Method and device for determining sedimentary environment of carbonate rock
By using a BP neural network model to generate a gradient matrix from well logging data to update weights, rapid and accurate identification of carbonate sedimentary environments is achieved. This solves the problem of high data requirements in existing technologies, improves identification efficiency and accuracy, and reduces exploration and development costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA PETROLEUM & CHEMICAL CORP
- Filing Date
- 2024-11-11
- Publication Date
- 2026-05-12
AI Technical Summary
Identifying carbonate sedimentary environments using existing technologies requires a variety of data, resulting in limited applicability and low efficiency and accuracy.
A sedimentary environment determination model based on BP neural network is adopted. First-order and second-order gradient matrices are generated using well logging data to update the weights of the input layer and hidden layer. The sedimentary environment of carbonate rocks is automatically analyzed through well logging data.
It improves the efficiency and accuracy of determining carbonate sedimentary environments, reduces exploration and development costs, and is particularly suitable for carbonate strata with similar background geological conditions.
Smart Images
Figure CN122020106A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of sedimentary environment identification technology, particularly to the field of identifying sedimentary environments using geophysical logging technology, specifically to a method, apparatus, equipment, storage medium, and computer program for determining the sedimentary environment of carbonate rocks. Background Technology
[0002] In existing technologies, there are two main categories of methods for studying the sedimentary facies of carbonate rocks: geological analysis methods and geophysical methods. Based on differences in data and implementation methods, geological analysis methods can be further subdivided into microfacies analysis, biofacies methods, experimental analysis methods, and geochemical methods; geophysical methods are divided into mathematical logging and seismic methods (Li Fengfeng et al., 2021). Microfacies analysis refers to a comprehensive sedimentological analysis method and means that, in addition to conventional direct field observation of outcrops, mainly relies on microscopy and other methods to observe and analyze rocks, in order to determine the rock's genetic mechanism, formation environment, and sedimentary process information (Bao Hongping et al., 1999). Biofacies analysis utilizes the specific adaptability of each organism to water quality conditions and water energy to analyze the sedimentary environment. Experimental analysis methods are mainly used to clarify the diagenetic processes experienced by rocks and reconstruct the original sedimentary state. The geological basis is as follows: diagenesis has a certain inheritance relationship with sedimentation. Sedimentation lays the material foundation for diagenesis and controls the type of diagenesis. Diagenetic mechanisms can be determined through diagenetic facies, and the original sedimentary state can be restored after diagenetic evolution, thus inferring the sedimentary environment. The principle of elemental analysis is that the distribution, ratio changes, and combinations of elements in strata indicate the evolutionary process of paleoclimate environment. Minerals are the information carriers of paleoenvironmental research, and elements are the basic units that make up minerals and rocks. The geochemical characteristics of elements can be used to identify sedimentary environments (Wang Kaiming et al., 2009). The mathematical-logging method determines the characteristic parameters of carbonate sedimentary facies through rock-electrical calibration, establishes well logging response functions, and optimizes mathematical algorithms to interpret single-well sedimentary facies. The seismic method establishes sedimentary facies seismic response models through well-seismic calibration, and converts seismic facies into sedimentary facies through qualitative or quantitative methods. The seismic-sedimentary facies response ultimately boils down to differences in lithology or paleogeography forming different seismic reflection structures, i.e., seismic lithology and seismic geomorphology (Zeng Hongliu, 2011).
[0003] The above methods all require a large amount of data, such as rock cores, outcrops, rock fragments, sedimentary structures, and paleontological assemblages, so their application is not universal. Summary of the Invention
[0004] This disclosure provides a method, apparatus, equipment, storage medium, and computer program for determining the sedimentary environment of carbonate rocks. It effectively solves the technical pain point of existing technologies that require multiple data to classify the sedimentary microfacies of carbonate rocks. This method only requires well logging data to quickly analyze the sedimentary environment, which greatly improves the efficiency and accuracy of determining the sedimentary environment of carbonate rocks and reduces exploration and development costs.
[0005] In a first aspect, this disclosure provides a method for determining the sedimentary environment of carbonate rocks, including:
[0006] Obtain well logging data of carbonate rocks within the research area;
[0007] The depositional environment of the carbonate rocks is determined based on the well logging data and a pre-generated depositional environment determination model. The depositional environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the depositional environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
[0008] In some embodiments of this disclosure, the steps of generating the first-order gradient matrix and the second-order gradient matrix based on the well logging data include:
[0009] Generate a multi-dimensional vector based on multiple well logging data;
[0010] Calculate the covariance matrix of the multidimensional vector;
[0011] The attenuation rate is determined based on the covariance matrix.
[0012] The initial matrix of the first-order gradient matrix is updated according to the decay rate to generate the first-order gradient matrix;
[0013] The initial matrix of the second-order gradient matrix is updated according to the decay rate to generate the second-order gradient matrix.
[0014] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0015] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the sedimentary environment determination model;
[0016] The characteristics of the logging data corresponding to the sedimentary environment are determined according to the order of importance.
