CN110618082A - Reservoir micro-pore structure evaluation method and device based on neural network - Google Patents
Reservoir micro-pore structure evaluation method and device based on neural network Download PDFInfo
- Publication number
- CN110618082A CN110618082A CN201911037944.5A CN201911037944A CN110618082A CN 110618082 A CN110618082 A CN 110618082A CN 201911037944 A CN201911037944 A CN 201911037944A CN 110618082 A CN110618082 A CN 110618082A
- Authority
- CN
- China
- Prior art keywords
- pore structure
- data
- core
- neural network
- logging
- 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.)
- Granted
Links
- 239000011148 porous material Substances 0.000 title claims abstract description 196
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 27
- 238000011156 evaluation Methods 0.000 title claims abstract description 26
- 238000000547 structure data Methods 0.000 claims abstract description 89
- 238000000034 method Methods 0.000 claims abstract description 57
- 238000003062 neural network model Methods 0.000 claims abstract description 55
- 238000012549 training Methods 0.000 claims abstract description 30
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000002474 experimental method Methods 0.000 claims abstract description 9
- QSHDDOUJBYECFT-UHFFFAOYSA-N mercury Chemical compound [Hg] QSHDDOUJBYECFT-UHFFFAOYSA-N 0.000 claims abstract description 9
- 229910052753 mercury Inorganic materials 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 75
- 238000012545 processing Methods 0.000 claims description 31
- 239000010410 layer Substances 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 25
- 238000005457 optimization Methods 0.000 claims description 21
- 238000010606 normalization Methods 0.000 claims description 20
- 230000035699 permeability Effects 0.000 claims description 17
- 238000003064 k means clustering Methods 0.000 claims description 16
- 230000003595 spectral effect Effects 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 230000009467 reduction Effects 0.000 claims description 11
- 230000004913 activation Effects 0.000 claims description 10
- 238000000556 factor analysis Methods 0.000 claims description 7
- 230000035945 sensitivity Effects 0.000 claims description 6
- 238000010219 correlation analysis Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 5
- 238000011478 gradient descent method Methods 0.000 claims description 4
- 238000012804 iterative process Methods 0.000 claims description 4
- 239000002356 single layer Substances 0.000 claims description 4
- 238000003860 storage Methods 0.000 description 16
- 238000010586 diagram Methods 0.000 description 10
- 238000013461 design Methods 0.000 description 8
- 239000011435 rock Substances 0.000 description 7
- 238000011176 pooling Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000000418 atomic force spectrum Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007417 hierarchical cluster analysis Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/08—Investigating permeability, pore-volume, or surface area of porous materials
- G01N15/088—Investigating volume, surface area, size or distribution of pores; Porosimetry
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Dispersion Chemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Analysing Materials By The Use Of Radiation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a reservoir micro-pore structure evaluation method and device based on a neural network, wherein the method comprises the following steps: obtaining core pore structure data obtained based on a core high-pressure mercury intrusion experiment, and the category of a core pore structure corresponding to the core pore structure data; aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and continuously obtaining the logging data after preprocessing; taking the preprocessed logging data as input and the category of the core pore structure as output, and training the initial neural network model to obtain a trained neural network model; when the reservoir microstructure of the target well needs to be evaluated, target logging data are input into the trained neural network model, the pore category corresponding to the target logging data is obtained, and the reservoir microstructure can be evaluated accurately and quickly.
Description
Technical Field
The embodiment of the invention relates to the technical field of oil and gas field development, in particular to a reservoir micro-pore structure evaluation method and device based on a neural network.
Background
The oil gas is stored and flows in the pores of the rock, so the shape, size, development degree and communication condition of the pores of the rock, namely the pore structure of the rock can directly influence the storage amount and the productivity of the oil gas, particularly the micro pore structure of a reservoir is the basis for researching the physical property and the storage property of the rock, and has extremely important significance for the exploration and development of oil gas resources.
At present, the methods for researching the microstructure of a reservoir are mainly divided into two types, the first type is a core analysis method, and the methods mainly comprise a capillary force curve method, a cast body slice method, a scanning electron microscope method and the like. The second category is field evaluation of well log data, such as the use of resistivity log data or nmr logging to study the pore structure of rock.
However, the core analysis method is generally manual identification and has slow analysis speed; the field evaluation method of logging information mainly reflects the macroscopic pore structure of rock and has low identification precision. Therefore, a method for evaluating the micro-pore structure of the reservoir rapidly and accurately is needed.
Disclosure of Invention
The embodiment of the invention provides a reservoir stratum micro-pore structure evaluation method and device based on a neural network, which can be used for evaluating the reservoir stratum micro-pore structure quickly and accurately.
In a first aspect, an embodiment of the present invention provides a reservoir micro-pore structure evaluation method based on a neural network, including:
obtaining core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially carrying out optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of a core pore structure corresponding to the core pore structure data according to a clustering processing result;
aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and performing optimization and normalization pretreatment on the logging data to obtain pretreated logging data;
taking the preprocessed logging data as input and the category of the core pore structure as output, and training the initial neural network model to obtain a trained neural network model;
and inputting the target logging data into the trained neural network model to obtain the pore category corresponding to the target logging data.
In a possible design, the optimizing the core pore structure data specifically includes:
taking the core permeability as a reference series of grey scale correlation analysis, taking each core pore structure parameter as a comparison series, and carrying out non-dimensionalization on the core permeability and the core pore structure parameters to obtain a parameter matrix of the non-dimensionalized core permeability and the core pore structure parameters;
and calculating the gray correlation degrees between each comparison array and the reference array according to the parameter matrix, sorting the gray correlation degrees, selecting the core pore structure parameters with the gray correlation degrees ranked at the top in the preset number, and completing the optimization of the core pore structure data.
