CN114925606A - Fracture parameter optimization method based on Gaussian process regression - Google Patents
Fracture parameter optimization method based on Gaussian process regression Download PDFInfo
- Publication number
- CN114925606A CN114925606A CN202210554894.3A CN202210554894A CN114925606A CN 114925606 A CN114925606 A CN 114925606A CN 202210554894 A CN202210554894 A CN 202210554894A CN 114925606 A CN114925606 A CN 114925606A
- Authority
- CN
- China
- Prior art keywords
- fracture
- gaussian process
- process regression
- daily
- parameters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 85
- 230000008569 process Effects 0.000 title claims abstract description 49
- 238000005457 optimization Methods 0.000 title claims abstract description 30
- 230000006870 function Effects 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 12
- 230000002068 genetic effect Effects 0.000 claims abstract description 12
- 238000004088 simulation Methods 0.000 claims abstract description 12
- 230000035699 permeability Effects 0.000 claims abstract description 8
- 238000012360 testing method Methods 0.000 claims abstract description 8
- 239000011159 matrix material Substances 0.000 claims description 26
- 230000001186 cumulative effect Effects 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 9
- 238000004519 manufacturing process Methods 0.000 claims description 9
- 230000005465 channeling Effects 0.000 claims description 6
- 238000007476 Maximum Likelihood Methods 0.000 claims description 5
- 230000035772 mutation Effects 0.000 claims description 4
- 238000012952 Resampling Methods 0.000 claims description 3
- 230000000295 complement effect Effects 0.000 claims 1
- 238000004364 calculation method Methods 0.000 abstract description 7
- 239000007789 gas Substances 0.000 description 21
- 238000011161 development Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 4
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 4
- 230000001133 acceleration Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 239000003345 natural gas Substances 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012938 design process Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000003129 oil well Substances 0.000 description 1
- 239000003209 petroleum derivative Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Data Mining & Analysis (AREA)
- Computer Hardware Design (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a fracture parameter optimization method based on Gaussian process regression, which comprises the following steps of: obtaining a group of daily output data and accumulated output data under different fracture parameters through simulation, using the daily output data and the accumulated output data as a basic data set, and establishing implicit functions of the fracture parameters, the daily output and the accumulated output in the basic data set; dividing the basic data set into a training set and a testing set, and respectively obtaining a Gaussian process regression daily yield model and a Gaussian process regression accumulated yield model; establishing an objective function according to two Gaussian process regression models; setting an initial population, preferably selecting better individuals in the initial population, and performing iterative computation on the better individuals by using a genetic algorithm to preferably select optimal crack parameters; the method can effectively optimize fracture parameters of the compact gas fracturing well, including optimal fracture half-seam length, optimal fracture seam spacing and optimal fracture permeability, and meanwhile can greatly improve calculation efficiency and achieve real-time optimization.
Description
Technical Field
The invention relates to the technical field of gas reservoir development, in particular to a fracture parameter optimization method based on Gaussian process regression.
Background
The dense gas is a very important unconventional oil gas resource, 80% of natural gas reservoirs in China are located in the dense gas reservoirs, and the enhancement of the development efficiency of the dense gas reservoirs is an important guarantee for ensuring the energy safety in China. For low-permeability and ultra-low-permeability compact gas reservoirs, fracturing production is often required to be put into operation or fracturing production increase is carried out after the production is put into operation, fracturing parameter optimization is a key link of fracturing design, and in order to maximize the production capacity of a fracturing modified well, fracture parameters such as fracture interval (number of fractures), fracture half-fracture length and fracture permeability need to be efficiently optimized.
The core of applying machine learning to the technical field of air reservoir development is two aspects, namely big data analysis and expensive optimization acceleration. The big data analysis mainly comprises two contents, namely, forecasting development benefits and monitoring reservoir dynamics, and the acceleration of expensive optimization mainly refers to physical model optimization. By "proxy model optimization" is meant building a proxy model (e.g., an approximation learning model such as a machine learning model) for an expensive (e.g., "expensive" in this context means computationally expensive, such as commercial software or numerical simulation models) objective function (e.g., those complex and time consuming numerical simulation models in the analysis and optimization design process), and using the proxy model to replace the original true function in the optimization process. At first, the prototype of the proxy model is a polynomial response surface model, and various methods such as polynomial response surface, kriging interpolation, radial basis function interpolation and the like have been developed at present, and are commonly used for some unknown problems of mathematical models. If the exact mathematical model between the design parameters (vehicle speed, frame structure, braking, etc.) and the degree of collision is not known in the vehicle design, then a surrogate model can be used to optimize the design parameters. At present, the application of a proxy model is very lacking in the technical field of gas reservoir development.
When the traditional fracture parameter optimization is carried out, a large amount of numerical simulation is used, the parameter optimization calculation efficiency based on the numerical simulation is low, the optimization of unconventional reservoir hydraulic fracture parameters is usually expensive in calculation cost, and hours or even days are often spent. And by constructing a machine learning model to substitute for the traditional numerical simulation model to assist in optimizing the crack parameters, the calculation efficiency can be greatly improved, and real-time optimization is realized. However, according to the traditional method, the optimization of the fracturing parameters based on commercial software or numerical simulation takes tens of hours or even days, the efficiency is low, and the rapid and real-time optimization of the parameters is difficult to achieve on site.
Disclosure of Invention
Aiming at the problems, the invention provides a fracture parameter optimization method based on Gauss process regression, which can effectively optimize fracture parameters of a compact gas fracturing well, including optimal fracture half-seam length, optimal fracture seam spacing and optimal fracture permeability, and assist in optimizing the fracture parameters by constructing a machine learning model and acting on a traditional numerical simulation model, can greatly improve the calculation efficiency, realize real-time optimization, and can make up for the defects of the fracture parameter optimization method of the compact gas fracturing well in the prior art.
The invention adopts the following technical scheme that,
a fracture parameter optimization method based on Gaussian process regression is characterized by comprising the following steps,
step 2, dividing a basic data set into a training set and a testing set, obtaining a covariance matrix according to the training set, and carrying out Gaussian process regression training by using training set data based on a maximum likelihood estimation method to respectively obtain a Gaussian process regression daily yield model and a Gaussian process regression accumulated yield model;
step 3, establishing a target function based on maximum yield according to two Gaussian process regression models;
and 4, setting an initial population, preferably selecting a better individual in the initial population by using a roulette method, coding the better individual, and performing iterative computation on the better individual by using a genetic algorithm to preferably select an optimal crack parameter.
The invention has the beneficial effects that:
on the basis of ensuring convenient field implementation, the method can effectively optimize the fracture parameters of the compact gas fracturing well, including optimal fracture half-seam length, optimal fracture seam spacing and optimal fracture permeability, and the fracture parameters are assisted and optimized by constructing a machine learning model and acting on a traditional numerical simulation model, so that the calculation efficiency can be greatly improved, real-time optimization is realized, the defects of the fracture parameter optimization method of the compact gas fracturing well in the prior art can be overcome, and the method is favorable for wide application;
meanwhile, the calculation process of the new method provided by the patent only needs several seconds, the optimization speed is improved by 4-5 orders of magnitude, and the new method has considerable potential economic value in the aspect of efficiency improvement and speed increase. In addition, machine learning and artificial intelligence are introduced into petroleum and natural gas engineering, the intelligent oil well fracturing system is in accordance with the development direction of the intelligent industry, and the method has strong prospect in the aspect of fracturing well fracture parameter optimization.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments will be briefly described below, and it is apparent that the drawings in the following description only relate to some embodiments of the present invention and are not limiting on the present invention.
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a block diagram of the Gaussian process regression and genetic algorithm fusion (GPR-GA) operation of the present invention;
FIG. 3 is a schematic diagram showing comparison between a fitting value of the daily output of dense gas and a true value of the daily output of dense gas of the model of the Gaussian process regression daily output of the present invention;
FIG. 4 is a schematic diagram showing comparison between a fitting value of cumulative yield of dense gas and a true value of cumulative yield of dense gas in a Gaussian process regression cumulative yield model according to the present invention;
FIG. 5 is a schematic diagram of the genetic algorithm roulette method of the present invention for selecting superior individuals.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1 to 5, the present invention provides a technical solution, a fracture parameter optimization method based on gaussian process regression, comprising the following steps,
(1) dividing a matrix grid, and dispersing the cracks into a plurality of crack segments through the matrix grid;
(2) calculating the channeling coefficient between the matrix and the crack (the connection between the crack and the matrix is called type I NNC), wherein the channeling coefficient between the matrix mesh and the crack segments is,
(3) calculating the channeling coefficient between the cracks (the connection between the cracks is called II-type NNC), wherein the channeling coefficient between the matrix grids and the crack segments is,
wherein,
(4) constructing a seepage equation, wherein the seepage equation of the gas in the matrix is,
the seepage equation of the gas in the fracture is,
daily production data and accumulated production data under different fracture parameters are calculated to form a basic data set, and the results are shown in table 1,
TABLE 1 basic data set
(5) And (3) establishing implicit functions of the crack parameters, the daily yield and the accumulated yield according to the basic data set:
y 1 ＝f 1 (x)
y 2 ＝f 2 (x)
in the formula, y 1 For daily output, y 2 For cumulative production, x is the fracture parameters including fracture permeability, fracture length, fracture spacing.
Step 2, dividing the basic data set into a training set and a testing set, obtaining a covariance matrix according to the training set, carrying out Gaussian process regression training by using the training set data based on a maximum likelihood estimation method, respectively obtaining a Gaussian process regression daily yield model and a Gaussian process regression accumulated yield model,
(1) in this step, a Bootstrap resampling method is used to select a new set from the basic data set, and the basis is assumedThe base data set S contains n different samples
Each time a sample is taken back from the basic data set S, a total of n times, forming a new set S 1 Then set S 1 Does not contain a certain sample
The probability of (a) being,
when n → ∞ is present
Remove S 1 Middle repeated samples, then new set S 1 Samples containing about 1-36.8% ═ 63.2% of the base data set S were analyzed for S 1 As training set, let new set S 2 ＝S-S 1 As a test set.
The results of the training set and test set partitioning are shown in table 2,
TABLE 2 results of training set and test set partitioning
(2) Calculating an initial covariance matrix according to a covariance matrix formula and a training set selected by Bootstrap resampling, wherein the covariance matrix formula is,
(3) the basic data set comprises fracture parameter data x and daily output data y 1 Cumulative production data y 2 The relation of an implicit function is satisfied,
y 1 ＝f 1 (x)
y 2 ＝f 2 (x)
then when the mean of the distribution is zero, a prior gaussian distribution of the implicit function is constructed from the covariance matrix K,
f(x)＝GP(0,K)
(4) after the Gaussian distribution relation is obtained, the fitting value of the daily yield of the dense gas and the accumulated yield of the dense gas can be calculated. Optimizing the regression parameters of the Gaussian process by using a maximum likelihood estimation method and calculating an optimized covariance matrix, wherein the optimal regression parameters of the Gaussian process are shown in Table 3,
TABLE 3 iterative final values of regression parameters for Gaussian process
| Signal standard deviation (sigma) f ) | Length Scale (σ) l ) |
| e 3.22 | e 0.41 |
(6) Obtaining an optimized Gaussian distribution relation according to the optimized covariance matrix, training a Gaussian process regression model to respectively obtain a Gaussian process regression daily yield model and a Gaussian process regression cumulative yield model, wherein the fitting result of the yield of the 365 th day is shown in Table 4,
TABLE 4 fitting results of daily yields on day 365
The results of the fit to the cumulative yield at day 365 are shown in table 5,
TABLE 5 cumulative yield on day 365 fitted results
(6) Comparing the fitting value calculated by the regression model of the Gaussian process after the training with the real value of the training set, calculating the mean square error by the following formula,
the fitting MSE of the gaussian process regression model is shown in table 6, and it can be seen from table 6 that the optimized gaussian process regression model has higher precision.
TABLE 6 fitting RMSE of Gaussian Process regression models
| Daily output model MSE | Cumulative yield model MSE |
| 5.2368 | 13.2729 |
Step 3, establishing a target function for genetic algorithm optimization by using the obtained two Gaussian process regression models,
max G＝lgG 1 (n f ,l half ,K)·lgG 2 (n f ,l half ,K)
step 4, assuming the population scale of the genetic algorithm, randomly setting an initial value of population iteration, defining a fitness function and calculating the fitness, and selecting excellent individuals in the population by using a roulette method according to the fitness,
(1) the population size of the genetic algorithm was set to 20 (population size refers to the number of individuals in the population, each individual being a set of fracture parameter combinations), and the initial value of the population iteration was randomly given, as shown in table 7,
TABLE 7 initial population of genetic algorithms
(2) Defining a fitness function (because the objective function value is always non-negative, the objective function is directly taken as the fitness function), and taking the value of the population (namely all crack parameter combinations n in the population) f ，l half And K) substituting into the step 3 by two Gaussian process regression models G 1 ，G 2 In the established target function G, each individual in the population (namely a crack parameter combination n) is obtained f ，l half K), the fitness corresponding to the initial population is shown in Table 8,
TABLE 8 fitness value of the initial population
| Serial number | Fitness | Serial number | Degree of |
| 1 | 20.48 | 11 | 24.32 |
| 2 | 17.97 | 12 | 22.76 |
| 3 | 22.45 | 13 | 19.87 |
| 4 | 16.89 | 14 | 19.36 |
| 5 | 21.87 | 15 | 19.47 |
| 6 | 16.84 | 16 | 18.07 |
| 7 | 18.69 | 17 | 20.09 |
| 8 | 21.26 | 18 | 20.11 |
| 9 | 22.81 | 19 | 23.18 |
| 10 | 15.82 | 20 | 22.95 |
(3) According to the fitness, excellent individuals in the population are selected by using a roulette method, namely, the optimal fracture parameter combination serial numbers of the current population are 11, 19, 20, 9, 12, 3, 5, 8, 1 and 18 (a roulette method schematic diagram is shown in fig. 5).
Step 5, converting the initial value of the excellent individuals into binary codes, and obtaining a new population through crossing (exchanging partial binary codes of two excellent individuals) and variation (changing a certain value of a certain binary code), which specifically comprises the following steps,
(1) converting the values of the excellent fracture parameter combinations selected in step 4 into binary codes, for example, a binary code of 75 is 01001011, and a binary code of 149 is 10010101;
(2) exchanging the partial binary codes of the superior fracture parameter combinations with each other, i.e., "crossover", e.g., "crossover" 01001011 and 10010101, results in 01000101 and 10011011, i.e., 69 and 165 from 75 and 149;
(3) changing a value of the binary code of the outstanding fracture parameter combination, i.e., "variation", e.g., 11001101, to obtain 11101101, the corresponding decimal value varies from 205 to 237;
(3) and re-decoding the binary codes subjected to the cross and the mutation into 10-system data as a new population for the next genetic algorithm iteration.
The new populations obtained after "crossover" and "mutation" are shown in table 9,
TABLE 9 New populations obtained after "crossover", "mutation
Step 7, finally repeating step 4 and step 5, continuously iterating the population in the genetic algorithm to obtain an optimal individual (i.e. a group of optimal fracture parameters), wherein the final optimal fracture parameters are shown in table 10,
TABLE 10 optimal fracture parameters
| Parameter(s) | Gap interval | Half seam length | Permeability of crack |
| Optimum value | 141.62m | 166.35m | 28.54D |
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (5)
1. A fracture parameter optimization method based on Gaussian process regression is characterized by comprising the following steps,
step 1, simulating a group of daily output data and accumulated output data of a compact gas fracturing well under different fracture parameters by using a fracture numerical simulation model as a basic data set, and establishing implicit functions of the fracture parameters, the daily output and the accumulated output in the basic data set;
step 2, dividing the basic data set into a training set and a test set, obtaining a covariance matrix according to the training set, and carrying out Gaussian process regression training by using the training set data based on a maximum likelihood estimation method to respectively obtain a Gaussian process regression daily yield model and a Gaussian process regression accumulated yield model;
step 3, establishing a target function based on maximum yield according to two Gaussian process regression models;
and 4, setting an initial population, preferably selecting a better individual in the initial population by using a roulette method, coding the better individual, and performing iterative computation on the better individual by using a genetic algorithm to preferably select an optimal crack parameter.
2. The method for optimizing fracture parameters based on Gaussian process regression as claimed in claim 1, wherein the step 1 specifically comprises the following steps,
step 101, dividing a matrix grid, and dispersing cracks into a plurality of crack segments through the matrix grid;
102, calculating a channeling coefficient between a matrix and a fracture;
103, calculating a channeling coefficient between the cracks;
104, constructing and solving a seepage equation, and calculating daily output data and accumulated output data under different fracture parameters to form a group of basic data sets;
step 105, establishing implicit functions of crack parameters, daily yield and accumulated yield:
y 1 ＝f 1 (x)
y 2 ＝f 2 (x)
in the formula, y 1 For daily output, y 2 For cumulative production, x is the fracture parameters including fracture permeability, fracture length, fracture spacing.
3. The method for optimizing fracture parameters based on Gaussian process regression as claimed in claim 1, wherein the step 2 specifically comprises the following steps,
step 201, extracting a plurality of samples from a basic data set by adopting a Bootstrap resampling method to form a new set, taking the new set as a training set after removing repeated samples in the new set, and taking a complementary set of the basic data set and the new set as a test set;
step 202, calculating an initial covariance matrix according to a covariance matrix formula and a training set;
step 203, making the distribution mean equal to zero, and constructing a prior gaussian distribution model of the implicit function through the covariance matrix:
f(x)＝GP(0,K)
wherein GP (0, K) represents a Gaussian distribution, wherein K is a covariance matrix;
204, optimizing parameters of Gaussian process regression by using a maximum likelihood estimation method, calculating an optimized covariance matrix, and obtaining an optimized Gaussian distribution model according to the optimized covariance matrix;
and step 205, obtaining an optimized Gaussian distribution relation according to the optimized covariance matrix, and training a Gaussian process regression model to respectively obtain a Gaussian process regression daily yield model and a Gaussian process regression cumulative yield model.
4. The method for optimizing fracture parameters based on Gaussian process regression as claimed in claim 1, wherein the objective function for genetic algorithm optimization established in step 3 is,
max G＝lgG 1 (n f ,l half ,σ)·lgG 2 (n f ,l half ,σ)
in the formula, G 1 (n f ,l half σ) represents a gaussian process regression daily yield model; g 2 (n f ,l half σ) represents a Gaussian process regression cumulative yield model; n is f ，l half And sigma respectively represents the crack gap, the crack half-crack length and the crack permeability.
5. The method for optimizing fracture parameters based on Gaussian process regression as claimed in claim 1, wherein the step 4 comprises the following sub-steps,
step 401, randomly generating an initial population;
step 402, substituting the individuals of the population into an objective function to obtain the fitness corresponding to each individual in the population, and preferably selecting the better individual according to a roulette method;
step 403, performing binary coding on the better individuals, generating new individuals through crossing and mutation, decoding the new individuals and calculating the fitness value of the new individuals;
and step 404, continuously iterating until the preset iteration times are reached, and preferably selecting the optimal individual to obtain the target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554894.3A CN114925606A (en) | 2022-05-19 | 2022-05-19 | Fracture parameter optimization method based on Gaussian process regression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554894.3A CN114925606A (en) | 2022-05-19 | 2022-05-19 | Fracture parameter optimization method based on Gaussian process regression |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114925606A true CN114925606A (en) | 2022-08-19 |
Family
ID=82809711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210554894.3A Pending CN114925606A (en) | 2022-05-19 | 2022-05-19 | Fracture parameter optimization method based on Gaussian process regression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114925606A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100250216A1 (en) * | 2009-03-24 | 2010-09-30 | Chevron U.S.A. Inc. | System and method for characterizing fractures in a subsurface reservoir |
| CN112922582A (en) * | 2021-03-15 | 2021-06-08 | 西南石油大学 | Gas well wellhead choke tip gas flow analysis and prediction method based on Gaussian process regression |
-
2022
- 2022-05-19 CN CN202210554894.3A patent/CN114925606A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100250216A1 (en) * | 2009-03-24 | 2010-09-30 | Chevron U.S.A. Inc. | System and method for characterizing fractures in a subsurface reservoir |
| CN112922582A (en) * | 2021-03-15 | 2021-06-08 | 西南石油大学 | Gas well wellhead choke tip gas flow analysis and prediction method based on Gaussian process regression |
Non-Patent Citations (3)
| Title |
|---|
| 刘开云;方昱;刘保国;: "基于进化高斯过程回归算法的隧道工程弹塑性模型参数反演", 岩土工程学报, no. 06, 15 June 2011 (2011-06-15), pages 883 - 889 * |
| 樊冬艳 等: "基于自适应遗传算法的压裂水平井参数优化", 油气地质与采收率, vol. 18, no. 05, 25 September 2011 (2011-09-25), pages 85 - 88 * |
| 陈倩: "致密砂岩气藏水平井分段压裂裂缝参数优化研究", 中国优秀硕士学位论文全文数据库（电子期刊）, 15 June 2020 (2020-06-15), pages 019 - 123 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106295199B (en) | Automatic history matching method and system based on autocoder and multiple-objection optimization | |
| CN113807432B (en) | Air temperature forecast data correction method based on deep learning | |
| CN111160520A (en) | BP neural network wind speed prediction method based on genetic algorithm optimization | |
| CN110807544B (en) | Oil field residual oil saturation distribution prediction method based on machine learning | |
| CN111899254A (en) | Method for automatically labeling industrial product appearance defect image based on semi-supervised learning | |
| CN114840900B (en) | Automatic generation method of derivative BIM (building information modeling) component based on i-GBDT (generic block transformation) technology | |
| CN111695666A (en) | Wind power ultra-short term conditional probability prediction method based on deep learning | |
| CN104865827B (en) | Oil pumping unit oil extraction optimization method based on multi-working-condition model | |
| CN110334438A (en) | A kind of air pollutant emission inventory inversion method and equipment | |
| CN104951847A (en) | Rainfall forecast method based on kernel principal component analysis and gene expression programming | |
| CN109583588B (en) | Short-term wind speed prediction method and system | |
| CN110852496A (en) | Natural gas load prediction method based on LSTM recurrent neural network | |
| CN105608295A (en) | Multi-objective evolutionary algorithm (MOEA) and radial basis function (RBF) neural network optimization modeling method of coking furnace pressure | |
| CN112819225A (en) | Carbon market price prediction method based on BP neural network and ARIMA model | |
| CN114819374A (en) | Regional new energy ultra-short term power prediction method and system | |
| CN109840335A (en) | Based on the radial forging pit prediction optimization method for strengthening T-S fuzzy neural network | |
| CN113172097B (en) | Cold-rolled strip shape pattern recognition method and system based on depth self-coding | |
| CN114118401A (en) | Neural network-based power distribution network flow prediction method, system, device and storage medium | |
| CN114925606A (en) | Fracture parameter optimization method based on Gaussian process regression | |
| CN105427583B (en) | A kind of highway traffic data compression method encoded based on LZW | |
| CN111859807A (en) | Initial pressure optimizing method, device, equipment and storage medium for steam turbine | |
| CN116011319A (en) | Urban expansion simulation method based on driving factor analysis | |
| CN114862007A (en) | Short-period gas production rate prediction method and system for carbonate gas well | |
| CN114492964A (en) | Photovoltaic power ultra-short term probability prediction method based on wavelet decomposition and optimized deep confidence network | |
| CN110704956B (en) | Cold rolling mill data-driven technological parameter optimization method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |