-06-07 | ExxonMobil Upstream Research Company | Procédé de génération de maillage volumétrique dans un domaine ayant des propriétés de matériau hétérogènes |
AU2015339883B2 (en) | 2014-10-31 | 2018-03-29 | Exxonmobil Upstream Research Company | Methods to handle discontinuity in constructing design space for faulted subsurface model using moving least squares |
EP3213126A1 (fr) | 2014-10-31 | 2017-09-06 | Exxonmobil Upstream Research Company | Gestion de discontinuité de domaine dans un modèle de grille de sous-surface à l'aide de techniques d'optimisation de grille |
CA3043231C (fr) | 2016-12-23 | 2022-06-14 | Exxonmobil Upstream Research Company | Procede et systeme de simulation de reservoir stable et efficace a l'aide d'indicateurs de stabilite |
CN111344710A (zh) | 2017-09-26 | 2020-06-26 | 沙特阿拉伯石油公司 | 使用基于机器学习的模型进行成本有效的热力学流体特性预测的方法 |
WO2019132864A1 (fr) * | 2017-12-26 | 2019-07-04 | Landmark Graphics Corporation | Représentation efficace de résultats de simulation tridimensionnelle complexe pour opérations en temps réel |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6208983B1 (en) * | 1998-01-30 | 2001-03-27 | Sarnoff Corporation | Method and apparatus for training and operating a neural network for detecting breast cancer |
EP1217474A1 (fr) * | 2000-12-22 | 2002-06-26 | Institut Francais Du Petrole | Méthode pour former un module à réseaux neuronaux optimisé, destiné à simuler le mode d'écoulement d'une veine de fluides polyphasiques |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5550761A (en) * | 1994-02-08 | 1996-08-27 | Institut Francais Du Petrole | Method for modelling multiphase flows in pipelines |
FR2756044B1 (fr) * | 1996-11-18 | 1998-12-24 | Inst Francais Du Petrole | Methode pour constituer un modele representatif d'ecoulements polyphasiques dans des conduites de production petroliere |
FR2812389B1 (fr) * | 2000-07-27 | 2002-09-13 | Inst Francais Du Petrole | Methode et systeme pour estimer en temps reel le mode d'ecoulement d'une veine fluide polyphasique, en tous points d'une conduite |
-
2002
- 2002-12-10 FR FR0215570A patent/FR2848320B1/fr not_active Expired - Fee Related
-
2003
- 2003-12-03 US US10/538,089 patent/US7177787B2/en not_active Expired - Fee Related
- 2003-12-03 WO PCT/FR2003/003583 patent/WO2004063983A2/fr active Application Filing
- 2003-12-03 GB GB0513744A patent/GB2412466B/en not_active Expired - Fee Related
- 2003-12-03 BR BR0317152-3A patent/BR0317152A/pt not_active IP Right Cessation
-
2005
- 2005-06-07 NO NO20052731A patent/NO20052731L/no not_active Application Discontinuation
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6208983B1 (en) * | 1998-01-30 | 2001-03-27 | Sarnoff Corporation | Method and apparatus for training and operating a neural network for detecting breast cancer |
EP1217474A1 (fr) * | 2000-12-22 | 2002-06-26 | Institut Francais Du Petrole | Méthode pour former un module à réseaux neuronaux optimisé, destiné à simuler le mode d'écoulement d'une veine de fluides polyphasiques |
Non-Patent Citations (3)
Title |
---|
LIN-CHENG WANG ET AL: "A modular neural network vector predictor for predictive VQ" PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP) LAUSANNE, SEPT. 16 - 19, 1996, NEW YORK, IEEE, US, vol. VOL. 1, 16 septembre 1996 (1996-09-16), pages 431-434, XP010202423 ISBN: 0-7803-3259-8 * |
TABUSE M ET AL: "Recurrent neural network using mixture of experts for time series processing" SYSTEMS, MAN, AND CYBERNETICS, 1997. COMPUTATIONAL CYBERNETICS AND SIMULATION., 1997 IEEE INTERNATIONAL CONFERENCE ON ORLANDO, FL, USA 12-15 OCT. 1997, NEW YORK, NY, USA,IEEE, US, 12 octobre 1997 (1997-10-12), pages 536-541, XP010248978 ISBN: 0-7803-4053-1 * |
WALTER P ET AL: "3D object recognition with a specialized mixtures of experts architecture" NEURAL NETWORKS, 1999. IJCNN '99. INTERNATIONAL JOINT CONFERENCE ON WASHINGTON, DC, USA 10-16 JULY 1999, PISCATAWAY, NJ, USA,IEEE, US, 10 juillet 1999 (1999-07-10), pages 3563-3568, XP010373074 ISBN: 0-7803-5529-6 * |