In one possible design, the dimension reduction on the core pore structure data specifically includes:
and forming the optimized core pore structure data into a matrix form as follows:
wherein N is the number of samples, M is the number of sample characteristics, i.e. the number of pore structure parameters, xijRepresenting the jth characteristic data corresponding to the ith element;
normalizing the characteristic data, and calculating a covariance matrix of the characteristic data, and an eigenvalue and an eigenvector of the covariance matrix;
and selecting a matrix formed by eigenvectors corresponding to the first K eigenvalues of the covariance matrix as a main component of the core pore structure data, wherein K is less than M.
In one possible design, the clustering process is performed on the core pore structure data, and specifically includes:
carrying out hierarchical clustering on the dimension-reduced core pore structure data to obtain a hierarchical clustering result;
performing spectral clustering on the dimensionality-reduced core pore structure data to obtain a spectral clustering result;
and carrying out K-means clustering on the dimension-reduced core pore structure data to obtain a K-means clustering result.
In a possible design, the obtaining of the category of the core pore structure corresponding to the core pore structure data according to the clustering result specifically includes:
establishing a coordinate system according to main components of the core pore structure data, putting the core pore structure data into the coordinate system to establish a three-dimensional intersection graph, and classifying the core pore structure data according to the three-dimensional intersection graph, wherein the classification basis is a hierarchical clustering result, a spectral clustering result and a K-means clustering result.
In one possible design, the pre-processing of the well log data preferably includes:
optimizing the logging data by adopting a factor analysis method, and extracting a logging curve with strong sensitivity to the micro-pore structure of the reservoir as a main factor;
carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the main factors of which the contribution rates are larger than a preset threshold value, rotating the main factor matrix to obtain a rotating component matrix, and optimizing the logging curve according to the rotating component matrix.
In one possible design, the normalization preprocessing is performed on the well logging data, and specifically includes:
if the well logging data after the pre-processing is preferably linear characteristic data, a formula is adoptedCarrying out normalization pretreatment; wherein X is the logging data after normalization, X is the logging data, X isminIs the minimum value, X, of the well log datamaxIs the maximum value of the logging data;
if the well logging data after the pre-processing is preferably linear characteristic data, a formula is adopted Carrying out normalization pretreatment; wherein X is the well log data after normalization, lgX is the well log data, lgXminIs the log minimum of the log data, lgXmaxIs the logarithm of the maximum value of the logging data.
In one possible design, the initial neural network model is a fully-connected neural network, wherein the fully-connected neural network adopts an input layer, three hidden layers and an output layer, a softmax layer is added behind the output layer to convert score values output by the network into probability distribution of each pore structure, weight initialization adopts gaussian initialization, bias is initialized by 0.1, an activation function adopts a relu activation function, a loss function adopts a cross entropy loss function, and a learning rate is set to be 0.01;
the method comprises the following steps of taking the preprocessed logging data as input and the category of a core pore structure as output, training an initial neural network model, and obtaining a trained neural network model, wherein the method specifically comprises the following steps:
and (3) performing batch training by adopting a gradient descent method, forming one batch by every 4 samples, finishing one-time training by all the batches into an iterative process, continuously reducing the loss function in the training process until a set threshold value is reached or the iterative times reach the set times, and storing the weight and the bias parameters after the training is finished to obtain the trained neural network model.
In one possible design, the initial neural network model is a convolutional neural network model, where the convolutional neural network model includes a single layer convolutional neural network model or a double layer convolutional network model.
In a second aspect, an embodiment of the present invention provides a reservoir micro-pore structure evaluation apparatus based on a neural network, including:
the core pore structure data processing module is used for acquiring core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially carrying out optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of a core pore structure corresponding to the core pore structure data according to a clustering processing result;
the logging data processing module is used for aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and performing optimization and normalization pretreatment on the logging data to obtain pretreated logging data;
the neural network model training module is used for taking the preprocessed logging data as input and the category of the core pore structure as output, training the initial neural network model and obtaining the trained neural network model;
and the pore evaluation module is used for inputting the target logging data into the trained neural network model to obtain the pore category corresponding to the target logging data.
According to the reservoir micro pore structure evaluation method and device based on the neural network, provided by the embodiment of the invention, firstly, core pore structure data obtained based on a core high-pressure mercury intrusion experiment are obtained, and the type of a core pore structure corresponding to the core pore structure data is obtained; aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and continuously obtaining the logging data after preprocessing; finally, the preprocessed logging data is used as input, the category of the core pore structure is used as output, and the initial neural network model is trained to obtain a trained neural network model; when the reservoir microstructure of the target well needs to be evaluated, target logging data are input into the trained neural network model, the pore category corresponding to the target logging data is obtained, and the reservoir microstructure can be evaluated accurately and quickly.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is an architecture diagram of a system for evaluating a micro-pore structure of a reservoir based on a neural network according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a reservoir micro-pore structure evaluation method based on a neural network according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a fully-connected neural network model architecture according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a convolutional neural network model architecture according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a reservoir micro-pore structure evaluation device based on a neural network according to an embodiment of the present invention;
fig. 6 is a schematic hardware structure diagram of a reservoir micro-pore structure evaluation device based on a neural network according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. 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.
