CN111783825B - Logging lithology recognition method based on convolutional neural network learning - Google Patents
Logging lithology recognition method based on convolutional neural network learning Download PDFInfo
- Publication number
- CN111783825B CN111783825B CN202010456323.7A CN202010456323A CN111783825B CN 111783825 B CN111783825 B CN 111783825B CN 202010456323 A CN202010456323 A CN 202010456323A CN 111783825 B CN111783825 B CN 111783825B
- Authority
- CN
- China
- Prior art keywords
- lithology
- neural network
- convolutional neural
- data
- training
- 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
Classifications
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/30—Assessment of water resources
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a logging lithology recognition method based on convolutional neural network learning, which comprises the following steps of 1, taking a data curve acquired by drilling and coring as an input characteristic; taking a drilling lithology result as an input characteristic label, cleaning sample data and establishing a learning data sample; 2. sequentially arranging three-porosity, three-resistivity and three-lithology curves in sequence, classifying the drilling lithology into four types, and classifying the learning data sample into a training set and a testing set; 3. extracting characteristic parameters by adopting primary convolution and primary pooling, linking a Softmax regression layer, and establishing a convolutional neural network model; 4. training a convolutional neural network model, testing the accuracy of the convolutional neural network model by using a test set, putting the convolutional neural network model into practice if the required accuracy is met, and increasing the training amount if the required accuracy is not met; 5. and identifying the lithology of the new well by using the trained convolutional neural network model. The rock stratum information can be accurately identified, and the convergence speed is high.
Description
Technical Field
The invention belongs to the field of rock stratum exploration, and relates to a logging lithology recognition method based on convolutional neural network learning.
Background
Lithology is an overall reflection of the deposition, structure, construction and mineral composition of underground rock, and accurate identification of lithology is of great significance to reservoir partitioning, hydrocarbon reservoir identification and reservoir evaluation.
The formation lithology identification includes various methods such as field outcrop, well drilling coring, seismic inversion, well logging interpretation and the like, the well logging interpretation is usually based on one or two empirical formulas of well logging curves, lithology is judged by calculating the content of components such as argillaceous, coal, calcite, dolomite and the like, and the formation lithology identification method includes a cross-plot method, a formation element well logging method and the like, but the methods cannot fully excavate lithology information in all well logging curves and have certain limitations. Secondly, the lithology of the logging curve is automatically identified by a support vector machine, a random forest, a BP neural network and other methods, but the methods have a relatively slow convergence rate, and the methods are easy to enter gradient disappearance, gradient explosion and the like, so that generalization is not ideal.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a logging lithology recognition method based on convolutional neural network learning, which can more accurately recognize rock formation information and has high convergence rate.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme:
A logging lithology recognition method based on convolutional neural network learning comprises the following steps of;
Taking a data curve acquired by drilling and coring as an input characteristic, wherein the data curve comprises natural potential, natural gamma, borehole diameter, deep induction, middle induction, eight lateral directions, acoustic wave time difference, and compensation neutron and volume density; taking a drilling lithology result as an input characteristic label, cleaning sample data and establishing a learning data sample;
Sequentially arranging three-porosity, three-resistivity and three-lithology curves in sequence, classifying the drilling lithology into four types, and classifying the learning data sample into a training set and a testing set;
thirdly, extracting characteristic parameters by adopting primary convolution and primary pooling, linking a Softmax regression layer, and establishing a convolutional neural network model;
Training a convolutional neural network model, testing the accuracy of the convolutional neural network model by using a test set, putting the convolutional neural network model into practice if the required accuracy is met, and increasing the training amount if the required accuracy is not met;
and fifthly, identifying the lithology of the new well by using the trained convolutional neural network model.
Preferably, in the first step, when the sample data are cleaned, the data samples of the thin layer, the lithology mutation section and the well wall collapse section are removed.
Preferably, in the first step, the data curve is subjected to depth correction and then is discretized into data with a sampling interval of 0.125 m.
Further, the porosity analysis result and the acoustic moveout calculation result are matched for depth correction.