Also Published As
Publication number | Publication date |
---|---|
US7177787B2 (en) | 2007-02-13 |
FR2848320B1 (fr) | 2005-01-28 |
BR0317152A (pt) | 2005-11-01 |
FR2848320A1 (fr) | 2004-06-11 |
GB0513744D0 (en) | 2005-08-10 |
GB2412466B (en) | 2006-12-20 |
GB2412466A (en) | 2005-09-28 |
NO20052731L (no) | 2005-07-05 |
US20060025975A1 (en) | 2006-02-02 |
WO2004063983A3 (fr) | 2005-05-12 |
NO20052731D0 (no) | 2005-06-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP1176481B1 (fr) | Methode et système pour estimer en temps réel le mode d'écoulement d'une veine fluide polyphasique, en tous points d'une conduite | |
EP1217474B1 (fr) | Méthode pour former un module à réseaux neuronaux optimisé, destiné à simuler le mode d'écoulement d'une veine de fluides polyphasiques | |
WO2004063983A2 (fr) | Methode pour modeliser des caracteristiques hydrodynamiques d'ecoulements polyphasiques par reseaux de neurones | |
Ahmadi et al. | Evolving predictive model to determine condensate-to-gas ratio in retrograded condensate gas reservoirs | |
CN107087161B (zh) | 视频业务中基于多层神经网络的用户体验质量的预测方法 | |
EP3877912A1 (fr) | Procédé de construction de réseau de neurones pour la simulation de systèmes physiques | |
Fruhwirth et al. | Hybrid simulation using neural networks to predict drilling hydraulics in real time | |
EP3953662B1 (fr) | Procede de definition d'un chemin | |
CN110796306A (zh) | 一种外汇时间序列预测的构建方法 | |
CN109001211A (zh) | 基于卷积神经网络的长输管道焊缝检测系统及方法 | |
CA2519184C (fr) | Methode pour former rapidement un modele stochastique representatif de la distribution d`une grandeur physique dans un milieu heterogene par une selection appropriee de realisations geostatistiques | |
EP2650471A1 (fr) | Procédé de sélection des positions de puits a forer pour l'exploitation d'un gisement pétrolier | |
US20050092161A1 (en) | Song search system and song search method | |
WO2020094995A1 (fr) | Procédé de construction de réseau de neurones pour la simulation de systèmes physiques | |
Hegeman et al. | Application of artificial neural networks to downhole fluid analysis | |
Yang et al. | Machine Learning Based Predictive Models for CO2 Corrosion in Pipelines With Various Bending Angles | |
Akbari et al. | Dewpoint pressure estimation of gas condensate reservoirs, using artificial neural network (ANN) | |
FR2881857A1 (fr) | Outil informatique de prevision | |
WO2019211367A1 (fr) | Procede de generation automatique de reseaux de neurones artificiels et procede d'evaluation d'un risque associe | |
EP1473627A2 (fr) | Procédé pour la modélisation de données référentielles et son utilisation pour la localisation de données référentielles dans un système d'informations | |
EP0401926B1 (fr) | Procédé de traitement, stucture de réseau de neurones mettant en oeuvre le procédé et ordinateur pour simuler ladite structure de réseau de neurones | |
Rey-Fabret et al. | Neural networks tools for improving tacite hydrodynamic simulation of multiphase flow behavior in pipelines | |
Mukhanbet et al. | Parallel Implementation of Neural Networks for Solving the Problem of Oil Production. | |
FR2735568A1 (fr) | Procede de reconstruction d'un champ dense de caracteristiques associees a un phenomene physique | |
WO1999001825A1 (fr) | Procede de construction d'un reseau de neurones pour la modelisation d'un phenomene |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A2 Designated state(s): BR GB NO US |
|
ENP | Entry into the national phase |
Ref document number: 2006025975 Country of ref document: US Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10538089 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 0513744 Country of ref document: GB Kind code of ref document: A Free format text: PCT FILING DATE = 20031203 |
|
ENP | Entry into the national phase |
Ref document number: PI0317152 Country of ref document: BR |
|
WWP | Wipo information: published in national office |
Ref document number: 10538089 Country of ref document: US |