Referring to fig. 1, fig. 1 is an architectural diagram of a system for evaluating a micro-pore structure of a reservoir based on a neural network according to an embodiment of the present invention. As shown in fig. 1, the system provided by the present embodiment includes a terminal 101 and a server 102. The terminal 101 may be a personal computer, a mobile phone, a tablet, or the like. The embodiment does not particularly limit the implementation manner of the terminal 101 as long as the terminal 101 can interact with the user. The server 102 may be one or a cluster of several servers.
Referring to fig. 2, fig. 2 is a schematic flow chart of a reservoir micro-pore structure evaluation method based on a neural network according to an embodiment of the present invention. The execution subject of this embodiment may be the terminal in the embodiment shown in fig. 1, or may also be the server in the embodiment shown in fig. 1, and this embodiment is not limited herein. As shown in fig. 2, the method includes:
s201: obtaining core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially carrying out optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of the core pore structure corresponding to the core pore structure data according to a clustering processing result.
The core pore structure data includes core permeability and core pore structure parameters.
In this embodiment, the rock permeability and the obtained core pore structure data may be subjected to gray level correlation analysis, the core pore structure data may be sorted according to the magnitude of the correlation, and n core pore structure parameters with the top correlation may be selected to obtain the core pore structure data after optimization. The initial data volume is reduced, the operation process can be reduced, and the evaluation efficiency is improved.
In this embodiment, principal component analysis is performed on the optimized pore structure data, feature dimension reduction is performed, and the collinear problem is solved.
In this embodiment, the clustering process may include hierarchical clustering analysis, spectral clustering analysis, and K-means clustering analysis.
S202: aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and performing optimization and normalization pretreatment on the logging data to obtain the pretreated logging data.
In the embodiment, a factor analysis method is adopted to optimize logging data, and a logging curve with strong sensitivity to a micro-pore structure of a reservoir is extracted as a main factor;
carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the main factors of which the contribution rates are larger than a preset threshold value, rotating the main factor matrix to obtain a rotating component matrix, and optimizing the logging curve according to the rotating component matrix.
S203: and training the initial neural network model by taking the preprocessed logging data as input and the category of the core pore structure as output to obtain the trained neural network model.
In an embodiment of the present invention, the initial neural network model is a fully-connected neural network model, referring to fig. 3, and fig. 3 is a schematic diagram of a fully-connected neural network model architecture provided in an embodiment of the present invention. The fully-connected neural network model adopts an input layer, three hidden layers and an output layer, a softmax layer is added behind the output layer, score values output by the network are converted into probability distribution of each pore structure, weight initialization adopts Gaussian initialization, bias is initialized by 0.1, an activation function adopts a relu activation function, a loss function adopts a cross entropy loss function, and the learning rate is set to be 0.01;
in step S203, training the initial neural network model by using the preprocessed well logging data as input and the category of the core pore structure as output, to obtain a trained neural network model, specifically including:
and (3) performing batch training by adopting a gradient descent method, forming one batch by every 4 samples, finishing one-time training by all the batches into an iterative process, continuously reducing the loss function in the training process until a set threshold value is reached or the iterative times reach the set times, and storing the weight and the bias parameters after the training is finished to obtain the trained neural network model.
In another embodiment of the present invention, the initial neural network model is a convolutional neural network model, wherein the convolutional neural network model comprises a single layer convolutional neural network model or a double layer convolutional network model.
Referring to fig. 4, fig. 4 is a schematic diagram of a convolutional neural network model architecture according to an embodiment of the present invention.
The single-layer convolution neural network model simulates logging data into a 3 x 3 pixel matrix as the input of the network, 4 convolution kernels are adopted, the size is 2 x 1, and the step length is 1; the pooling layer adopts 2 multiplied by 2 maximum pooling, and the step length is 1; two full-connection layers are adopted, and the number of the neurons is 36 and 5 respectively. And adding a softmax layer behind an output layer of the network, converting the score value output by the network into probability distribution of each pore structure, initializing the weight by adopting Gaussian initialization, and initializing the bias by using 0.1. The activation function adopts a relu activation function, and the loss function adopts a cross entropy loss function.
The double-layer convolution network model simulates well logging data into a 9 x 9 pixel matrix as the input of the network, the first layer of convolution adopts 4 convolution kernels, the size of the convolution kernels is 3 x 1, and the step length is 1; the pooling layer adopts 2 multiplied by 2 maximum pooling, and the step length is 1; the second layer convolution adopts 8 convolution kernels, the size is 3 multiplied by 4, the step length is 1, and finally two full-connection layers are adopted, and the number of the neurons is 36 and 5 respectively. And adding a softmax layer behind an output layer of the network, converting the score value output by the network into probability distribution of each pore structure, initializing the weight by adopting Gaussian initialization, and initializing the bias by using 0.1. The activation function adopts a relu activation function, and the loss function adopts a cross entropy loss function.
Step S203, training the initial neural network model by using the preprocessed well logging data as input and the category of the core pore structure as output, to obtain a trained neural network model, which specifically includes:
and (3) performing batch training by adopting a gradient descent method, forming one batch by every 4 samples, finishing one-time training by all the batches into an iterative process, continuously reducing the loss function in the training process until a set threshold value is reached or the iterative times reach the set times, and storing the weight and the bias parameters after the training is finished to obtain the trained neural network model.
S204: and inputting the target logging data into the trained neural network model to obtain the pore category corresponding to the target logging data.
And inputting the target logging data into a trained neural network model according to the target logging data, and directly outputting the pore category corresponding to the target logging data.
