Data processing: structural design – modeling – simulation – and em – Modeling by mathematical expression
Reexamination Certificate
2001-12-21
2004-11-23
Phan, Thai (Department: 2128)
Data processing: structural design, modeling, simulation, and em
Modeling by mathematical expression
C703S005000
Reexamination Certificate
active
06823296
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method for forming a neural network module for real-time simulation of the flow mode, at any point of a pipe, of a multiphase fluid stream comprising at least a liquid phase and at least a gas phase, that is best suited to operating conditions and to a set of fixed physical quantities.
2. Description of the Prior Art
Transporting hydrocarbons from production sites to treating plants constitutes an important link in the petroleum chain. It is a delicate link because of the complex interactions between the phases forming the transported effluents. The basic objective for operators is to reach an optimum productivity under the best safety conditions. They therefore have to control as best they can the velocity and the temperature of the transported hydrocarbons so as to avoid unnecessary pressure drops, unwanted deposits and unsteady-state flows. The method that is generally used models in the best possible way the transportation of complex multiphase streams so as to provide at all times an image of the flows in the various parts of the production chain, taking into account the precise constitution of the effluent, the flow rates, the pressures and the flow modes.
There are currently various software modules for simulating the transport of complex multiphase streams, allowing to design suitable production equipments at an early stage.
U.S. Pat. Nos. 5,550,761, 6,028,992 and 5,960,187 filed by the applicant notably describe modelling modules forming the TACITE model known in the art, allowing to simulate the transport of complex multiphase streams as a steady or transient flow and accounting for instability phenomena that occur because of the irregular geometry of the formation crossed by the pipe or of the topography thereof, referred to by specialists as “terrain slugging” or “severe slugging”.
The simulation modules are as complex as the modelled phenomena. Precision and performance can only be obtained after a relatively long modelling time, which is not really compatible with real-time management. That is the reason why the modelling modules cannot be used as they are for real-time management of the production. It therefore appears necessary to use modelling methods offering a good compromise between calculating speed and accuracy of results.
French Patent application 00/09,889 filed by the applicant describes a method of real-time estimation of the flow mode, at any point of a pipe having a structure that can be defined by a certain number of structure parameters, of a multiphase fluid stream defined by several physical quantities and comprising liquid and gas phases. According to this method, the flow mode is modelled:
by forming a non-linear neural network with an input layer having as many inputs as there are structure parameters and physical quantities necessary for good estimation of the output, an output layer with as many outputs as there are quantities necessary for estimation of the flow mode, and at least one intermediate layer,
by forming a learning base with predetermined tables connecting various values obtained for the output data to the corresponding values of the input data, and
by determining, by iterations, weighting factors of the activation function allowing to properly connect the values in the input and output data tables.
Output data of the neural network is preferably analysed so as to sort, among the values of the output data of the neural network, only the pertinent data to be taken into account for iterative determination of the weighting coefficients of the activation function.
SUMMARY OF THE INVENTION
The method according to the invention forms a module (hydrodynamic or thermodynamic for example) intended for real-time simulation of the flow mode, at any point of a pipe, of a multiphase fluid stream comprising at least a liquid phase and at least a gas phase, that is best suited to fixed operating conditions concerning a certain number of determined structure and physical parameters relative to the pipe, and a set of determined physical quantities (hydrodynamic or thermodynamic quantities for example), with fixed variation ranges for the parameters and the physical quantities.
The method of the invention comprises using a modelling system based on non-linear neural networks each having inputs for structure parameters and physical quantities, outputs where quantities necessary for estimation of the flow mode are available, and at least one intermediate layer. The neural networks are determined iteratively so as to adjust to the values of a learning base with predetermined tables connecting various values obtained for the output data to the corresponding values of the input data.
The method is forms a learning base suited to the imposed operating conditions and optimized neural networks best adjusted to the imposed operating conditions are generated.
In the case, for example, where the module is to be integrated in a general multiphase flow simulation model, both hydrodynamic and thermodynamic, the model is used to form the learning base so as to select the set of physical quantities that is best suited to the model, as well as the variation ranges fixed for the parameters and the physical quantities, and the optimized neural networks that best adjust to the learning base formed are generated.
REFERENCES:
patent: 4874963 (1989-10-01), Alspector
patent: 5619433 (1997-04-01), Wang et al.
patent: 5649064 (1997-07-01), Jorgensen et al.
patent: 5924086 (1999-07-01), Mathur et al.
patent: 6047221 (2000-04-01), Piche et al.
patent: 6243696 (2001-06-01), Keeler et al.
patent: 6411944 (2002-06-01), Ulyanov
Bishop et al., “Analysis of multiphase flows using dual-energy gamma densitometry and neural networks”,Nuclear Instruments and Methods in Physics ResearchA327 (1993) 580-593.
Duret Emmanuel
Heintze Eric
Henriot Veronique
Rey-Fabret Isabelle
Institut Francais du Pe'trole
Phan Thai
LandOfFree
Method for forming an optimized neural network module... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for forming an optimized neural network module..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for forming an optimized neural network module... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3307388