[0017] In some embodiments of this disclosure, determining the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model includes:
[0018] The feature values of the well logging data are randomly shuffled to generate shuffled feature values;
[0019] The shuffled feature values are input into the sedimentation environment determination model, and the differences in the computational performance of the sedimentation environment determination model are determined.
[0020] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the differences in computational performance.
[0021] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0022] During the forward propagation training of the deposition environment determination model, the input layer data is generated based on the input layer data, the weights, and the preset bias.
[0023] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0024] The dimension of the sedimentary environment determination model is determined based on the number of sedimentary environments.
[0025] Based on the dimension and the sedimentary environment, the probability of the model's predicted samples being positive examples of the corresponding sedimentary environment is determined, generating the loss function in the forward propagation training process.
[0026] Secondly, this disclosure provides an apparatus for determining the sedimentary environment of carbonate rocks, comprising:
[0027] The well logging data acquisition module is used to acquire well logging data of carbonate rocks in the research area;
[0028] The sedimentary environment determination module is used to determine the sedimentary environment of the carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model. The sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
[0029] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0030] A gradient matrix generation module is used to generate the first-order gradient matrix and the second-order gradient matrix based on the well logging data.
[0031] The gradient matrix generation module includes:
[0032] A multi-dimensional vector generation unit is used to generate multi-dimensional vectors based on multiple well logging data.
[0033] A covariance matrix calculation unit is used to calculate the covariance matrix of the multidimensional vector;
[0034] The attenuation rate determination unit is used to determine the attenuation rate based on the covariance matrix.
[0035] A first-order gradient matrix generation unit is used to update the initial matrix of the first-order gradient matrix according to the decay rate to generate the first-order gradient matrix.
[0036] The second-order gradient matrix generation unit is used to update the initial matrix of the second-order gradient matrix according to the decay rate to generate the second-order gradient matrix.
[0037] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0038] The sequence determination unit is used to determine the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model.
[0039] The well logging data feature determination unit is used to determine the features of the well logging data corresponding to the sedimentary environment according to the importance order.
[0040] In some embodiments of this disclosure, the sequence determination unit includes:
[0041] The feature value shuffling unit is used to randomly shuffle the feature values of the well logging data to generate shuffled feature values;
[0042] The performance change difference determination unit is used to input the shuffled feature values into the deposition environment determination model and determine the performance change difference of the deposition environment determination model.
[0043] The importance order determination unit is used to determine the importance order of multiple well logging data corresponding to the sedimentary environment based on the differences in computing performance.
[0044] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0045] The input layer data generation module is used to generate the input layer data based on the input layer data, the weights, and a preset bias during the forward propagation training of the model in the deposition environment determination process.
[0046] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0047] A dimension determination module is used to determine the dimension of the sedimentation environment determination model based on the number of sedimentation environments.
[0048] The loss function generation module is used to determine the probability that the model's predicted samples are positive examples of the corresponding sedimentary environment based on the dimension and the sedimentary environment, and to generate the loss function during the forward propagation training process.
[0049] Thirdly, this disclosure provides a computer device including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the foregoing aspects.
[0050] Fourthly, this disclosure provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the methods described in the above aspects.
[0051] Fifthly, this disclosure provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the methods described in the foregoing aspects.
[0052] This disclosure provides a method, apparatus, equipment, storage medium, and computer program for determining the sedimentary environment of carbonate rocks. The corresponding method for determining the sedimentary environment of carbonate rocks includes: first, acquiring well logging data of carbonate rocks in the study area; then, determining the sedimentary environment of carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model; wherein the sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on pre-generated first-order gradient matrices and second-order gradient matrices, which are generated based on the well logging data.
[0053] The corresponding device includes: a well logging data acquisition module for acquiring well logging data of carbonate rocks in the study area; and a sedimentary environment determination module for determining the sedimentary environment of carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model. The sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
[0054] In summary, this invention automatically classifies and identifies carbonate sedimentary environments using only conventional well logging curves. Furthermore, it can continuously improve its identification accuracy through learning. This method is particularly suitable for carbonate formations with similar background geological conditions. It can quickly and systematically analyze well logging curve information of target strata, accurately converting it into carbonate sedimentary environments, significantly improving the efficiency of exploration and card-making work and reducing labor costs. Attached Figure Description
[0055] The present disclosure will be described in more detail below based on embodiments and with reference to the accompanying drawings:
[0056] Figure 1 This is a schematic flowchart illustrating a method for determining the sedimentary environment of carbonate rocks, as provided in an embodiment of this disclosure.
[0057] Figure 2 This is a second flowchart illustrating a method for determining the sedimentary environment of carbonate rocks, provided in an embodiment of this disclosure.
[0058] Figure 3 This is a flowchart illustrating step 300 of a method for determining the sedimentary environment of carbonate rocks, as provided in an embodiment of this disclosure.
[0059] Figure 4 This is a schematic diagram of a third process for determining the sedimentary environment of carbonate rocks, provided in an embodiment of this disclosure.