As can be seen from the above embodiment, first, core pore structure data obtained based on a core high-pressure mercury intrusion experiment is obtained, and the category of the core pore structure corresponding to the core pore structure data is obtained; aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and continuously obtaining the logging data after preprocessing; finally, the preprocessed logging data is used as input, the category of the core pore structure is used as output, and the initial neural network model is trained to obtain a trained neural network model; when the reservoir microstructure of the target well needs to be evaluated, target logging data are input into the trained neural network model, the pore category corresponding to the target logging data is obtained, and the reservoir microstructure can be evaluated accurately and quickly.
In one embodiment of the present invention,
the optimization of the core pore structure data specifically comprises the following steps:
s2011: taking the core permeability as a reference series of grey scale correlation analysis, taking each core pore structure parameter as a comparison series, and carrying out non-dimensionalization on the core permeability and the core pore structure parameters to obtain a parameter matrix of the non-dimensionalized core permeability and the core pore structure parameters; and calculating the gray correlation degrees between each comparison array and the reference array according to the parameter matrix, sorting the gray correlation degrees, selecting the core pore structure parameters with the gray correlation degrees ranked at the top in the preset number, and completing the optimization of the core pore structure data.
First, the core pore structure data is formed into the following matrix:
wherein n is the number of pore structure parameters, and m is the number of samples.
Further, taking the core permeability as a reference series for grey correlation analysis, taking each pore structure parameter as a comparison series, and carrying out non-dimensionalization processing on the core permeability and the pore structure data to solve the influence caused by different data dimensions, wherein the adopted formula is as follows:
obtaining a nondimensionalized core permeability array and a pore structure parameter matrix;
wherein, X0For reference sequence, (X)1,…,Xn) Compare the series.
Further, solving a grey correlation coefficient between the comparison number series and the reference number series, wherein the formula is as follows:
wherein, Delta (min) is the minimum value of the difference between the comparison sequence and the reference sequence, Delta (max) is the maximum value of the difference between the comparison sequence and the reference sequence, and Delta (min) is the minimum value of two stagesoi(k) For comparison of the difference between each point in the series and each point in the reference series, ρ is a resolution factor, typically between 0 and 1, and is usually 0.5.
Further, the gray correlation degree of the comparison number series and the reference number series is calculated, and the formula is as follows:
and sorting according to the obtained correlation degree from large to small, selecting the first pore structure parameters with larger correlation degree, and finishing parameter optimization.
In one embodiment of the present invention,
the dimension reduction of the core pore structure data specifically comprises the following steps:
s2012: and forming the optimized core pore structure data into a matrix form as follows:
wherein N is the number of samples, M is the number of sample characteristics, i.e. the number of pore structure parameters, xijRepresenting the jth characteristic data corresponding to the ith element;
normalizing the characteristic data, and calculating a covariance matrix of the characteristic data, and an eigenvalue and an eigenvector of the covariance matrix;
and selecting a matrix formed by eigenvectors corresponding to the first K (K < M) eigenvalues of the covariance matrix as a main component of the core pore structure data.
Wherein, the characteristic data is normalized by the following formula:
in the above two formulas, the first and second groups,is the average value of each row of elements, X'N×MIs XN×MZero mean normalized results of (a). The above two-equation process is to subtract the average value of each column element from each element in each column, and is an essential step for solving the covariance matrix of X.
Further, a covariance matrix is calculated, and the formula adopted is as follows:
further, the eigenvalue and eigenvector of the covariance matrix are solved by the following formula:
C＝ζλζ-1 (10)
the column of the ζ matrix is an eigenvector of the covariance matrix C, which is an orthogonal matrix, and λ is a diagonal matrix of eigenvalues of the matrix, elements on a diagonal line of the matrix are arranged in descending order, the first k (k < M) eigenvalues of the diagonal matrix with the largest magnitude are taken, and a matrix formed by eigenvectors corresponding to the k eigenvalues is the principal component to be transformed. Each component of the transformed sample vector is irrelevant, and the problem of collinearity among pore structure parameters is solved.
In one embodiment of the present invention,
the clustering processing is carried out on the core pore structure data, and the method specifically comprises the following steps:
s2013: and performing hierarchical clustering on the dimension-reduced core pore structure data to obtain a hierarchical clustering result.
And clustering the pore structure data subjected to dimensionality reduction by using a hierarchical clustering method by taking the data subjected to dimensionality reduction as input data of hierarchical clustering. The clustering method can divide the data sets on different levels, so that a tree-shaped clustering structure is formed. In the implementation, a "bottom-up" aggregation strategy is used, in which each sample is first treated as a cluster, and then the clusters are combined into larger and larger clusters until all samples are in a cluster. The class distance is taken as the class distance, i.e., the average distance of each object in the class from all objects in other classes is taken as the class distance. The method can make the sample data of any shape converge to the global optimum and is insensitive to noise data; besides, another great advantage of this clustering method is that the number of clusters may not be specified.
S2014: and performing spectral clustering on the dimension-reduced core pore structure data to obtain a spectral clustering result.
The spectral clustering method is a graph theory-based clustering method, and the purpose of clustering sample data is achieved by clustering Laplacian eigenvectors of the sample data. The specific process is as follows:
calculating a similarity matrix of the sample data, wherein the adopted formula is as follows:
wherein, WijIs the element in the ith row and j column of the similarity matrix.
The calculation degree matrix D adopts the formula as follows:
wherein D is represented byiA diagonal matrix is formed.
Calculating a Laplace matrix L, and standardizing the L by adopting a formula as follows:
L＝D-W (13)
wherein L issymIs a normalized laplacian matrix.