Preferably, in the second step, the lithology of the well is classified into fine sandstone, argillaceous siltstone, siltstone and mudstone.
Preferably, in the third step, the activating function adopts Sigmoid, the gradient descent adopts a self-adaptive gradient descent method, the loss function adopts a square difference function, and the regularization adopts L2 regularization.
Preferably, in the third step, the sample convolution matrix is a 3×3 matrix, and the convolution kernels of four types of drilling lithology adopt a2×2 matrix; the output layer is a four-dimensional probability matrix, fine sandstone is [1, 0], the argillite siltstone is [0,1, 0], the siltstone is [0,1, 0], and the mudstone is [0, 1].
Preferably, in the fourth step, the training sample size Bachsize of a single batch is 128, and the training round Epoch is 40000.
Compared with the prior art, the invention has the following beneficial effects:
According to the method, nine data curves are obtained, so that the extracted logging curve features are more, the reflected formation lithology information is more comprehensive, and the formation lithology can be identified in a higher dimension. And secondly, a convolutional neural network model is adopted, nine data curves are sequenced and classified, so that the convergence rate of the convolutional neural network model is higher, over-fitting and under-fitting can be effectively prevented, and the applicability of the model is improved.
Drawings
FIG. 1 is a schematic diagram of a logging lithology recognition process of the present invention;
FIG. 2 is a schematic diagram of a convolutional neural network architecture of the present invention;
FIG. 3 is a schematic diagram of three activation functions according to the present invention;
FIG. 4 is a three-dimensional visualization screenshot of the training process of the present invention;
FIG. 5 is a graph comparing the automatic predicted outcome of new well lithology with logging coring according to the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the following detailed description of the embodiments of the present invention will be given with reference to the accompanying drawings. The exemplary embodiments of the present invention and their descriptions herein are for the purpose of explaining the present invention, but are not to be construed as limiting the invention.
The embodiment uses the programming interface API provided by TensorFlow in fig. 1 to define a convolutional neural network, tensorFlow is an open source software library for numerical calculation, which is developed for google and adopts a data flow diagram. Based on the source program, the flow of the logging lithology automatic identification method based on convolutional neural network machine learning designed by the invention is shown in figure 1, and all steps can be automatically operated by a person skilled in the art by adopting a computer software technology. The embodiment specifically realizes the following steps:
Step 1, taking a conventional 9 curves collected by a logging instrument as input characteristics, taking lithology results obtained by drilling and coring as labels, removing data samples of a thin layer, lithology abrupt sections and a well wall collapse section, and cleaning sample data to establish machine learning data samples;
Step 1.1, sample data acquisition: acquiring natural potential, natural gamma, well diameter, deep induction, middle induction, eight lateral direction, acoustic time difference, compensated neutron and volume density of a logging curve, and dispersing the logging curve into data with a sampling interval of 0.125m after depth correction;
Step 1.2, sample label making: and calibrating a logging interpretation result according to the lithology description result of the drilling coring. The core data is matched with the acoustic time difference calculation result by utilizing the porosity analysis result to carry out depth correction;
Step 1.3, sample data cleaning: and removing the data samples of the thin layer, the lithology mutation section and the well wall collapse section, and reserving the data of the stratum deposition stable interval.
Step 2, arranging three-porosity, three-resistivity and three-lithology curves in sequence, dividing the drilling lithology into four types, so that the samples are uniformly distributed, and dividing the data samples into a training set and a testing set;
step 2.1, arranging the data sequence according to three porosities, three resistivities and three lithologies, and facilitating the development of convolution operation;
Step 2.2, according to geological characteristics of a research area, lithology data are divided into fine sandstone, argillaceous siltstone and siltstone; redundant samples are deleted, so that four lithology samples are evenly distributed, and the tendency of machine learning results is prevented; the data is divided into a training set and a verification set to form a machine learning sample library. The labels of four lithologies are represented by a matrix: fine sandstone: [1, 0], argillaceous siltstone: [0,1, 0], silty mudstone: [0,1, 0], mudstone: [0, 1];
Step 3, extracting characteristic parameters by adopting primary convolution and primary pooling, linking a Softmax regression layer, and establishing a convolutional neural network model;
fig. 2 is a schematic diagram of activation functions of the pooling layer and the Softmax layer, and three general functions are Sigmoid, relu and Tanh functions respectively, and this training finds that the Sigmoid function has good applicability, so that the Sigmoid function is adopted for carrying out network convergence training.