[0060] Figure 5 This is a flowchart illustrating step 400 of a method for determining the depositional environment of carbonate rocks, as provided in an embodiment of this disclosure.
[0061] Figure 6 This is a schematic diagram of the fourth process for determining the sedimentary environment of carbonate rocks, provided in an embodiment of this disclosure.
[0062] Figure 7 This is a fifth flowchart illustrating a method for determining the sedimentary environment of carbonate rocks, as provided in an embodiment of this disclosure.
[0063] Figure 8 A flowchart illustrating a method for determining the sedimentary environment of carbonate rocks, provided as an application example of this disclosure.
[0064] Figure 9 A schematic diagram of the structure of the deep neural network carbonate sedimentary environment classification model provided as an application example of this disclosure.
[0065] Figure 10 A schematic diagram illustrating the results of the application example in Region A provided for this public application example.
[0066] Figure 11 A block diagram of an apparatus for determining the sedimentary environment of carbonate rocks provided in an embodiment of this disclosure.
[0067] In the accompanying drawings, the same parts are referred to by the same reference numerals, and the drawings are not drawn to scale. Detailed Implementation
[0068] To enable those skilled in the art to better understand the technical solutions of this disclosure, and to fully understand and implement the process of how this disclosure applies technical means to solve technical problems and achieve corresponding technical effects, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, not all embodiments. The embodiments of this disclosure and the various features within them can be combined with each other without conflict, and the resulting technical solutions are all within the protection scope of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort should fall within the protection scope of this disclosure.
[0069] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0070] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0071] Example 1
[0072] This disclosure provides a method for determining the depositional environment of carbonate rocks. Figure 1 This is a schematic flowchart illustrating a method for determining the sedimentary environment of carbonate rocks, provided as an embodiment of this disclosure. Figure 1 As shown, a method for determining the sedimentary environment of carbonate rocks includes:
[0073] Step 100: Obtain well logging data of carbonate rocks in the study area;
[0074] Step 200: Determine the depositional environment of the carbonate rock based on the well logging data and the pre-generated depositional environment determination model; wherein the depositional environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the depositional environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, and the first-order gradient matrix and the second-order gradient matrix are generated based on the well logging data.
[0075] This disclosure provides a method for determining the sedimentary environment of carbonate rocks, comprising: first, acquiring well logging data of carbonate rocks in the study area; then, determining the sedimentary environment of carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model; wherein the sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on pre-generated first-order gradient matrices and second-order gradient matrices, and the first-order gradient matrices and second-order gradient matrices are generated based on the well logging data.
[0076] In summary, this invention automatically classifies and identifies carbonate sedimentary environments using only conventional well logging curves. Furthermore, it can continuously improve its identification accuracy through learning. This method is particularly suitable for carbonate formations with similar background geological conditions. It can quickly and systematically analyze well logging curve information of target strata, accurately converting it into carbonate sedimentary environments, significantly improving the efficiency of exploration and card-making work and reducing labor costs.
[0077] Example 2
[0078] Regarding steps 100 and 200, it is understood that many large oil and gas fields and high-yield oil and gas wells worldwide produce oil and gas from carbonate reservoirs, accounting for over 60% of global production. Therefore, accurate classification of the sedimentary environment of carbonate reservoirs is crucial. To classify the sedimentary environment of carbonate rocks, it is first necessary to understand reservoir genesis, establish a reservoir genetic model, and clarify the relationship between rock structural components and pore type characteristics. Based on the rock physical response mechanism of rock composition, structure, and pores, a comprehensive well logging response model is established. Well logging data is then comprehensively processed to obtain the identification results of the lithology and sedimentary environment of the carbonate reservoir. The machine analysis results are further corrected using accurate data such as core samples. The accuracy of machine analysis is gradually improved through a backpropagation (BP) neural network, ultimately establishing a geological-geophysical transformation model and dataset. This allows the machine to accurately describe the sedimentary environment of carbonate reservoirs by analyzing geophysical information such as well logging data. This invention improves the interpretation efficiency of carbonate sedimentary environments and reduces corresponding exploration and development costs.
[0079] For step 200, BP neural network technology is a nonlinear intelligent information processing technology. Currently, neural network technology has been widely applied in oil well logging interpretation (Zhao Junlong et al., 2010), with major achievements including: lithological interpretation and lithofacies identification (Dai Libo et al., 2009); reservoir parameter interpretation (Chen Rong et al., 2009); reservoir fluid property identification (Zhang Baiyuan, 2002); low resistivity and fracture research (Shen Huilin et al., 2007); formation pressure calculation and prediction, etc. (Yang Bin et al., 2005). Therefore, this application of BP neural networks in the interpretation of carbonate sedimentary environments will further improve interpretation efficiency and reduce corresponding exploration and development costs.
[0080] Example 3
[0081] Based on the above embodiments, see Figure 2 A method for determining the sedimentary environment of carbonate rocks, further comprising:
[0082] Step 300: Generate the first-order gradient matrix and the second-order gradient matrix based on the well logging data. Then, see... Figure 3 Step 300 includes:
[0083] Step 301: Generate a multidimensional vector based on multiple well logging data;
[0084] Specifically, n logging data points are selected from the relevant logging data, and these data are combined into an n-dimensional vector y, y = (y1, y2, ..., y3). n ) T Furthermore, the dimension of the multidimensional vector is equal to the number of categories in the well logging data.
[0085] Step 302: Calculate the covariance matrix of the multidimensional vector;
[0086] Specifically, the mean vector of the multidimensional vector is first calculated, and then the mean vector is subtracted from the eigenvalue corresponding to each element in the multidimensional vector to obtain the centered data matrix (multidimensional vector), and the covariance matrix is calculated based on the data matrix.
[0087] Step 303: Determine the attenuation rate based on the covariance matrix;
[0088] Specifically, if the diagonal elements of the covariance matrix (i.e., the variance of each feature) are large, it indicates that the data varies significantly across these features, requiring a smaller decay rate to accommodate this variation. Therefore, the initial decay rate should be reduced. Alternatively, the covariance matrix can be eigenvalued to obtain a set of eigenvalues and eigenvectors. Eigenvalues represent the main directions of data variation, while eigenvectors represent these directions. If the variance of some eigenvalues is high, the decay rate can be appropriately reduced. Conversely, if the variance of some eigenvalues is low, the decay rate can be appropriately increased.
[0089] Step 304: Update the initial matrix of the first-order gradient matrix according to the decay rate to generate the first-order gradient matrix;
[0090] First, initialize the first-order moment estimation variables as zero vectors or zero tensors, with the same dimension as the optimization parameters. Then, for each time step, calculate the current gradient and update the first-order gradient matrix using the decay rate. Specifically: update the first-order gradient matrix: m_t = β*m_(t-1) + (1-β)*g_t; where β is the decay rate, ε is a very small constant, g_t is the current gradient, and t is the time step.
[0091] Step 305: Update the initial matrix of the second-order gradient matrix according to the decay rate to generate the second-order gradient matrix.
[0092] This step is similar to the implementation process of step 304, and will not be repeated here.
[0093] Based on the above embodiments, see Figure 4 A method for determining the sedimentary environment of carbonate rocks, further comprising:
[0094] Step 400: Determine the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model;
[0095] Step 500: Determine the characteristics of the logging data corresponding to the sedimentary environment according to the order of importance.
[0096] Specifically, the characteristics and differences of well logging curves for different carbonate sedimentary environments are obtained based on the ranking of importance.
[0097] Based on the above embodiments, see Figure 5 Step 400 includes:
[0098] Step 401: Randomly shuffle the feature values of the well logging data to generate shuffled feature values;
[0099] Step 402: Input the shuffled feature values into the sedimentation environment determination model, and determine the differences in the computational performance of the sedimentation environment determination model;
[0100] The computational performance includes the baseline performance of the model determined by the deposition environment (e.g., accuracy, mean squared error, etc.). In steps 401 and 402, the feature values are randomly shuffled (i.e., the feature is permuted). The permuted data is then input into the model, and the model's performance on the validation or test set is calculated.
[0101] Step 403: Determine the importance order of the multiple logging data corresponding to the sedimentary environment based on the differences in computing performance.
[0102] For each feature value, calculate the performance difference before and after the permutation. The larger the difference, the greater the contribution of that feature value to the model performance, and the higher its importance.
[0103] Example 4
[0104] Based on the above embodiments, see Figure 6 A method for determining the sedimentary environment of carbonate rocks, further comprising:
[0105] Step 600: During the forward propagation training of the deposition environment determination model, the input layer data is generated based on the input layer data, the weights, and the preset bias.
[0106] The sedimentary environment determination model consists of three parts: an input layer, multiple hidden layers, and an output layer. The input layer receives the input logging data, and its number of nodes equals the number of types of input logging data. During forward propagation, the calculation formula from the input layer to the hidden layers is as follows:
[0107]
[0108] Among them, W i X represents the weight of the connection between the i-th node in the input layer and the node in the hidden layer. i represents the input data of the i-th node in the input layer, and b is the bias.
[0109] Based on the above embodiments, see Figure 7 A method for determining the sedimentary environment of carbonate rocks, further comprising:
[0110] Step 700: Determine the dimension of the sedimentary environment determination model based on the number of sedimentary environments;
[0111] Step 800: Based on the dimension and the sedimentation environment, determine the positive probability of the model's predicted samples being positive examples of the corresponding sedimentation environment, and generate the loss function in the forward propagation training process.
[0112] In steps 700 and 800, the input values of the hidden layer, after being calculated by the activation function, are passed to the next layer as the output values of the hidden layer. The activation function expression is:
[0113]
[0114] After forward propagation is complete, the error (Loss) between the calculated result and the sample label is evaluated using a loss function, the expression of which is:
[0115]
[0116] Where n is the dimension of the model output value (i.e., the number of sedimentary environments), y is the probability that the model predicts a sample is a positive example in the corresponding sedimentary environment i. i It is the label of the sample on the corresponding deposition environment i. If the sample is a positive example, the value is 1, otherwise the value is 0.