Calculating LsymFirst k minimumAnd (3) forming a n-x-k dimensional matrix F by using the k eigenvectors of the eigenvalue, then taking each row in the matrix F as a k dimensional sample, counting n samples in total, and clustering by using a traditional method to obtain a clustering result. When the number of the clustered categories is small, the clustering effect is good, and the method can effectively process sparse data, but has low anti-interference capability.
S2015: and carrying out K-means clustering on the dimension-reduced core pore structure data to obtain a K-means clustering result.
And clustering the processed pore structure data by adopting K-means clustering, wherein the target function is as follows:
wherein the content of the first and second substances,is a cluster CiThe objective of K-means clustering is to decrease the above equation until the cluster center of each class does not change with the increase of the number of iterations, and the above equation reaches the minimum value. The K-means clustering has the advantage of higher computational efficiency, but is only applicable to convex or spherical data and is easy to converge on a locally optimal solution.
In one embodiment of the present invention,
the obtaining of the category of the core pore structure corresponding to the core pore structure data according to the clustering result specifically includes:
s2016: establishing a coordinate system according to main components of the core pore structure data, putting the core pore structure data into the coordinate system to establish a three-dimensional intersection graph, and classifying the core pore structure data according to the three-dimensional intersection graph, wherein the classification basis is a hierarchical clustering result, a spectral clustering result and a K-means clustering result.
In one embodiment of the present invention,
performing optimal preprocessing on the logging data, which specifically comprises the following steps:
optimizing the logging data by adopting a factor analysis method, and extracting a logging curve with strong sensitivity to the micro-pore structure of the reservoir as a main factor;
carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the main factors of which the contribution rates are larger than a preset threshold value, rotating the main factor matrix to obtain a rotating component matrix, and optimizing the logging curve according to the rotating component matrix.
Rotating the main factor matrix to obtain a rotation component matrix, wherein the numerical value on the rotation component matrix is the correlation coefficient between the logging curve and the extracted main factors, and optimizing the logging curve according to the correlation coefficient between the logging curve and each main factor
The method comprises the following steps of (1) optimizing logging data by adopting a factor analysis method, and extracting a logging curve with strong sensitivity to a micro pore structure of a reservoir, wherein the specific process comprises the following steps:
the mathematical representation of the factorial analysis is a matrix: x ═ AF + B, i.e.:
wherein, the vector X (X)1,x2,…,xp) Is an observable random variable, i.e., an original observed variable; f (F)1,f2,…,fk) Is X (X)1,x2,…,xp) The common factors of (1), i.e. the factors which appear together in the expression of each original observed variable, are independent non-observable theoretical variables; beta (. beta.)1,β2,…,βp) Is a special factor; alpha is alphaijIs a factor load, is the load of the ith original variable on the jth factor, and can be expressed as alphaijConsidering the weight of the ith variable on the jth common factor; alpha is alphaijA factor load matrix is constructed.
Carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the first main factors with larger contribution rates, carrying out factor rotation on a component matrix obtained after factor analysis to obtain a rotation component matrix, and carrying out optimization on the logging curve according to the rotation component matrix.
Further, the optimized logging data is normalized, and the formula is as follows:
wherein equation (17) applies to the linear characteristic data; equation (18) is applied to nonlinear characteristic data such as resistivity.
Referring to fig. 5, fig. 5 is a schematic structural diagram of a reservoir micro-pore structure evaluation device based on a neural network according to an embodiment of the present invention. As shown in fig. 5, the neural network-based reservoir micro-pore structure evaluation apparatus 50 includes: the system comprises a core pore structure data processing module 501, a logging data processing module 502, a neural network model training module 503 and a pore evaluation module 504.
The core pore structure data processing module 501 is used for acquiring core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially performing optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of a core pore structure corresponding to the core pore structure data according to a clustering processing result;
the logging data processing module 502 is configured to align the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and perform optimization and normalization preprocessing on the logging data to obtain preprocessed logging data;
a neural network model training module 503, configured to train an initial neural network model by using the preprocessed logging data as input and using the category of the core pore structure as output, so as to obtain a trained neural network model;
and the pore evaluation module 504 is configured to input the target logging data into the trained neural network model to obtain a pore category corresponding to the target logging data.
The apparatus provided in this embodiment may be used to implement the technical solutions of the above method embodiments, and the implementation principles and technical effects are similar, which are not described herein again.
In an embodiment of the present invention, the core pore structure data processing module 501 is specifically configured to use the core permeability as a reference sequence for grey scale associated analysis, use each core pore structure parameter as a comparison sequence, and perform non-dimensionalization processing on the core permeability and the core pore structure parameter to obtain a parameter matrix of the non-dimensionalized core permeability and core pore structure parameter; and calculating the gray correlation degrees between each comparison array and the reference array according to the parameter matrix, sorting the gray correlation degrees, selecting the core pore structure parameters with the gray correlation degrees ranked at the top in the preset number, and completing the optimization of the core pore structure data.
In an embodiment of the present invention, the core pore structure data processing module 501 is specifically configured to form the preferred core pore structure data into a matrix form, where the matrix form is as follows:
wherein N is the number of samples, M is the number of sample characteristics, i.e. the number of pore structure parameters, xijRepresenting the jth characteristic data corresponding to the ith element;
normalizing the characteristic data, and calculating a covariance matrix of the characteristic data, and an eigenvalue and an eigenvector of the covariance matrix;
and selecting a matrix formed by eigenvectors corresponding to the first K eigenvalues of the covariance matrix as a main component of the core pore structure data, wherein K is less than M.
In an embodiment of the present invention, the core pore structure data processing module 501 is specifically configured to perform hierarchical clustering on the dimensionality-reduced core pore structure data to obtain a hierarchical clustering result; performing spectral clustering on the dimensionality-reduced core pore structure data to obtain a spectral clustering result; and carrying out K-means clustering on the dimension-reduced core pore structure data to obtain a K-means clustering result.
In an embodiment of the present invention, the core pore structure data processing module 501 is specifically configured to establish a coordinate system according to a principal component of the core pore structure data, put the core pore structure data into the coordinate system to establish a three-dimensional cross plot, and classify the core pore structure data according to the three-dimensional cross plot, where the classification basis is a hierarchical clustering result, a spectral clustering result, and a K-means clustering result.
In an embodiment of the present invention, the log data processing module 502 is specifically configured to perform optimization on log data by using a factor analysis method, and extract a log curve with strong sensitivity to a micro-pore structure of a reservoir as a main factor;
carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the main factors of which the contribution rates are larger than a preset threshold value, rotating the main factor matrix to obtain a rotating component matrix, and optimizing the logging curve according to the rotating component matrix.
In an embodiment of the present invention, the log data processing module 502 is specifically configured to adopt a formula if the pre-processed log data is linear characteristic dataCarrying out normalization pretreatment; wherein X is the well logging data after normalization, X is the well logging data, X isminIs the minimum value, X, of the well log datamaxIs the maximum value of the logging data;
if the well logging data after the pre-processing is preferably linear characteristic data, a formula is adopted Carrying out normalization pretreatment; wherein X is the well log data after normalization, lgX is the well log data, lgXminIs the measurementMinimum log of well data, lgXmaxIs the logarithm of the maximum value of the logging data.
The apparatus provided in this embodiment may be used to implement the technical solutions of the above method embodiments, and the implementation principles and technical effects are similar, which are not described herein again.
Referring to fig. 6, fig. 6 is a schematic hardware structure diagram of a neural network-based reservoir micro-pore structure evaluation apparatus according to an embodiment of the present invention. As shown in fig. 6, the neural network-based reservoir micro-pore structure evaluation apparatus 60 of the present embodiment includes: a processor 601 and a memory 602; wherein
A memory 602 for storing computer-executable instructions;
the processor 601 is configured to execute the computer execution instructions stored in the memory to implement the steps performed by the terminal or the server in the above embodiments. Reference may be made in particular to the description relating to the method embodiments described above.
Alternatively, the memory 602 may be separate or integrated with the processor 601.
When the memory 602 is provided independently, the neural network-based reservoir micro-pore structure evaluation device further includes a bus 603 for connecting the memory 602 and the processor 601.
The embodiment of the invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when a processor executes the computer-executable instructions, the method for evaluating the micro-pore structure of the reservoir based on the neural network is realized.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described device embodiments are merely illustrative, and for example, the division of the modules is only one logical division, and other divisions may be realized in practice, for example, a plurality of modules may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or modules, and may be in an electrical, mechanical or other form.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to implement the solution of the present embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each module may exist alone physically, or two or more modules are integrated into one unit. The unit formed by the modules can be realized in a hardware form, and can also be realized in a form of hardware and a software functional unit.
The integrated module implemented in the form of a software functional module may be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) or a processor to execute some steps of the methods described in the embodiments of the present application.
It should be understood that the Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.
The memory may comprise a high-speed RAM memory, and may further comprise a non-volatile storage NVM, such as at least one disk memory, and may also be a usb disk, a removable hard disk, a read-only memory, a magnetic or optical disk, etc.
The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.
The storage medium may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. Of course, the storage medium may also be integral to the processor. The processor and the storage medium may reside in an Application Specific Integrated Circuits (ASIC). Of course, the processor and the storage medium may reside as discrete components in an electronic device or host device.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A reservoir micro-pore structure evaluation method based on a neural network is characterized by comprising the following steps:
obtaining core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially carrying out optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of a core pore structure corresponding to the core pore structure data according to a clustering processing result;
aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and performing optimization and normalization pretreatment on the logging data to obtain pretreated logging data;
taking the preprocessed logging data as input and the category of the core pore structure as output, and training the initial neural network model to obtain a trained neural network model;
and inputting the target logging data into the trained neural network model to obtain the pore category corresponding to the target logging data.
2. The method according to claim 1, wherein the optimizing the core pore structure data specifically comprises:
taking the core permeability as a reference series of grey scale correlation analysis, taking each core pore structure parameter as a comparison series, and carrying out non-dimensionalization on the core permeability and the core pore structure parameters to obtain a parameter matrix of the non-dimensionalized core permeability and the core pore structure parameters;
and calculating the gray correlation degrees between each comparison array and the reference array according to the parameter matrix, sorting the gray correlation degrees, selecting the core pore structure parameters with the gray correlation degrees ranked at the top in the preset number, and completing the optimization of the core pore structure data.
3. The method according to claim 2, wherein the reducing the dimension of the core pore structure data specifically comprises:
and forming the optimized core pore structure data into a matrix form as follows:
wherein N is the number of samples, M is the number of sample characteristics, i.e. the number of pore structure parameters, xijRepresenting the jth characteristic data corresponding to the ith element;
normalizing the characteristic data, and calculating a covariance matrix of the characteristic data, and an eigenvalue and an eigenvector of the covariance matrix;
and selecting a matrix formed by eigenvectors corresponding to the first K eigenvalues of the covariance matrix as a main component of the core pore structure data, wherein K is less than M.
4. The method as claimed in claim 3, wherein the clustering process is performed on the core pore structure data, and specifically comprises:
carrying out hierarchical clustering on the dimension-reduced core pore structure data to obtain a hierarchical clustering result;
performing spectral clustering on the dimensionality-reduced core pore structure data to obtain a spectral clustering result;
and carrying out K-means clustering on the dimension-reduced core pore structure data to obtain a K-means clustering result.
5. The method according to claim 4, wherein the obtaining of the category of the core pore structure corresponding to the core pore structure data according to the clustering result specifically comprises:
establishing a coordinate system according to main components of the core pore structure data, putting the core pore structure data into the coordinate system to establish a three-dimensional intersection graph, and classifying the core pore structure data according to the three-dimensional intersection graph, wherein the classification basis is a hierarchical clustering result, a spectral clustering result and a K-means clustering result.
6. The method of claim 1, wherein the pre-processing of the well log data preferably comprises:
optimizing the logging data by adopting a factor analysis method, and extracting a logging curve with strong sensitivity to the micro-pore structure of the reservoir as a main factor;
carrying out variance analysis on the logging data to obtain variance contribution rates of all the main factors, extracting the main factors of which the contribution rates are larger than a preset threshold value, rotating the main factor matrix to obtain a rotating component matrix, and optimizing the logging curve according to the rotating component matrix.
7. The method of claim 6, wherein the performing a normalization pre-processing on the well log data comprises:
if the well logging data after the pre-processing is preferably linear characteristic data, a formula is adoptedCarrying out normalization pretreatment; wherein X is the well logging data after normalization, X is the well logging data, X isminIs the minimum value, X, of the well log datamaxIs the maximum value of the logging data;
if the well logging data after the pre-processing is preferably linear characteristic data, a formula is adopted Carrying out normalization pretreatment; where X is the log data after normalization, lgXFor said log data, lgXminIs the log minimum of the log data, lgXmaxIs the logarithm of the maximum value of the logging data.
8. The method according to any one of claims 1 to 7, wherein the initial neural network model is a fully-connected neural network, wherein the fully-connected neural network adopts an input layer, three hidden layers and an output layer, the output layer is added with a softmax layer, score values of network outputs are converted into probability distributions of pore structures, weight initialization adopts Gaussian initialization, bias is initialized with 0.1, an activation function adopts a relu activation function, a loss function adopts a cross entropy loss function, and a learning rate is set to be 0.01;
the method comprises the following steps of taking the preprocessed logging data as input and the category of a core pore structure as output, training an initial neural network model, and obtaining a trained neural network model, wherein the method specifically comprises the following steps:
and (3) performing batch training by adopting a gradient descent method, forming one batch by every 4 samples, finishing one-time training by all the batches into an iterative process, continuously reducing the loss function in the training process until a set threshold value is reached or the iterative times reach the set times, and storing the weight and the bias parameters after the training is finished to obtain the trained neural network model.
9. The method of any one of claims 1 to 7, wherein the initial neural network model is a convolutional neural network model, wherein the convolutional neural network model comprises a single layer convolutional neural network model or a double layer convolutional network model.
10. A neural network-based reservoir micro-pore structure evaluation device based on a neural network is characterized by comprising the following components:
the core pore structure data processing module is used for acquiring core pore structure data obtained based on a core high-pressure mercury intrusion experiment, sequentially carrying out optimization, dimensionality reduction and clustering processing on the core pore structure data, and obtaining the category of a core pore structure corresponding to the core pore structure data according to a clustering processing result;
the logging data processing module is used for aligning the core depth with the logging depth according to a density curve in the core density and the logging data to obtain logging data corresponding to the core, and performing optimization and normalization pretreatment on the logging data to obtain pretreated logging data;
the neural network model training module is used for taking the preprocessed logging data as input and the category of the core pore structure as output, training the initial neural network model and obtaining the trained neural network model;
and the pore evaluation module is used for inputting the target logging data into the trained neural network model to obtain the pore category corresponding to the target logging data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911037944.5A CN110618082B (en) | 2019-10-29 | 2019-10-29 | Reservoir micro-pore structure evaluation method and device based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911037944.5A CN110618082B (en) | 2019-10-29 | 2019-10-29 | Reservoir micro-pore structure evaluation method and device based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110618082A true CN110618082A (en) | 2019-12-27 |
| CN110618082B CN110618082B (en) | 2021-06-04 |
Family
ID=68927077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911037944.5A Active CN110618082B (en) | 2019-10-29 | 2019-10-29 | Reservoir micro-pore structure evaluation method and device based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110618082B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN111583148A (en) * | 2020-05-07 | 2020-08-25 | 苏州闪掣智能科技有限公司 | Rock core image reconstruction method based on generation countermeasure network |
| CN111950192A (en) * | 2020-07-15 | 2020-11-17 | 中海油田服务股份有限公司 | Method and device for modeling pore network model based on convolutional neural network |
| CN112862169A (en) * | 2021-01-28 | 2021-05-28 | 中国石油大学(北京) | Method and device for predicting content of free oil in continental facies shale |
| CN113239955A (en) * | 2021-04-07 | 2021-08-10 | 长江大学 | Carbonate reservoir rock classification method |
| CN115830873A (en) * | 2023-01-10 | 2023-03-21 | 西南交通大学 | Urban road traffic event classification method, device, equipment and readable storage medium |
| RU2812143C1 (en) * | 2020-07-15 | 2024-01-23 | Чайна Ойлфилд Сервисез Лимитед | Method and device for measuring characteristics of rock column to create model of pore system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844993A (en) * | 2017-02-09 | 2017-06-13 | 朱亚婷 | A kind of method of oil well classification and oil reservoir subregion based on SPSS |
| CN107576772A (en) * | 2017-07-25 | 2018-01-12 | 中国地质大学（北京） | A kind of method using log data quantitative assessment coal body structure type coal |
| CN109711429A (en) * | 2018-11-22 | 2019-05-03 | 中国石油天然气股份有限公司 | A kind of evaluating reservoir classification method and device |
| CN110020714A (en) * | 2018-01-10 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Model training and data analysing method, device, equipment and storage medium |
-
2019
- 2019-10-29 CN CN201911037944.5A patent/CN110618082B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844993A (en) * | 2017-02-09 | 2017-06-13 | 朱亚婷 | A kind of method of oil well classification and oil reservoir subregion based on SPSS |
| CN107576772A (en) * | 2017-07-25 | 2018-01-12 | 中国地质大学（北京） | A kind of method using log data quantitative assessment coal body structure type coal |
| CN110020714A (en) * | 2018-01-10 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Model training and data analysing method, device, equipment and storage medium |
| CN109711429A (en) * | 2018-11-22 | 2019-05-03 | 中国石油天然气股份有限公司 | A kind of evaluating reservoir classification method and device |
Non-Patent Citations (5)
| Title |
|---|
| 吕晓光等: "储层分类方法的应用及评价", 《大庆石油地质与开发》 * |
| 安鹏 等: "基于 LSTM 循环神经网络的孔隙度预测方法研究", 《中国地球科学联合学术年会2018》 * |
| 柴瑞泽: "基于神经网络的工业大数据分类模型与拟合模型研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 段友祥 等: "卷积神经网络在储层预测中的应用研究", 《通信学报》 * |
| 蔺景龙等: "基于BP神经网络的储层微孔隙结构类型预测", 《测井技术》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN111583148A (en) * | 2020-05-07 | 2020-08-25 | 苏州闪掣智能科技有限公司 | Rock core image reconstruction method based on generation countermeasure network |
| CN111950192A (en) * | 2020-07-15 | 2020-11-17 | 中海油田服务股份有限公司 | Method and device for modeling pore network model based on convolutional neural network |
| WO2022011894A1 (en) * | 2020-07-15 | 2022-01-20 | 中海油田服务股份有限公司 | Convolutional neural network-based modeling method and device for pore network model |
| CN111950192B (en) * | 2020-07-15 | 2023-10-24 | 中海油田服务股份有限公司 | Modeling method and device for pore network model based on convolutional neural network |
| RU2812143C1 (en) * | 2020-07-15 | 2024-01-23 | Чайна Ойлфилд Сервисез Лимитед | Method and device for measuring characteristics of rock column to create model of pore system |
| CN112862169A (en) * | 2021-01-28 | 2021-05-28 | 中国石油大学(北京) | Method and device for predicting content of free oil in continental facies shale |
| CN112862169B (en) * | 2021-01-28 | 2024-03-29 | 中国石油大学(北京) | Method and device for predicting free oil content of continental phase shale |
| CN113239955A (en) * | 2021-04-07 | 2021-08-10 | 长江大学 | Carbonate reservoir rock classification method |
| CN115830873A (en) * | 2023-01-10 | 2023-03-21 | 西南交通大学 | Urban road traffic event classification method, device, equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110618082B (en) | 2021-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110618082B (en) | Reservoir micro-pore structure evaluation method and device based on neural network | |
| CN108764195B (en) | Handwriting model training method, handwritten character recognition method, device, equipment and medium | |
| WO2019179403A1 (en) | Fraud transaction detection method based on sequence width depth learning | |
| Chacko et al. | Handwritten character recognition using wavelet energy and extreme learning machine | |
| CN112464638B (en) | Text clustering method based on improved spectral clustering algorithm | |
| CN109086653B (en) | Handwriting model training method, handwritten character recognition method, device, equipment and medium | |
| CN110276679B (en) | Network personal credit fraud behavior detection method for deep learning | |
| CN108932455B (en) | Remote sensing image scene recognition method and device | |
| CN112437053B (en) | Intrusion detection method and device | |
| CN114048468A (en) | Intrusion detection method, intrusion detection model training method, device and medium | |
| CN106503743B (en) | Self-adaptive clustering method for local feature points of images with large number and high dimension | |
| CN112749737A (en) | Image classification method and device, electronic equipment and storage medium | |
| CN113744251A (en) | Method for predicting instability of microsatellite from pathological picture based on self-attention mechanism | |
| CN111371611B (en) | Weighted network community discovery method and device based on deep learning | |
| CN113052577A (en) | Method and system for estimating category of virtual address of block chain digital currency | |
| CN110232154B (en) | Random forest-based product recommendation method, device and medium | |
| CN109902167B (en) | Interpretation method and device of embedded result | |
| CN115659823A (en) | Wing profile aerodynamic coefficient prediction method based on attribute reduction, electronic device and storage medium | |
| Zhuang et al. | A handwritten Chinese character recognition based on convolutional neural network and median filtering | |
| Chavan et al. | Mini batch k-means clustering on large dataset | |
| Shayegan et al. | A New Dataset Size Reduction Approach for PCA‐Based Classification in OCR Application | |
| CN113011532A (en) | Classification model training method and device, computing equipment and storage medium | |
| CN107533672A (en) | Pattern recognition device, mode identification method and program | |
| KR100842215B1 (en) | Method and apparatus for multi-class classification using support vector domain description, and computer-readable storage medium used thereto | |
| CN110837853A (en) | Rapid classification model construction 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |