CN114117898B - Machine learning algorithm-based gamma logging while drilling forward modeling method - Google Patents
Machine learning algorithm-based gamma logging while drilling forward modeling method Download PDFInfo
- Publication number
- CN114117898B CN114117898B CN202111338652.2A CN202111338652A CN114117898B CN 114117898 B CN114117898 B CN 114117898B CN 202111338652 A CN202111338652 A CN 202111338652A CN 114117898 B CN114117898 B CN 114117898B
- Authority
- CN
- China
- Prior art keywords
- gamma
- stratum
- detector
- drilling
- network
- 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.)
- Active
Links
- 238000005553 drilling Methods 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 19
- 238000010801 machine learning Methods 0.000 title claims abstract description 19
- 230000004044 response Effects 0.000 claims abstract description 29
- 238000001514 detection method Methods 0.000 claims abstract description 15
- 238000013528 artificial neural network Methods 0.000 claims abstract description 9
- 230000008569 process Effects 0.000 claims abstract description 7
- 230000005251 gamma ray Effects 0.000 claims abstract description 6
- 238000012549 training Methods 0.000 claims description 26
- 230000006870 function Effects 0.000 claims description 24
- 230000002285 radioactive effect Effects 0.000 claims description 13
- 210000002569 neuron Anatomy 0.000 claims description 12
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 10
- 230000015572 biosynthetic process Effects 0.000 claims description 10
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000008689 nuclear function Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 3
- 239000006185 dispersion Substances 0.000 claims description 3
- 230000005284 excitation Effects 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 238000003672 processing method Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 239000003208 petroleum Substances 0.000 abstract description 5
- 238000004088 simulation Methods 0.000 abstract description 3
- 238000003062 neural network model Methods 0.000 abstract 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000005259 measurement Methods 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
- E21B47/00—Survey of boreholes or wells
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
- E21B49/00—Testing the nature of borehole walls; Formation testing; Methods or apparatus for obtaining samples of soil or well fluids, specially adapted to earth drilling or wells
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
-
- 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
- G06N3/048—Activation functions
-
- 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
- G06N3/08—Learning methods
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
- E21B2200/00—Special features related to earth drilling for obtaining oil, gas or water
- E21B2200/22—Fuzzy logic, artificial intelligence, neural networks or the like
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mining & Mineral Resources (AREA)
- Geology (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Environmental & Geological Engineering (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Fluid Mechanics (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geochemistry & Mineralogy (AREA)
- Geophysics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a gamma logging while drilling forward modeling method based on a machine learning algorithm, and relates to the technical field of petroleum well detection. The method specifically comprises the following steps: establishing stratum models with different radioactivity intensity, layer thickness, density and different inclination angles; simulating the transportation process of gamma rays in the stratum, and acquiring the response relation of the count of the gamma detector while drilling with depth under different geological conditions; establishing response relations between depth coordinates of the detector and counts of the gamma detector while drilling under different stratum thicknesses; establishing a response relation between the depth coordinates of the detector and the count of the gamma detector while drilling under different stratum densities; establishing response relations between depth coordinates of the detector and counts of the gamma-ray detector while drilling under different stratum dip angles; establishing response relations between different stratum interface distances and gamma counts while drilling; based on the response relation, a database is established, the simulation data are trained by a neural network algorithm, a neural network model suitable for gamma forward modeling while drilling is constructed, and a gamma logging while drilling rapid forward modeling method is formed.
Description
Technical Field
The invention relates to the technical field of petroleum well detection, in particular to a gamma logging while drilling forward modeling method based on a machine learning algorithm.
Background
Petroleum resources are important resources in the industry. In recent years, the high-inclination well and the horizontal well in the petroleum drilling field are continuously increased, and the importance and difficulty of well logging and exploration are also increased. Gamma-ray logging while drilling is carried out, a gamma detector is placed in a drill bit, and formation information is obtained by measuring gamma-ray radioactivity of a formation in the drilling process. The method is an important means in logging of highly deviated wells and horizontal wells, forward modeling is to simulate the detection count of a gamma detector in underground stratum on the basis of known stratum information, forward modeling is carried out by utilizing stratum information of an adjacent well before logging, and stratum is readjusted after comparison with an actual measurement result so as to be in line with the actual measurement result, thereby achieving the purpose of obtaining a new stratum model, and being a particularly important step in actual drilling operation.