[0117] In summary, this invention provides a method for determining the sedimentary environment of carbonate rocks, comprising: collecting well logging curve data of the target stratigraphic level in the study area; analyzing the sedimentary environment of carbonate rocks based on the petroleum geological conditions and core data of the study area to obtain a classification scheme for carbonate rock sedimentary environments; assigning labels to each well logging curve data according to the classification scheme to construct a dataset suitable for supervised learning; constructing a deep neural network model; training and optimizing the deep neural network model based on the dataset to obtain an optimized deep neural network model; and analyzing the well logging curves using the optimized deep neural network model to obtain the results of predicting the sedimentary environment of carbonate rocks using the well logging curves. In conclusion, this invention, by employing a deep neural network model and well logging data, achieves the analysis of the sedimentary environment of carbonate rocks, greatly accelerating the research progress and improving the accuracy of research.
[0118] Example 5
[0119] To further illustrate the solution, based on the above embodiments, this embodiment takes region A as an example to provide an application example to further explain a method for determining the sedimentary environment of carbonate rocks.
[0120] In this application example, a carbonate sedimentary environment analysis system based on a deep neural network model is first provided. The system includes:
[0121] The data acquisition module is used to collect well logging curve data of the target strata in the study area;
[0122] The sedimentary environment analysis module is used to analyze the carbonate rock sedimentary environment based on the petroleum geological conditions and core data of the study area, and to obtain a classification scheme for carbonate rock sedimentary environments.
[0123] The dataset construction module is used to assign labels to each of the well logging curve data according to the carbonate rock sedimentary environment classification scheme, and construct a dataset suitable for supervised learning.
[0124] A deep neural network model building module is used to build a deep neural network model; and to train and optimize the deep neural network model based on the dataset to obtain an optimized deep neural network model;
[0125] Optionally, a carbonate sedimentary environment analysis system based on a deep neural network model further includes:
[0126] The sensitivity analysis module is used to interpret the optimized deep neural network model according to the permutation feature importance algorithm to obtain the importance ranking of the logging curve parameters;
[0127] The well logging curve characteristics and differences module is used to sort the well logging curves according to their importance and obtain the characteristics and differences of different carbonate sedimentary environments.
[0128] See Figure 8 Based on the above-mentioned carbonate rock sedimentary environment analysis system based on a deep neural network model, the application example of the method for determining the sedimentary environment of carbonate rocks provided by this invention includes the following steps:
[0129] S1: Well logging curves of carbonate rocks from different sedimentary environments in the target strata of the study area.
[0130] Specifically, well logging curves of the target carbonate rock formations in the study area were collected, including: lithological logging sequences (natural gamma curve GR, spontaneous potential SP, caliper curve CAL); resistivity logging curves (deep lateral resistivity curve RD, shallow lateral resistivity curve RS); and porosity logging curves (neutron curve CNL, density curve DNL, sonic transit time curve AC) to predict the subsequent carbonate rock depositional environment.
[0131] S2: Based on the petroleum geological conditions and core data of the study area, the sedimentary environment of the target strata is analyzed to obtain a classification scheme for carbonate rock sedimentary environments;
[0132] Specifically, based on the petroleum geological conditions and core data of the study area, the sedimentary environments of the target strata were classified to obtain a classification scheme for carbonate sedimentary environments. The classification scheme refers to the comprehensive model diagram of carbonate sedimentary facies (Wilson, 1975).
[0133] S3: Based on the obtained sedimentary environment classification scheme, assign labels to each of the well logging data to construct a dataset suitable for supervised learning;
[0134] Preferably, the feature data of the dataset can also be processed using a formula. Perform Min-Max normalization;
[0135] Where x and x' represent the original data and the normalized data, respectively, and min(x) and max(x) represent the minimum and maximum values in a set of elements waiting to be normalized.
[0136] Preferably, step S3 further includes: interpreting the optimized deep neural network model according to the permutation feature importance algorithm to obtain the importance ranking of different logging curve parameters; and obtaining the logging curve characteristics and differences of different carbonate sedimentary environments according to the importance ranking.
[0137] S4: Construct a deep neural network model.
[0138] Specifically, see Figure 9 The figure shows a schematic diagram of the deep neural network sedimentary environment classification model constructed in this invention. The deep neural network model comprises three parts: an input layer, multiple hidden layers, and an output layer. The input layer receives the input logging curve data, and its number of nodes equals the number of types of input logging curves. During forward propagation, the calculation formula from the input layer to the hidden layers is as follows:
[0139]
[0140] Among them, W i X represents the weight of the connection between the i-th node in the input layer and the node in the hidden layer. i represents the input data of the i-th node in the input layer, and b is the bias.
[0141] The input values of the hidden layer, after being processed by the activation function, are passed as the output values to the next layer. The activation function used is the Softmax function, suitable for multi-class classification problems, and its expression is:
[0142]
[0143] After forward propagation, the error (Loss) between the calculated result and the sample label is evaluated using the Categorical Crossentropy loss function, which is expressed as the formula:
[0144]
[0145] Where n is the dimension of the model output value (i.e., the number of sedimentary environments), y is the probability that the model predicts a sample is a positive example in the corresponding sedimentary environment i. i It is the label of the sample on the corresponding deposition environment i. If the sample is a positive example, the value is 1, otherwise the value is 0.
[0146] The deep neural network model is trained using an adaptive moment estimation optimization algorithm. The weight update rule of the deep neural network model is as follows:
[0147]
[0148] Where Vt is the first-order gradient matrix and St is the second-order gradient matrix. and These are the deviation corrections for the original values; 'a' is the learning rate, set to 0.001; and 'ε' is a very small constant with a value of 10. -8 To avoid division by zero, θ is the weight before the update. t+1 This is the updated weight.
[0149] S5: Train and optimize the deep neural network model based on the dataset to obtain an optimized deep neural network model.
[0150] Specifically, the test sample dataset is divided into a training set and a test set at a ratio of 70% to 30%, with the samples randomly distributed between the two sets to ensure that their data features are identical. The training set is used to train the model, and the test set is used to test the classification accuracy of the trained model.
[0151] S6: Analyze the sedimentary environment based on the optimized deep neural network model to obtain the prediction results of the well logging curve.
[0152] The well logging curves are analyzed based on the optimized deep neural network model to obtain the predicted carbonate sedimentary environment. (See [link to relevant documentation]). Figure 10 .
[0153] This disclosure provides a method for determining the sedimentary environment of carbonate rocks, comprising: first, acquiring well logging data of carbonate rocks in the study area; then, determining the sedimentary environment of carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model; wherein the sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, and the first-order gradient matrix and second-order gradient matrix are generated based on the well logging data.
[0154] Specifically, firstly, well logging data of the target stratigraphic level in the study area are collected; secondly, the sedimentary environment of carbonate rocks is analyzed based on the petroleum geological conditions and core data of the study area to obtain a sedimentary environment classification scheme; thirdly, labels are assigned to each well logging data according to the sedimentary environment classification scheme to construct a dataset suitable for supervised learning; fourthly, a deep neural network model is constructed; fifthly, the deep neural network model is trained and optimized based on the dataset to obtain an optimized deep neural network model; and finally, well logging curves are analyzed based on the optimized deep neural network model to obtain sedimentary environment prediction results. Compared with traditional methods of manually classifying carbonate rock sedimentary microfacies using cores, outcrops, rock fragments, sedimentary structures, and paleontological assemblages, this invention can quickly analyze the sedimentary environment using only well logging data, greatly improving analysis efficiency and accuracy, and reducing exploration and development costs.
[0155] Example 6
[0156] Based on the same inventive concept, this application also provides a device for determining the sedimentary environment of carbonate rocks, which can be used to implement the method described in the above embodiments, as shown in the following embodiments. Since the principle of the device for determining the sedimentary environment of carbonate rocks is similar to that of the method for determining the sedimentary environment of carbonate rocks, the implementation of the device for determining the sedimentary environment of carbonate rocks can refer to the implementation of the method for determining the sedimentary environment of carbonate rocks, and repeated details will not be elaborated further. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.
[0157] The embodiments of the present invention provide a specific implementation of a carbonate rock sedimentary environment determination apparatus capable of realizing a method for determining the sedimentary environment of carbonate rocks, wherein, see [link to specific implementation details]. Figure 11 An apparatus for determining the sedimentary environment of carbonate rocks includes:
[0158] The well logging data acquisition module 10 is used to acquire well logging data of carbonate rocks in the research area;
[0159] The sedimentary environment determination module 20 is used to determine the sedimentary environment of the carbonate rocks based on the well logging data and the pre-generated sedimentary environment determination model; wherein the sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and the hidden layer of the sedimentary environment determination model are updated based on the pre-generated first-order gradient matrix and the second-order gradient matrix, and the first-order gradient matrix and the second-order gradient matrix are generated based on the well logging data.
[0160] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0161] A gradient matrix generation module is used to generate the first-order gradient matrix and the second-order gradient matrix based on the well logging data.
[0162] The gradient matrix generation module includes:
[0163] A multi-dimensional vector generation unit is used to generate multi-dimensional vectors based on multiple well logging data.
[0164] A covariance matrix calculation unit is used to calculate the covariance matrix of the multidimensional vector;
[0165] The attenuation rate determination unit is used to determine the attenuation rate based on the covariance matrix.
[0166] A first-order gradient matrix generation unit is used to update the initial matrix of the first-order gradient matrix according to the decay rate to generate the first-order gradient matrix.
[0167] The second-order gradient matrix generation unit is used to update the initial matrix of the second-order gradient matrix according to the decay rate to generate the second-order gradient matrix.
[0168] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0169] The sequence determination unit is used to determine the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model.
[0170] The well logging data feature determination unit is used to determine the features of the well logging data corresponding to the sedimentary environment according to the importance order.
[0171] In some embodiments of this disclosure, the sequence determination unit includes:
[0172] The feature value shuffling unit is used to randomly shuffle the feature values of the well logging data to generate shuffled feature values;
[0173] The performance change difference determination unit is used to input the shuffled feature values into the deposition environment determination model and determine the performance change difference of the deposition environment determination model.
[0174] The importance order determination unit is used to determine the importance order of multiple well logging data corresponding to the sedimentary environment based on the differences in computing performance.
[0175] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0176] The input layer data generation module is used to generate the input layer data based on the input layer data, the weights, and the preset bias during the forward propagation training of the deposition environment determination model.
[0177] In some embodiments of this disclosure, an apparatus for determining the depositional environment of carbonate rocks further includes:
[0178] A dimension determination module is used to determine the dimension of the sedimentation environment determination model based on the number of sedimentation environments.
[0179] The loss function generation module is used to determine the probability that the model's predicted samples are positive examples of the corresponding sedimentary environment based on the dimension and the sedimentary environment, and to generate the loss function during the forward propagation training process.
[0180] This disclosure provides a device for determining the sedimentary environment of carbonate rocks, comprising: a well logging data acquisition module for acquiring well logging data of carbonate rocks in a study area; and a sedimentary environment determination module for determining the sedimentary environment of carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model. The sedimentary environment determination model is generated based on a BP neural network, and the weights of the input and hidden layers of the sedimentary environment determination model are updated based on pre-generated first-order gradient matrices and second-order gradient matrices, which are generated based on the well logging data.
[0181] In summary, this invention automatically classifies and identifies carbonate sedimentary environments using only conventional well logging curves. Furthermore, it can continuously improve its identification accuracy through learning. This method is particularly suitable for carbonate formations with similar background geological conditions. It can quickly and systematically analyze well logging curve information of target strata, accurately converting it into carbonate sedimentary environments, significantly improving the efficiency of exploration and card-making work and reducing labor costs.
[0182] Example 7
[0183] Based on the above embodiments, this embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the above embodiments.
[0184] In some embodiments of this example, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program implements the steps of the method described in the above embodiments, specifically including the following:
[0185] Obtain well logging data of carbonate rocks within the research area;
[0186] The depositional environment of the carbonate rocks is determined based on the well logging data and a pre-generated depositional environment determination model. The depositional environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the depositional environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
[0187] In some embodiments of this disclosure, the steps of generating the first-order gradient matrix and the second-order gradient matrix based on the well logging data include:
[0188] Generate a multi-dimensional vector based on multiple well logging data;
[0189] Calculate the covariance matrix of the multidimensional vector;
[0190] The attenuation rate is determined based on the covariance matrix.
[0191] The initial matrix of the first-order gradient matrix is updated according to the decay rate to generate the first-order gradient matrix;
[0192] The initial matrix of the second-order gradient matrix is updated according to the decay rate to generate the second-order gradient matrix.
[0193] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0194] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the sedimentary environment determination model;
[0195] The characteristics of the logging data corresponding to the sedimentary environment are determined according to the order of importance.
[0196] In some embodiments of this disclosure, determining the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model includes:
[0197] The feature values of the well logging data are randomly shuffled to generate shuffled feature values;
[0198] The shuffled feature values are input into the sedimentation environment determination model, and the differences in the computational performance of the sedimentation environment determination model are determined.
[0199] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the differences in computational performance.
[0200] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0201] During the forward propagation training of the deposition environment determination model, the input layer data is generated based on the input layer data, the weights, and the preset bias.
[0202] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0203] The dimension of the sedimentary environment determination model is determined based on the number of sedimentary environments.
[0204] Based on the dimension and the sedimentary environment, the probability of the model's predicted samples being positive examples of the corresponding sedimentary environment is determined, generating the loss function in the forward propagation training process.
[0205] In some embodiments of this example, a computer program product is provided, including a computer program / instructions, which, when executed by a processor, implements the steps of the method described in the above embodiments, specifically including the following:
[0206] Obtain well logging data of carbonate rocks within the research area;
[0207] The depositional environment of the carbonate rocks is determined based on the well logging data and a pre-generated depositional environment determination model. The depositional environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the depositional environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
[0208] In some embodiments of this disclosure, the steps of generating the first-order gradient matrix and the second-order gradient matrix based on the well logging data include:
[0209] Generate a multi-dimensional vector based on multiple well logging data;
[0210] Calculate the covariance matrix of the multidimensional vector;
[0211] The attenuation rate is determined based on the covariance matrix.
[0212] The initial matrix of the first-order gradient matrix is updated according to the decay rate to generate the first-order gradient matrix;
[0213] The initial matrix of the second-order gradient matrix is updated according to the decay rate to generate the second-order gradient matrix.
[0214] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0215] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the sedimentary environment determination model;
[0216] The characteristics of the logging data corresponding to the sedimentary environment are determined according to the order of importance.
[0217] In some embodiments of this disclosure, determining the importance order of multiple well logging data corresponding to the sedimentary environment based on the sedimentary environment determination model includes:
[0218] The feature values of the well logging data are randomly shuffled to generate shuffled feature values;
[0219] The shuffled feature values are input into the sedimentation environment determination model, and the differences in the computational performance of the sedimentation environment determination model are determined.
[0220] The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the differences in computational performance.
[0221] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0222] During the forward propagation training of the deposition environment determination model, the input layer data is generated based on the input layer data, the weights, and the preset bias.
[0223] In some embodiments of this disclosure, a method for determining the depositional environment of carbonate rocks further includes:
[0224] The dimension of the sedimentary environment determination model is determined based on the number of sedimentary environments.
[0225] Based on the dimension and the sedimentary environment, the probability of the model's predicted samples being positive examples of the corresponding sedimentary environment is determined, generating the loss function in the forward propagation training process.
[0226] The processor may include, but is not limited to, one or more processors or microprocessors. Each processor may be implemented as an Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor, or other electronic component, for executing the methods in the above embodiments.
[0227] Computer-readable storage media can be implemented by any type of volatile or non-volatile storage device or a combination thereof. Computer-readable storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, and computer storage media (e.g., hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.).
[0228] Computer-readable storage media may also store at least one computer-executable program / instruction, such as computer-readable instructions. Computer-readable storage media include, but are not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Computer-readable storage media may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the computer-readable storage medium, the various methods described above can be performed.
[0229] In addition, the computer device may include (but is not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (e.g., keyboard, mouse, speakers, etc.).
[0230] The processor can communicate with external devices via the I / O bus through wired or wireless networks.
[0231] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product / computer program product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.
[0232] In the embodiments provided in this disclosure, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.
[0233] It should be noted that, in this disclosure, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element limited by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0234] While the embodiments disclosed herein are as described above, the foregoing content is merely for the purpose of facilitating understanding of this disclosure and is not intended to limit this disclosure. Any person skilled in the art to which this disclosure pertains may make any modifications and changes in form and detail of the implementation without departing from the spirit and scope of this disclosure; however, the scope of patent protection of this disclosure shall still be determined by the scope defined in the appended claims.
Claims
1. A method for determining the sedimentary environment of carbonate rocks, characterized in that, include: Obtain well logging data of carbonate rocks within the research area; The depositional environment of the carbonate rocks is determined based on the well logging data and a pre-generated depositional environment determination model. The depositional environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the depositional environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
2. The method for determining the sedimentary environment according to claim 1, characterized in that, The steps of generating the first-order gradient matrix and the second-order gradient matrix based on the well logging data include: Generate a multi-dimensional vector based on multiple well logging data; Calculate the covariance matrix of the multidimensional vector; The attenuation rate is determined based on the covariance matrix. The initial matrix of the first-order gradient matrix is updated according to the decay rate to generate the first-order gradient matrix; The initial matrix of the second-order gradient matrix is updated according to the decay rate to generate the second-order gradient matrix.
3. The method for determining the sedimentary environment according to claim 1, characterized in that, Also includes: The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the sedimentary environment determination model; The characteristics of the logging data corresponding to the sedimentary environment are determined according to the order of importance.
4. The method for determining the sedimentary environment according to claim 3, characterized in that, The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the sedimentary environment determination model, including: The feature values of the well logging data are randomly shuffled to generate shuffled feature values; The shuffled feature values are input into the sedimentation environment determination model, and the differences in the computational performance of the sedimentation environment determination model are determined. The importance order of multiple well logging data corresponding to the sedimentary environment is determined based on the differences in computational performance.
5. The method for determining the sedimentary environment according to claim 1, characterized in that, Also includes: During the forward propagation training of the deposition environment determination model, the input layer data is generated based on the input layer data, the weights, and the preset bias.
6. The method for determining the sedimentary environment according to claim 5, characterized in that, Also includes: The dimension of the sedimentary environment determination model is determined based on the number of sedimentary environments. Based on the dimension and the sedimentary environment, the probability of the model's predicted samples being positive examples of the corresponding sedimentary environment is determined, generating the loss function in the forward propagation training process.
7. A device for determining the sedimentary environment of carbonate rocks, characterized in that, include: The well logging data acquisition module is used to acquire well logging data of carbonate rocks in the research area; The sedimentary environment determination module is used to determine the sedimentary environment of the carbonate rocks based on the well logging data and a pre-generated sedimentary environment determination model. The sedimentary environment determination model is generated based on a BP neural network, and the weights of the input layer and hidden layer of the sedimentary environment determination model are updated based on the pre-generated first-order gradient matrix and second-order gradient matrix, which are generated based on the well logging data.
8. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method for determining the sedimentary environment of carbonate rocks according to any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method for determining the depositional environment of carbonate rocks according to any one of claims 1 to 6.
10. A computer program product comprising a computer program / instructions, characterized in that, When executed by a processor, the computer program implements the steps of the method for determining the depositional environment of carbonate rocks according to any one of claims 1 to 6.