Step 3.1, building a convolutional neural network according to the input parameter form, selecting an optimal global variable, setting an important parameter adjustment table, and manually searching an optimal parameter;
step 3.2, connecting a pooling layer by adopting a convolution layer, and then connecting a network structure of a Softmax regression layer, wherein an activation function adopts Sigmoid, gradient descent adopts a self-adaptive gradient descent method Adagrad, a loss function adopts a square difference function, and regularization adopts L2 regularization;
Fig. 2 is a schematic diagram of a convolutional neural network framework of the present invention, including a convolutional layer, a pooling layer, and a Softmax layer. 9 log curves are arranged according to the sequence in the figure to form a matrix capable of convolution, and a convolution kernel adopts a 2 multiplied by 2 matrix; the output layer is a four-dimensional probability matrix, and the corresponding relation is as follows: fine sandstone: [1, 0], argillaceous siltstone: [0,1, 0], silty mudstone: [0,1, 0], mudstone: [0, 1];
1) The sample convolution matrix is a3×3 matrix, the convolution kernel is a2×2 matrix, and the arrangement sequence is as follows: input matrix: Convolution kernel:
2) The pooling layer parameter is 32, namely, 32 feature vectors are finally extracted;
3) The activation function adopts a Sigmoid function, and the calculation method is represented by the following formula:
wherein: x is a vector representing the value of the input layer;
f (x) is a vector representing the weight matrix of the output layer;
4) The gradient descent algorithm adopts an adaptive gradient descent method Adagrad, and the calculation method is represented by the following formula:
where G t is a diagonal matrix, each diagonal position i, i is the sum of squares of the gradient of the corresponding parameter θ i from round 1 to round t. The E is a smooth term, in order to avoid zero denominator; θ represents an argument, i.e., one of the 9 data curves.
5) The loss function is a square difference function, and the calculation method is represented by the following formula:
where C is the loss function x representing the sample, y (x) representing the output, Representing the actual value and n representing the total number of samples.
Three activation functions are schematically shown in fig. 3.
FIG. 4 is a process screenshot of the training process visual inspection of the model after the invention has been trained and optimized, stored, at Tensorboard. From the figure, 4 lithologies are gradually separated after 10000 rounds of training, and the classification effect is good.
And 4, training the network model, adjusting training parameters to enable the model to be converged rapidly, and testing the accuracy of the model by using a test set.
Step 4.1, feeding training set samples into a nerve network, and adjusting key super parameters including parameters such as learning rate, batchsize, epoch and the like; wherein, the training sample size Bachsize of a single batch is 128, the training round Epoch is 40000, the observation accuracy and the loss function ensure the model convergence, the loss function is smoothly reduced, the accuracy is steadily increased, and the stability level is reached;
step 4.2, feeding the test set sample data into a neural network to obtain a loss function and accuracy;
And 4.3, repeating the steps 4.1 and 4.2, and considering that the trained model has practical value when the test accuracy rate is not large and reaches more than 85%. Testing the sample number training of each group of parameter combination after each round of iterative training on the test set to obtain the error of the current model on the test set, and stopping training after the sample iteration number reaches Epoch number or the error is no longer reduced on the test set; finally, the ultra-parameter combination with the minimum error on the test set is taken to obtain an optimized model, the accuracy of the training set is 96.3%, and the test set is 85.2%.
FIG. 5 is a graph showing the comparison of the effect of the trained artificial intelligent model in the new well treatment, and the accuracy of prediction can be better by comparing the interpretation and lithology of a convolutional network algorithm and the lithology of logging and coring, most lithology is correctly identified, and only a few mudstone sections are identified and actually different, so that the production requirement can be basically met.
And 5, automatically predicting the lithology of the new well.
And 5.1, finding out new well data, arranging the data according to network setting, feeding the data into a network model to obtain lithology prediction results, comparing the lithology prediction results with drilling coring data, and evaluating the practicability of the lithology prediction results.
And 5.2, processing new wells in batches, and carrying out regional lithology prediction and reservoir evaluation.
The above is only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by this, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the claims of the present invention.
Claims (4)
1. The logging lithology recognition method based on convolutional neural network learning is characterized by comprising the following steps of;
Taking a data curve acquired by drilling and coring as an input characteristic, wherein the data curve comprises natural potential, natural gamma, borehole diameter, deep induction, middle induction, eight lateral directions, acoustic wave time difference, and compensation neutron and volume density; taking a drilling lithology result as an input characteristic label, cleaning sample data and establishing a learning data sample;
Sequentially arranging three-porosity, three-resistivity and three-lithology curves in sequence, classifying the drilling lithology into four types, and classifying the learning data sample into a training set and a testing set;
thirdly, extracting characteristic parameters by adopting primary convolution and primary pooling, linking a Softmax regression layer, and establishing a convolutional neural network model;
the sample convolution matrix is a3 multiplied by 3 matrix, and the convolution kernel of four types of drilling lithology adopts a2 multiplied by 2 matrix; the output layer is a four-dimensional probability matrix, fine sandstone is [1, 0], the argillaceous siltstone is [0,1, 0], the siltstone is [0,1, 0], and the mudstone is [0, 1];
the activation function adopts Sigmoid, the gradient descent adopts a self-adaptive gradient descent method, the loss function adopts a square difference function, and the regularization adopts L2 regularization;
Training a convolutional neural network model, testing the accuracy of the convolutional neural network model by using a test set, putting the convolutional neural network model into practice if the required accuracy is met, and increasing the training amount if the required accuracy is not met;
during training, the training sample size Bachsize of a single batch is 128, and the training round Epoch is 40000;
and fifthly, identifying the lithology of the new well by using the trained convolutional neural network model.
2. The method for identifying lithology of well logging based on convolutional neural network learning according to claim 1, wherein in the first step, when the sample data is cleaned, the data samples of the thin layer, lithology abrupt section and well wall collapse section are removed.
3. The method for identifying lithology of logging based on convolutional neural network learning according to claim 1, wherein in the first step, after the data curve is subjected to depth correction, the data curve is discretized into data with a sampling interval of 0.125 m.
4. The method for identifying logging lithology based on convolutional neural network learning according to claim 3, wherein the depth correction is performed by matching the result of the porosity analysis with the result of the acoustic moveout calculation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010456323.7A CN111783825B (en) | 2020-05-26 | 2020-05-26 | Logging lithology recognition method based on convolutional neural network learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010456323.7A CN111783825B (en) | 2020-05-26 | 2020-05-26 | Logging lithology recognition method based on convolutional neural network learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111783825A CN111783825A (en) | 2020-10-16 |
| CN111783825B true CN111783825B (en) | 2024-06-28 |
Family
ID=72753181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010456323.7A Active CN111783825B (en) | 2020-05-26 | 2020-05-26 | Logging lithology recognition method based on convolutional neural network learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111783825B (en) |
Families Citing this family (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112253087A (en) * | 2020-10-20 | 2021-01-22 | 河南理工大学 | Biological disturbance reservoir physical property calculation method based on multi-source logging data |
| CN114529110A (en) * | 2020-11-03 | 2022-05-24 | 中国石油天然气集团有限公司 | Lithofacies inversion method and system based on deep neural network model |
| CN114528746A (en) * | 2020-11-04 | 2022-05-24 | 中国石油化工股份有限公司 | Complex lithology identification method, identification system, electronic device and storage medium |
| CN112100930B (en) * | 2020-11-11 | 2021-02-02 | 中国石油大学(华东) | Formation pore pressure calculation method based on convolutional neural network and Eaton formula |
| CN112597826A (en) * | 2020-12-08 | 2021-04-02 | 核工业北京地质研究院 | Method for lithologic classification of hyperspectral SASI data |
| CN112396130A (en) * | 2020-12-09 | 2021-02-23 | 中国能源建设集团江苏省电力设计院有限公司 | Intelligent identification method and system for rock stratum in static sounding test, computer equipment and medium |
| CN112712025A (en) * | 2020-12-29 | 2021-04-27 | 东北石油大学 | Complex lithology identification method based on long-term and short-term memory neural network |
| CN112784980B (en) * | 2021-01-05 | 2024-05-28 | 中国石油天然气集团有限公司 | Intelligent logging horizon dividing method |
| CN112906465B (en) * | 2021-01-15 | 2023-12-22 | 阳泉煤业（集团）股份有限公司 | Coal measure stratum acoustic curve reconstruction method and system based on stratum factors |
| CN112796746B (en) * | 2021-02-26 | 2022-06-07 | 西安石油大学 | A drilling method for petroleum geological exploration |
| CN112990320A (en) * | 2021-03-19 | 2021-06-18 | 中国矿业大学(北京) | Lithology classification method and device, electronic equipment and storage medium |
| CN113159136B (en) * | 2021-03-30 | 2023-05-09 | 中铁第四勘察设计院集团有限公司 | Stratum partitioning method, device, equipment and storage medium for in-hole data fusion |
| CN113177919B (en) * | 2021-04-28 | 2022-08-05 | 成都艾立本科技有限公司 | Lithology classification and principal component element content detection method combining LIBS and deep learning |
| CN115327643B (en) * | 2021-05-11 | 2024-09-10 | 中国石油化工股份有限公司 | Machine learning training sample expansion and evaluation method for intelligent oil gas detection |
| CN115407424B (en) * | 2021-05-28 | 2024-12-20 | 中国石油化工股份有限公司 | An intelligent lithology identification method based on frequency-phase characteristics |
| CN113344050B (en) * | 2021-05-28 | 2024-03-26 | 中国石油天然气股份有限公司 | Lithology intelligent recognition method and system based on deep learning |
| CN113642698B (en) * | 2021-06-15 | 2024-04-02 | 中国科学技术大学 | Geophysical well logging intelligent interpretation methods, systems and storage media |
| CN113392924B (en) * | 2021-06-29 | 2023-05-02 | 中海油田服务股份有限公司 | Identification method of acoustic-electric imaging log and related equipment |
| CN114200524A (en) * | 2021-10-29 | 2022-03-18 | 五季数据科技(北京)有限公司 | A logging density curve correction method based on artificial intelligence deep learning |
| CN114114414A (en) * | 2021-11-18 | 2022-03-01 | 电子科技大学长三角研究院(湖州) | Artificial intelligence prediction method for 'dessert' information of shale reservoir |
| CN114444393A (en) * | 2022-01-26 | 2022-05-06 | 北京科技大学 | Construction method and device of well logging curve based on temporal convolutional neural network |
| CN114896468B (en) * | 2022-04-24 | 2024-02-02 | 北京月新时代科技股份有限公司 | File type matching method and data intelligent input method based on neural network |
| CN114863187B (en) * | 2022-06-13 | 2024-06-18 | 中国石油天然气集团有限公司 | Rock debris type identification method and device, electronic equipment and storage medium |
| CN115267935B (en) * | 2022-07-01 | 2024-08-20 | 中国石油大学(北京) | Logging reservoir evaluation method and device driven by combination of data and physical model |
| CN115659245A (en) * | 2022-10-24 | 2023-01-31 | 东华理工大学 | Sandstone-type uranium deposit rock stratum type identification method and device based on machine learning |
| CN115393656B (en) * | 2022-10-26 | 2023-01-24 | 中石化经纬有限公司 | An automatic classification method for stratigraphic classification of logging-while-drilling images |
| CN117093931A (en) * | 2023-07-26 | 2023-11-21 | 北京科技大学 | Automatic classification method and device for well test curves based on convolutional neural network |
| CN117235628B (en) * | 2023-11-10 | 2024-01-26 | 天津花栗鼠软件科技有限公司 | Well logging curve prediction method and system based on hybrid Bayesian deep network |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5251286A (en) * | 1992-03-16 | 1993-10-05 | Texaco, Inc. | Method for estimating formation permeability from wireline logs using neural networks |
| CN109389128B (en) * | 2018-08-24 | 2021-08-27 | 中国石油天然气股份有限公司 | Automatic extraction method and device for electric imaging logging image characteristics |
| CN109614883A (en) * | 2018-11-21 | 2019-04-12 | 瑾逸科技发展扬州有限公司 | A kind of tight sand crack intelligent identification Method based on convolutional neural networks |
| CN109670539A (en) * | 2018-12-03 | 2019-04-23 | 中国石油化工股份有限公司 | A kind of silt particle layer detection method based on log deep learning |
| CN111178441A (en) * | 2019-12-31 | 2020-05-19 | 中国矿业大学(北京) | Lithology identification method based on principal component analysis and full-connection neural network |
-
2020
- 2020-05-26 CN CN202010456323.7A patent/CN111783825B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 卷积神经网络在岩性识别中的应用;陈钢花等;测井技术;20190430;第43卷(第2期);第130-133页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111783825A (en) | 2020-10-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111783825B (en) | Logging lithology recognition method based on convolutional neural network learning | |
| CN112989708B (en) | A logging lithology identification method and system based on LSTM neural network | |
| US11599790B2 (en) | Deep learning based reservoir modeling | |
| CN107356958B (en) | A kind of fluvial depositional reservoir substep seismic facies prediction technique based on geological information constraint | |
| CN110674841B (en) | Logging curve identification method based on clustering algorithm | |
| CN109799533A (en) | A kind of method for predicting reservoir based on bidirectional circulating neural network | |
| Ali et al. | Classification of reservoir quality using unsupervised machine learning and cluster analysis: Example from Kadanwari gas field, SE Pakistan | |
| CN110619353B (en) | Multi-scale logging curve automatic identification method based on deep learning | |
| US6477469B2 (en) | Coarse-to-fine self-organizing map for automatic electrofacies ordering | |
| Lu et al. | Lithology identification using graph neural network in continental shale oil reservoirs: A case study in Mahu Sag, Junggar Basin, Western China | |
| CN111596978A (en) | Web page display method, module and system for lithofacies classification by artificial intelligence | |
| US11719851B2 (en) | Method and system for predicting formation top depths | |
| CN113610945A (en) | Ground stress curve prediction method based on hybrid neural network | |
| CN111914478A (en) | A comprehensive geological borehole logging lithology identification method | |
| CN114723095A (en) | Missing well logging curve prediction method and device | |
| Brown et al. | Machine learning on Crays to optimize petrophysical workflows in oil and gas exploration | |
| Akinyokun et al. | Well log interpretation model for the determination of lithology and fluid | |
| CN111626377A (en) | Lithofacies identification method, device, equipment and storage medium | |
| CN118656705B (en) | Compact sandstone reservoir rock phase intelligent identification method and system based on MLP-MTS | |
| CN114239937A (en) | Reservoir oil-gas-containing property prediction method and device, computer equipment and storage medium | |
| US20230141334A1 (en) | Systems and methods of modeling geological facies for well development | |
| DENG et al. | A Real‐time Lithological Identification Method based on SMOTE‐Tomek and ICSA Optimization | |
| CN117407841B (en) | Shale layer seam prediction method based on optimization integration algorithm | |
| Wang et al. | Auto recognition of carbonate microfacies based on an improved back propagation neural network | |
| CN111894551A (en) | Oil-gas reservoir prediction method based on LSTM |
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 |