The commonly used gamma logging while drilling forward modeling method is to build a stratum model by using stratum information, and then simulate the detection count of the stratum by using a Monte Carlo simulation method. Although the method has extremely high accuracy, the time consumed by operation is too long, and the efficiency in actual engineering is affected.
In recent years, as the development of computer technology, machine learning algorithms have been vigorous. Through training of mass data, the model generated by the machine learning algorithm can rapidly complete analysis of the studied problems while ensuring accuracy. According to the invention, a stratum model is established through the MCNP, key parameters in the stratum model are changed, a large amount of data is simulated by using a Monte Carlo method, analysis, training and verification are performed by using a machine learning algorithm, a network model relation between stratum parameters and gamma logging response is established, an efficient and accurate forward modeling method of gamma logging while drilling is formed, and guidance is provided for updating and geosteering of the stratum model.
Disclosure of Invention
The invention aims at overcoming the defect of the simulation speed of Monte Carlo simulation, and provides a gamma logging while drilling forward modeling method based on a machine learning algorithm, which can greatly improve the gamma forward modeling speed while drilling.
The invention adopts the following technical scheme:
A gamma logging while drilling forward modeling method based on a machine learning algorithm comprises the following steps:
a. building logging models with different stratum radioactivity sizes, layer thicknesses, densities and different dip angles;
b. Simulating the transportation process of gamma particles in the stratum by using Monte Carlo software to obtain response values of the gamma detector while drilling, which are counted and change with depth under different geological conditions;
c. Establishing a response relation between the formation radioactivity and the count of the gamma detector while drilling; establishing response relations between depth coordinates of the detector and counts of the gamma detector while drilling under different stratum thicknesses; establishing a response relation between the depth coordinates of the detector and the count of the gamma detector while drilling under different stratum densities; establishing response relations between depth coordinates of the detector and counts of the gamma-ray detector while drilling under different stratum dip angles; establishing response relations between different stratum interface distances and gamma counts while drilling;
d. According to the radioactivity size difference contained in the two strata, the data are divided into: three response relation classifications with overlarge radioactivity, smaller radioactivity and moderate radioactivity are established;
e. preprocessing a network, constructing a network, training the network and adjusting parameters;
f. Importing the other 25% of data of the database into a trained network for machine learning, performing regression prediction to obtain a forward result, and adjusting network parameters again;
g. and selecting the same stratum parameters which do not exist in the database, respectively carrying out regression prediction by using Monte Carlo simulation and a trained network model, and comparing the results of the Monte Carlo simulation and the trained network model.
Preferably, in step d, in order to eliminate the dimensional influence between the data, the data of the maximum detection range of the detector at different stratum positions is divided into: the classified data is split into two databases, namely 6 new databases in total, all under the influence of one radioactive stratum and under the cross influence of two different radioactive strata. And then carrying out data processing methods such as data dispersion standardization, standard deviation standardization, logarithmic transformation, reverse order, weight bias and the like on each database, and perfecting the databases.
Preferably, the specific process of step e is:
Preprocessing a network and constructing a network: the deep neural network is selected as a network model, the structure of the neural network is provided with n input layers, the number of the n input layers is determined by different databases, and input values, input shapes, output shapes and excitation functions are preset according to the data size. 4 layers of temporary hidden layers, namely 4, 8 and 12 temporary hidden layer neurons, and the activation function of the temporary hidden layer neurons is ReLU; the hidden layer neurons and the ReLU activation functions are divided into four groups, an output layer is arranged at last, and a random gradient descent algorithm is selected as an optimization algorithm; for the more complex label situation, a SVR model in a Support Vector Machine (SVM) can be selected to carry out regression prediction, different kernel functions are selected for different databases, and Linear can be selected as the kernel function no matter how different the radioactivity is, under the condition that the maximum detection range of the detector is completely in the influence of a radioactive stratum. For a database with the maximum detection range of the detector under the cross influence of two different radioactive strata, polynomial can be selected as a kernel function when the radioactivity of the strata has an excessive difference; when the radioactivity of the stratum is too small, RBF can be selected as a nuclear function; when the radioactivity phase difference is moderate, RBF can be selected as a kernel function.
Training network and adjusting parameters: randomly sampling the classified database, taking 75% of data in the database as a training set, training the selected machine learning model, and adjusting the number of neuron layers and the number of neurons during training, namely adjusting the depth and width of a DNN (direct current network) or adjusting parameters such as a kernel function, gamma, coef0, espilon and the like. The value of the parameter depends on the degree of network training, and the prediction accuracy of the DNN network adopts an average absolute error (MAE) as an index of calculation Loss (Loss), and a Mean Square Error (MSE) as a basis of measurement accuracy (Acc). Both of which are smaller and better. The accuracy of regression prediction of SVR is evaluated by not limited to R2_score, and the method is very suitable for the case of multi-label and multi-dimension, and the closer the value of R2_score is to 1, the more accurate the value is. And adjusting parameters according to the evaluation feedback of MAE, MSE, R2_score, and feeding the parameters back to the evaluation value until the network training is completed. The invention has the following beneficial effects:
The method trains an accurate model by utilizing a machine learning algorithm, and realizes the rapid forward generation of gamma curves of stratum with different API values, stratum with different densities and multiple strata.
By the method, the speed of gamma forward modeling while drilling is greatly improved, the accuracy of simulating gamma response while drilling is ensured, and the forward modeling method can be applied to production logging in real time. And the program has strong inclusion, after the network training is finished, the logging curve can be quickly generated by only inputting data, and logging data can be processed by non-petroleum logging professionals.
Drawings
FIG. 1 is a schematic diagram of a Monte Carlo calculation model of a well logging;
FIG. 2 is a schematic diagram of a gamma forward database while drilling;
FIG. 3 is a schematic diagram of a neural network;
FIG. 4 is a schematic diagram of the response of the gamma count while drilling to formation depth of a Monte Carlo simulation as the API changes;
FIG. 5 is a schematic diagram showing the comparison of the simulation forward result based on the algorithm of the present invention and Meng Ka under the same formation parameters;
fig. 6 is a flow chart of a forward modeling method of gamma logging while drilling based on a machine learning algorithm.
Detailed Description
The following description of the embodiments of the invention will be given with reference to the accompanying drawings and examples:
Referring to fig. 6, a gamma logging while drilling forward modeling method based on a machine learning algorithm includes the following steps:
a. Logging models of different formation radioactivity sizes, layer thicknesses, densities and different dip angles were created as shown in fig. 1.
B. Simulating the transportation process of gamma particles in the stratum by using Monte Carlo software to obtain response values of the gamma detector while drilling, which are counted and change with depth under different geological conditions; after formation modeling, a training dataset is generated using Monte Carlo simulation with API parameters set to upper layer APIs from 10 to 130 and lower layer APIs from 20 to 300. The data set includes API values of the upper and lower strata, depth of the instrument measurement points from the surface, and counts of the individual position detectors, together comprising a set of data 297200 as shown in fig. 2.
C. Establishing a response relation between the formation radioactivity and the count of the gamma detector while drilling; establishing response relations between depth coordinates of the detector and counts of the gamma detector while drilling under different stratum thicknesses; establishing a response relation between the depth coordinates of the detector and the count of the gamma detector while drilling under different stratum densities; establishing response relations between depth coordinates of the detector and counts of the gamma-ray detector while drilling under different stratum dip angles; establishing response relations between different stratum interface distances and gamma counts while drilling;
d. Classifying different response relations by a decision tree classification method, and establishing a database; in order to eliminate the dimensional influence among the data, the data of the maximum detection range of the detector at different stratum positions are divided into: the classified data is split into two databases, namely 6 new databases in total, all under the influence of one radioactive stratum and under the cross influence of two different radioactive strata. And then carrying out data processing methods such as data dispersion standardization, standard deviation standardization, logarithmic transformation, reverse order, weight bias and the like on each database, and perfecting the databases.
E. Preprocessing a network, constructing a network, training the network and adjusting parameters; the neural network structure is shown in fig. 3, a deep neural network is selected as a network model, n structural input layers of the neural network are adopted, the number n of the input layers is determined by different databases, and input values, input shapes, output shapes and excitation functions are preset according to the data size. 4 layers of temporary hidden layers, namely 4, 8 and 12 temporary hidden layer neurons, and the activation function of the temporary hidden layer neurons is ReLU; the hidden layer neurons and the ReLU activation functions are divided into four groups, an output layer is arranged at last, and a random gradient descent algorithm is selected as an optimization algorithm; for the more complex label situation, a SVR model in a Support Vector Machine (SVM) can be selected to carry out regression prediction, different kernel functions are selected for different databases, and Linear can be selected as the kernel function no matter how different the radioactivity is, under the condition that the maximum detection range of the detector is completely in the influence of a radioactive stratum. For a database with the maximum detection range of the detector under the cross influence of two different radioactive strata, polynomial can be selected as a kernel function when the radioactivity of the strata has an excessive difference; when the radioactivity of the stratum is too small, RBF can be selected as a nuclear function; when the radioactivity phase difference is moderate, RBF can be selected as a kernel function.
Training network and adjusting parameters: randomly sampling the classified database, taking 75% of data in the database as a training set, training the selected machine learning model, and adjusting the number of neuron layers and the number of neurons during training, namely adjusting the depth and width of a DNN (direct current network) or adjusting parameters such as a kernel function, gamma, coef0, espilon and the like. The value of the parameter depends on the degree of network training, and the prediction accuracy of the DNN network adopts an average absolute error (MAE) as an index of calculation Loss (Loss), and a Mean Square Error (MSE) as a basis of measurement accuracy (Acc). Both of which are smaller and better. The accuracy of regression prediction of SVR is evaluated by not limited to R2_score, and the method is very suitable for the case of multi-label and multi-dimension, and the closer the value of R2_score is to 1, the more accurate the value is. And adjusting parameters according to the evaluation feedback of MAE, MSE, R2_score, and feeding the parameters back to the evaluation value until the network training is completed.
For more complex label cases, training speed is accelerated. A SVR model in a Support Vector Machine (SVM) is selected to carry out regression prediction, different kernel functions are selected for different databases, and Linear can be selected as the kernel function no matter how different the radioactivity is, under the condition that the maximum detection range of the detector is completely in the influence of a radioactive stratum. For a database with the maximum detection range of the detector under the cross influence of two different radioactive strata, polynomial can be selected as a kernel function when the radioactivity of the strata has an excessive difference; when the radioactivity of the stratum is too small, RBF can be selected as a nuclear function; when the radioactivity phase difference is moderate, RBF can be selected as a kernel function.
F. and importing the other 25% of data of the database into a trained network for machine learning, carrying out regression prediction to obtain a forward result, and adjusting network parameters again, as shown in fig. 4.
G. The same formation parameters not existing in the database are selected, regression prediction is performed by using the Monte Carlo simulation and the trained network model respectively, and the results of the two are compared, as shown in FIG. 5.
It should be understood that the above description is not intended to limit the invention to the particular embodiments disclosed, but to limit the invention to the particular embodiments disclosed, and that the invention is not limited to the particular embodiments disclosed, but is intended to cover modifications, adaptations, additions and alternatives falling within the spirit and scope of the invention.
Claims (1)
1. The gamma logging while drilling forward modeling method based on the machine learning algorithm is characterized by comprising the following steps of:
a. building logging models with different stratum radioactivity sizes, layer thicknesses, densities and different dip angles;
b. Simulating the transportation process of gamma particles in the stratum by using Monte Carlo software to obtain theoretical values of the change of the count of the gamma detector while drilling with the depth under different geological conditions;
c. Establishing a response relation between the formation radioactivity and the count of the gamma detector while drilling; establishing response relations between depth coordinates of the detector and counts of the gamma detector while drilling under different stratum thicknesses; establishing a response relation between the depth coordinates of the detector and the count of the gamma detector while drilling under different stratum densities; establishing response relations between depth coordinates of the detector and counts of the gamma-ray detector while drilling under different stratum dip angles; establishing response relations between different stratum interface distances and gamma counts while drilling;
d. according to the radioactivity size difference contained in the two strata, the data are divided into: three response relation classifications with overlarge radioactivity, smaller radioactivity and moderate radioactivity are established;
e. preprocessing a network, constructing a network, training the network and adjusting parameters;
f. importing the other 25% of data of the database into a trained network for machine learning, performing regression prediction to obtain a forward result, and adjusting network parameters again;
g. selecting the same stratum parameters which do not exist in the database, respectively carrying out regression prediction by using Monte Carlo simulation and a trained network model, and comparing the results of the Monte Carlo simulation and the trained network model;
In step d, in order to eliminate the dimensional influence among the data, the data of the maximum detection range of the detector at different stratum positions are divided into: the classified data are divided into two databases under the cross influence of two different radioactive strata, namely 6 new databases are generated in total; performing a data processing method of dispersion standardization, standard deviation standardization, logarithmic transformation, reverse order and weight bias on the data of each database to perfect the database;
the specific process of the step e is as follows:
Preprocessing a network and constructing a network: selecting a deep neural network as a network model, wherein n structural input layers of the neural network are adopted, the number n of the input layers is determined by different databases, and an input value, an input shape, an output shape and an excitation function are preset according to the data size; for the more complex label situation, SVR models in a support vector machine SVM can be selected to carry out regression prediction, different kernel functions are selected for different databases, and Linear can be selected as the kernel function no matter how different the radioactivity is, under the condition that the maximum detection range of the detector is completely in the influence of a radioactive stratum; for a database with the maximum detection range of the detector under the cross influence of two different radioactive stratum, polynomial can be selected as a nuclear function when the radioactivity phase difference of the stratum is too large, RBF is selected as the nuclear function when the radioactivity phase difference of the stratum is too small, and RBF is selected as the nuclear function when the radioactivity phase difference is moderate;
Training network and adjusting parameters: randomly sampling the classified database, taking 75% of data in the database as a training set, training the selected machine learning model, and adjusting the number of neuron layers and the number of neurons during training, namely adjusting the depth and width of a DNN (direct current network) or adjusting the parameters of a kernel function, gamma, coef0 and espilon; the numerical value of the parameter depends on the degree of network training, the prediction accuracy of the DNN adopts an average absolute error MAE as an index for calculating Loss, and a mean square error MSE is used as a basis for measuring accuracy Acc; and adjusting parameters according to the evaluation feedback of MAE, MSE, R2_score, and feeding the parameters back to the evaluation value until the network training is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111338652.2A CN114117898B (en) | 2021-11-12 | 2021-11-12 | Machine learning algorithm-based gamma logging while drilling forward modeling method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111338652.2A CN114117898B (en) | 2021-11-12 | 2021-11-12 | Machine learning algorithm-based gamma logging while drilling forward modeling method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114117898A CN114117898A (en) | 2022-03-01 |
| CN114117898B true CN114117898B (en) | 2024-06-28 |
Family
ID=80379288
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111338652.2A Active CN114117898B (en) | 2021-11-12 | 2021-11-12 | Machine learning algorithm-based gamma logging while drilling forward modeling method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114117898B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115393656B (en) * | 2022-10-26 | 2023-01-24 | 中石化经纬有限公司 | Automatic classification method for stratum classification of logging-while-drilling image |
| CN115749724A (en) * | 2022-11-15 | 2023-03-07 | 核工业北京化工冶金研究院 | Drilling control method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103774988A (en) * | 2013-06-05 | 2014-05-07 | 中国石油大学(华东) | Real-time while-drilling gamma forward modeling geo-steering drilling method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108919331B (en) * | 2018-07-03 | 2020-07-14 | 上海交通大学 | Double-detector spiral gamma scanning measurement method for radioactive waste barrel |
| CN109657346A (en) * | 2018-12-18 | 2019-04-19 | 中国石油大学(华东) | Based on integral calculation with boring orientation gamma forward modeling method, device and equipment |
| WO2021075994A1 (en) * | 2019-10-16 | 2021-04-22 | Saudi Arabian Oil Company | Determination of elastic properties of a geological formation using machine learning applied to data acquired while drilling |
-
2021
- 2021-11-12 CN CN202111338652.2A patent/CN114117898B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103774988A (en) * | 2013-06-05 | 2014-05-07 | 中国石油大学(华东) | Real-time while-drilling gamma forward modeling geo-steering drilling method |
Non-Patent Citations (1)
| Title |
|---|
| 基于空间灵敏度函数的随钻伽马测井快速正演方法;宗畅;《中国核科学技术进展报告（第七卷）——中国核学会2021年学术年会论文集第7册》;20211031;第138-147 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114117898A (en) | 2022-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111783825B (en) | Logging lithology recognition method based on convolutional neural network learning | |
| CN114117898B (en) | Machine learning algorithm-based gamma logging while drilling forward modeling method | |
| Ghiasi-Freez et al. | Improving the accuracy of flow units prediction through two committee machine models: an example from the South Pars Gas Field, Persian Gulf Basin, Iran | |
| CN112989708B (en) | Well logging lithology identification method and system based on LSTM neural network | |
| Majdi et al. | Applying evolutionary optimization algorithms for improving fuzzy C-mean clustering performance to predict the deformation modulus of rock mass | |
| CN103824329B (en) | Geological exploration three-dimensional visual reserve estimation method | |
| CN114723095A (en) | Missing well logging curve prediction method and device | |
| CN112652066B (en) | Three-dimensional geological model-based geological characterization condition evaluation method and system | |
| CN108952699A (en) | A kind of complicated geological drilling process formation lithology intelligent identification Method | |
| CN113610945A (en) | Ground stress curve prediction method based on hybrid neural network | |
| CN113902861A (en) | Three-dimensional geological modeling method based on machine learning | |
| Jiang et al. | Development of a core feature identification application based on the Faster R-CNN algorithm | |
| CN114861519A (en) | Initial ground stress field acceleration optimization inversion method under complex geological condition | |
| CN107633556A (en) | It is a kind of quantitatively to obtain the probabilistic method of three dimensional ore deposit geological model | |
| CN111079783A (en) | Method for identifying stratum lithology parameters based on multi-core ensemble learning | |
| Goodwin et al. | Stochastic modeling of subseismic faults conditioned on displacement and orientation maps | |
| Ji et al. | An automated method to build 3D multi-scale geological models for engineering sedimentary layers with stratum lenses | |
| CN114114414A (en) | Artificial intelligence prediction method for 'dessert' information of shale reservoir | |
| CN116911216B (en) | Reservoir oil well productivity factor assessment and prediction method | |
| CN116721227A (en) | Automatic modeling method for three-dimensional geologic model of complex geologic body | |
| US20230141334A1 (en) | Systems and methods of modeling geological facies for well development | |
| CN115809411A (en) | Improved decision tree lithology identification method based on logging data | |
| Bertoncello | Conditioning surface-based models to well and thickness data | |
| Song et al. | Reservoir lithology identification based on improved adversarial learning | |
| CN108537883B (en) | Geological modeling method based on MapReduce framework |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |