Petroleum proppant raw material ratio automatic optimization method and system

By generating adversarial network and neural network models through conditions, combining historical and new geological information, dynamically adjusting the ratio of petroleum proppant, solving the ratio of traditional methods under new geological conditions, improving production efficiency and accuracy, and achieving more Efficient production of petroleum proppant.

CN118942569AActive Publication Date: 2024-11-12GUIYANG XINRUI MATERIAL TECH CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411429696.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-14
Publication Date
2024-11-12
Estimated Expiration
2044-10-14

AI Technical Summary

Technical Problem

Traditional petroleum proppant raw material rationing methods rely on experience and trial and error, making it difficult to achieve optimal ratio under new geological conditions, affecting the performance and cost-effectiveness of proppant, and the lack of sufficient experimental data leads to insufficient model training and insufficient prediction accuracy. .

Method used

Conditionally generated adversarial network model and neural network model are adopted. By collecting historical geological information and proportional data, the adversarial network and performance index prediction model are trained to generate new proportional data under new geological conditions, and the linear and nonlinear adjustment strategies are used. Dynamically adjust the ratio to ensure that the performance indicators meet the standards.

Benefits of technology

It improves the production efficiency and product quality of petroleum proppants, generates more accurate new proportioning data, improves the accuracy and efficiency of proportioning adjustment, adapts to complex geological conditions, and makes dynamic adjustment strategies more flexible and intelligent.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118942569A_ABST
    Figure CN118942569A_ABST
Patent Text Reader

Abstract

The invention discloses a petroleum proppant raw material ratio automatic optimization method and system, and relates to the technical field of petroleum proppant production, and the method comprises the steps: collecting geological information, historical ratio data and performance index data, training a cGAN model, obtaining first geological information and first ratio data through the cGAN model, and carrying out the training to obtain a performance index prediction model. Utilizing the performance index prediction model to predict first performance index data, setting a standard performance index, if the first performance index data is smaller than the standard performance index, adopting an optimal matching strategy to obtain optimal matching data, and if the first performance index data is larger than or equal to the standard performance index, adopting an optimal matching strategy to obtain optimal matching data; if yes, the first matching data serve as the optimal matching data to produce the proppant, actual geological information is obtained, preliminary matching data is generated, predicted performance index data is obtained, the preliminary matching data is adjusted according to the optimal matching strategy until the standard performance index is met, and the production efficiency and the product quality of the petroleum proppant are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of petroleum proppant production, and in particular to a method and system for automatically optimizing the raw material ratio of petroleum proppant. Background Art

[0002] In the process of oil extraction, the use of proppants is crucial, as they can help keep the cracks in the oil field open, thereby increasing the production of oil and gas. The performance of proppants depends largely on the ratio of their raw materials, including quartz sand, ceramsite, resin coating and other components. The traditional raw material ratio method relies on empirical formulas and laboratory tests. This method relies heavily on experience and trial and error and may not be optimal, especially when facing new geological conditions or mining technologies. This is not only time-consuming and labor-intensive, but also difficult to achieve the optimal ratio, affecting the performance and cost-effectiveness of the proppant.

[0003] With the development of artificial intelligence technology, especially the advancement of deep learning models and optimization algorithms, modern proppant ratio methods use more technical means to achieve automatic optimization of proppant raw material ratios, thereby improving proppant performance and reducing production costs.

[0004] In oilfield production, high-quality experimental data may be difficult to obtain, and modern proppant ratio methods rely heavily on the amount of available ratio data and performance indicators. Especially for new or complex geological conditions, insufficient and biased data may lead to insufficient model training and affect the accuracy of predictions. Summary of the invention

[0005] The present application aims to solve at least one of the technical problems in the related art to a certain extent. To this end, one purpose of the present application is to propose a method and system for automatically optimizing the ratio of raw materials of petroleum proppants, thereby improving the production efficiency and product quality of petroleum proppants.

[0006] One aspect of the present application provides a method for automatically optimizing the ratio of raw materials of petroleum proppants, comprising:

[0007] Step S100: Collecting geological information of each oil field during historical oil field exploitation, historical proportion data and performance index data of petroleum proppants corresponding to the geology;

[0008] The geological information refers to the geological information of the oil field, including but not limited to: stratigraphic information, pore structure information, and oil and gas characteristic data;

[0009] The ratio data refers to: the ratio information of the raw materials used to produce the petroleum proppant, which affects the performance and quality of the proppant;

[0010] The performance index data refers to: parameters used to measure the quality and effectiveness of proppants, reflecting the performance of proppants in oil and gas well fracturing operations;

[0011] The specific method of collecting the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data and performance index data of the petroleum proppant corresponding to the geology is as follows:

[0012] Step S110: In the historical oilfield exploitation process, using geological exploration equipment to collect geological information of each oilfield;

[0013] Step S120: Access the database to obtain all proppant components of historically produced petroleum proppants to obtain historical ratio data , obtain the performance index data corresponding to each proppant ;

[0014] Among them, λ=1, 2, ..., m, and the sum of the proportions of all the ratio data is 1, that is, ;

[0015] Among them, δ=1, 2, ..., g, and the performance index data contains a total of g performance indicators ;

[0016] Step S200: using geological information and historical matching data to train a conditional generative adversarial network model, and using the conditional generative adversarial network model to obtain first geological information and first matching data;

[0017] The specific method of using geological information and historical matching data to train a conditional generative adversarial network model and using the conditional generative adversarial network model to obtain the first geological information and the first matching data is:

[0018] Step S210: using the existing geological information and historical matching data to train a conditional generative adversarial network model cGAN;

[0019] The specific method of using the existing geological information and historical matching data to train the conditional generative adversarial network model cGAN is:

[0020] Step S211: taking geological information as feature data, taking historical proppant ratio data corresponding to the geological information as label data, and encoding the feature data to obtain a conditional vector;

[0021] Step S212: designing a generator structure of a conditional generative adversarial network model, wherein the network structure of the generator is composed of convolutional layers, each layer uses a ReLU activation function, and the last layer uses a sigmoid function as an activation function. The input of the generator is a conditional vector C and a noise vector z, and the output is the historical ratio data of the proppant;

[0022] The noise vector z is a vector obtained by random sampling, and is used together with the conditional vector as the input of the generator to ensure the diversity of data;

[0023] Step S213: Design a discriminator structure of a conditional generative adversarial network model. The network structure of the discriminator consists of convolutional layers. Each layer uses a LeakyReLU activation function. The last layer is a fully connected layer containing one neuron. The sigmoid function is used as the activation function. The input of the discriminator is the real historical ratio data or the ratio data predicted by the generator. The output of the discriminator is the probability that the input of the discriminator is the real historical ratio data.

[0024] Step S214: Taking the convergence of the loss function as the training goal, setting a loss function change threshold, and stopping the training when the change of the loss function is less than the loss function change threshold;

[0025] The loss function includes the generator loss and the discriminator loss , the generator loss The calculation formula is: , the discriminator loss The calculation formula is: , where x is the real historical ratio data, z is the noise vector, C is the conditional vector, G is the generator, D is the discriminator, and m is the batch size;

[0026] The value of the loss function change threshold is set according to the actual production situation;

[0027] Step S220: acquiring new geological information collected in the new oil field exploitation;

[0028] Step S230: using the trained conditional generative adversarial network model, taking the new geological information as input, and generating new proportion data of the petroleum proppant corresponding to the new geological information conditions;

[0029] The specific method of using the trained conditional generative adversarial network model and taking the new geological information as input to generate new proportion data of petroleum proppants corresponding to the new geological information conditions is as follows:

[0030] Step S231: Encode the new geological information and use the encoded new geological information as a conditional vector;

[0031] Step S232: randomly generate a noise vector, and input the noise vector and the conditional vector as input into the generator of the trained conditional generative adversarial network model;

[0032] Step S233: the generator generates new matching data according to the input noise vector and condition vector and outputs it;

[0033] Step S240: Integrate the historical geological information and historical matching data with the new geological information and new matching data to obtain first geological information and first matching data;

[0034] Step S300: using geological information, historical mix ratio data and performance index data from historically exploited oil fields to train a performance index prediction model;

[0035] Preferably, the performance indicator prediction model uses a neural network model as an initial model;

[0036] The specific method of using geological information, historical proportioning data and performance index data in historically produced oil fields to train a performance index prediction model is as follows:

[0037] Step S310: using geological information and historical proportion data of historically exploited oil fields as feature data, and using performance index data as feature labels corresponding to the feature data;

[0038] Step S320: Selecting a neural network model as an initial model, taking geological information, historical proportion data of petroleum proppants corresponding to the geological information, and performance index data of the historical proportion data under the geological information as a set of training samples;

[0039] Step S330: using geological information of historically exploited oil fields and historical proportioning data as input data for training a performance indicator prediction model, and using the performance indicator data as output data for training the performance indicator prediction model;

[0040] Step S340: taking accurate prediction of the performance indicator data in each group of training samples as the prediction target, taking converging the loss function Loss between the real performance indicator data and the predicted performance indicator data as the training target of the performance indicator prediction model, and training to obtain the performance indicator prediction model;

[0041] The calculation formula of the loss function Loss is: ;

[0042] Where a is the number of training samples, i is a variable, i=1, 2, ..., a, represents the actual performance index data in the i-th training sample, Represents the performance indicator data predicted by the i-th training sample;

[0043] Step S400: taking the first geological information and the first ratio data as input data, and using the performance index prediction model to predict and obtain first performance index data;

[0044] Step S500: setting a standard performance index. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by using an optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant.

[0045] The standard performance indicators are set according to the actual production needs and the experience of professionals;

[0046] The method of setting the standard performance index, if the first performance index data is less than the standard performance index, then adjusting the first ratio data by using the optimal ratio strategy to obtain the optimal ratio data, and if the first performance index data is greater than or equal to the standard performance index, then using the first ratio data as the optimal ratio data to produce the proppant is as follows:

[0047] Step S510: define the objective function , used to evaluate the first performance indicator data of the first ratio data, Indicates the first ratio data;

[0048] Step S520: defining constraint conditions, wherein the constraint conditions include performance index constraint, ratio constraint and sum constraint;

[0049] The expression of the performance indicator constraint is: , where T is the preset standard performance index;

[0050] The expression of the ratio constraint is: ,λ=1,2,...,m;

[0051] The expression of the sum constraint is: ;

[0052] Step S530: Calculate the difference ratio ΔQ between the first performance indicator data and the standard performance indicator. When the difference ratio ΔQ is less than or equal to the threshold value θ, the difference between the first performance indicator data and the standard performance indicator is determined to be a slight difference; when the difference ratio ΔQ is greater than the threshold value θ, the difference between the first performance indicator data and the standard performance indicator is determined to be a significant difference.

[0053] The difference ratio is a value that distinguishes the degree of difference between the first performance indicator data and the standard performance indicator, which is the ratio of the difference between the first performance indicator data and the standard performance indicator to the standard performance indicator, and the calculation formula is: ;

[0054] The selection of the threshold value θ depends on the specific production and the need for adjustment of the proportioning data;

[0055] Step S540: determining whether the gap between the first performance indicator data and the standard performance indicator is a small gap, if so, adjusting the first ratio data using the first adjustment strategy, if not, proceeding to step S550;

[0056] The first adjustment strategy is a linear adjustment strategy, and the expression is: ;

[0057] in, is the adjusted first ratio data, is the adjustment factor, is the difference ratio of the first performance indicator;

[0058] The adjustment factor Used to control the adjustment range and set according to actual production needs;

[0059] The difference ratio of the first performance indicator Used to dynamically adjust the ratio data;

[0060] Step S550: When the gap between the first performance indicator data and the standard performance indicator is significant, the first ratio data is adjusted using the second adjustment strategy;

[0061] The second adjustment strategy is a nonlinear adjustment strategy, and the expression is: ;

[0062] Among them, α is the nonlinear adjustment coefficient, which is used to control the sensitivity of the adjustment and is set according to actual production needs;

[0063] Step S560: The adjusted first ratio data and the first geological information are predicted using the performance index prediction model, and the predicted value is used to update the first performance index data, and the new first performance index data is compared with the standard performance index. If the new first performance index data is less than the standard performance index, return to step S540 to continue to adjust, and if the new first performance index data is greater than or equal to the standard performance index, the corresponding first ratio data is used as the optimal ratio data to produce proppant;

[0064] Step S600: obtaining current actual geological information, generating preliminary matching data, using the preliminary matching data and the actual geological information to obtain predicted performance index data, and adjusting the preliminary matching data according to the optimal matching strategy until the standard performance index is met;

[0065] The specific method of obtaining the current actual geological information, generating preliminary matching data, using the preliminary matching data and the actual geological information to obtain predicted performance index data, and adjusting the preliminary matching data according to the optimal matching strategy until the standard performance index is met is:

[0066] Step S610: using geological exploration equipment to collect actual geological information of the currently exploited oil field, and using a conditional generative adversarial network model to generate preliminary matching data;

[0067] Step S620: Integrate the preliminary ratio data and the actual geological information, input them into the performance index prediction model as input data, and output the predicted performance index data;

[0068] Step S630: If the predicted performance index data does not meet the standard performance index, the preliminary matching data is adjusted according to the optimal matching strategy to obtain the optimal matching data that ultimately meets the standard performance index.

[0069] One aspect of the present application provides an automatic optimization system for petroleum proppant raw material ratio, comprising:

[0070] The data collection module is used to collect the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data of the petroleum proppant corresponding to the geology, and the performance index data;

[0071] A generative adversarial network module is used to train a conditional generative adversarial network model using geological information and historical matching data, and obtain first geological information and first matching data using the conditional generative adversarial network model;

[0072] The performance indicator model training module is used to train the performance indicator prediction model using geological information, historical proportion data and performance indicator data in the historical oil field;

[0073] A performance indicator prediction module is used to use the first geological information and the first ratio data as input data and predict the first performance indicator data using the performance indicator prediction model;

[0074] The optimal ratio data acquisition module is used to set the standard performance index. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by the optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant.

[0075] The practical application module is used to obtain the current actual geological information, generate preliminary proportioning data, use the preliminary proportioning data and the actual geological information to obtain the predicted performance index data, and adjust the preliminary proportioning data according to the optimal proportioning strategy until the standard performance index is met.

[0076] One aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the method for automatically optimizing the ratio of petroleum proppant raw materials are implemented.

[0077] One aspect of the present application provides a readable storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor to execute the steps in a method for automatically optimizing the ratio of petroleum proppant raw materials.

[0078] The method for automatically optimizing the ratio of raw materials of petroleum proppants proposed in this application has the following advantages over the prior art:

[0079] By training the conditional generative adversarial network model, new ratio data under new geological information conditions can be generated, providing more data samples for the subsequent training of performance indicator prediction models;

[0080] By collecting new oilfield geological information and integrating it with historical matching data, new geological information and new matching data are obtained, which provide a basis for performance index prediction and optimization;

[0081] The neural network model is used as the performance indicator prediction model. Compared with the traditional regression method, the neural network model can better fit the complex nonlinear relationship and improve the accuracy of the prediction;

[0082] Different ratio adjustment strategies are adopted, including linear adjustment and nonlinear adjustment, and the adjustment strategy is dynamically selected according to the actual situation. Compared with the traditional fixed adjustment method, it is more flexible and intelligent. At the same time, the concepts of difference ratio and threshold are introduced to dynamically judge the degree of insufficient performance indicators, thereby improving the accuracy and efficiency of ratio adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 A method flow chart of the automatic optimization method for the petroleum proppant raw material ratio provided in this application;

[0084] Figure 2 A schematic diagram of the optimal ratio strategy for the automatic optimization method for the petroleum proppant raw material ratio provided in this application;

[0085] Figure 3 A timing diagram of the automatic optimization method for the petroleum proppant raw material ratio provided in this application;

[0086] Figure 4 Functional module diagram of the automatic optimization system for petroleum proppant raw material ratio provided in this application;

[0087] Figure 5 A schematic diagram of the structure of an electronic device provided in this application;

[0088] Figure 6 It is a schematic diagram of the structure of a readable storage medium provided by this application. DETAILED DESCRIPTION

[0089] In order to better understand the present application, various aspects of the present application will be described in more detail with reference to the accompanying drawings. It should be understood that these detailed descriptions are only descriptions of exemplary embodiments of the present application and do not limit the scope of the present application in any way. Throughout the specification, the same reference numerals refer to the same elements. The expression "and / or" includes any and all combinations of one or more of the associated listed items.

[0090] In the accompanying drawings, the size, dimensions and shapes of the elements have been slightly adjusted for ease of illustration. The drawings are for illustration only and are not strictly drawn to scale. As used herein, the terms "substantially", "approximately" and similar terms are used as terms of approximation, not as terms of degree, and are intended to illustrate the inherent deviations in measurements or calculations that will be recognized by those of ordinary skill in the art. In addition, in this application, the order in which the steps are described does not necessarily represent the order in which these processes occur in actual operation, unless otherwise specified or can be derived from the context.

[0091] It should also be understood that expressions such as "including", "comprising", "having", "containing" and / or "comprising" are open rather than closed expressions in this specification, which indicate the presence of the stated features, elements and / or components, but do not exclude the presence of one or more other features, elements, components and / or combinations thereof. In addition, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire list of features rather than just the individual elements in the list. In addition, when describing embodiments of the present application, "may" is used to mean "one or more embodiments of the present application". And, the term "exemplary" is intended to refer to an example or illustration.

[0092] Unless otherwise defined, all words (including engineering terms and scientific and technological terms) used in this article have the same meaning as those commonly understood by ordinary technicians in the field to which this application belongs. It should also be understood that unless there is a clear explanation in this application, words defined in commonly used dictionaries should be interpreted as having the same meaning as their meaning in the context of the relevant technology, and should not be interpreted in an idealized or overly formal sense.

[0093] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0094] Example 1

[0095] like Figure 1 As shown, the method for automatically optimizing the ratio of raw materials of petroleum proppant provided in this application comprises:

[0096] Step S100: Collecting geological information of each oil field during historical oil field exploitation, historical proportion data and performance index data of petroleum proppants corresponding to the geology;

[0097] The geological information refers to the geological information of the oil field, including but not limited to: stratigraphic information, pore structure information, and oil and gas characteristic data;

[0098] The ratio data refers to: the ratio information of the raw materials used to produce the petroleum proppant, which affects the performance and quality of the proppant;

[0099] The performance index data refers to: parameters used to measure the quality and effectiveness of proppants, reflecting the performance of proppants in oil and gas well fracturing operations;

[0100] The specific method of collecting the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data and performance index data of the petroleum proppant corresponding to the geology is as follows:

[0101] Step S110: In the historical oilfield exploitation process, using geological exploration equipment to collect geological information of each oilfield;

[0102] Step S120: Access the database to obtain all proppant components of historically produced petroleum proppants to obtain historical ratio data , obtain the performance index data corresponding to each proppant ;

[0103] Among them, λ=1, 2, ..., m, and the sum of the proportions of all the ratio data is 1, that is, ;

[0104] Among them, δ=1, 2, ..., g, and the performance index data contains a total of g performance indicators ;

[0105] The database is used to store the ratio data of the petroleum proppant produced in the experiment and production process in the historical time and the performance index data of the corresponding petroleum proppant;

[0106] This step builds a complete data set by collecting geological information, ratio data, and performance index data of historically mined oil fields, providing basic data for subsequent model training;

[0107] Step S200: using geological information and historical matching data to train a conditional generative adversarial network model, and using the conditional generative adversarial network model to obtain first geological information and first matching data;

[0108] The specific method of using geological information and historical matching data to train a conditional generative adversarial network model and using the conditional generative adversarial network model to obtain the first geological information and the first matching data is:

[0109] Step S210: using the existing geological information and historical matching data to train a conditional generative adversarial network model cGAN;

[0110] The specific method of using the existing geological information and historical matching data to train the conditional generative adversarial network model cGAN is:

[0111] Step S211: taking geological information as feature data, taking historical proppant ratio data corresponding to the geological information as label data, and encoding the feature data to obtain a conditional vector;

[0112] Step S212: designing a generator structure of a conditional generative adversarial network model, wherein the network structure of the generator is composed of convolutional layers, each layer uses a ReLU activation function, and the last layer uses a sigmoid function as an activation function. The input of the generator is a conditional vector C and a noise vector z, and the output is the historical ratio data of the proppant;

[0113] The noise vector z is a vector obtained by random sampling, and is used together with the conditional vector as the input of the generator to ensure the diversity of data;

[0114] Step S213: Design a discriminator structure of a conditional generative adversarial network model. The network structure of the discriminator consists of convolutional layers. Each layer uses a LeakyReLU activation function. The last layer is a fully connected layer containing one neuron. The sigmoid function is used as the activation function. The input of the discriminator is the real historical ratio data or the ratio data predicted by the generator. The output of the discriminator is the probability that the input of the discriminator is the real historical ratio data.

[0115] Step S214: Taking the convergence of the loss function as the training goal, setting a loss function change threshold, and stopping the training when the change of the loss function is less than the loss function change threshold;

[0116] The loss function includes the generator loss and the discriminator loss , the generator loss The calculation formula is: , the discriminator loss The calculation formula is: , where x is the real historical ratio data, z is the noise vector, C is the conditional vector, G is the generator, D is the discriminator, and m is the batch size;

[0117] The batch size refers to the amount of data processed in one training cycle;

[0118] The value of the loss function change threshold is set according to the actual production situation;

[0119] Step S220: acquiring new geological information collected in the new oil field exploitation;

[0120] Step S230: using the trained conditional generative adversarial network model, taking the new geological information as input, and generating new proportion data of the petroleum proppant corresponding to the new geological information conditions;

[0121] The specific method of using the trained conditional generative adversarial network model and taking the new geological information as input to generate new proportion data of petroleum proppants corresponding to the new geological information conditions is as follows:

[0122] Step S231: Encode the new geological information and use the encoded new geological information as a conditional vector;

[0123] Step S232: randomly generate a noise vector, and input the noise vector and the conditional vector as input into the generator of the trained conditional generative adversarial network model;

[0124] Step S233: the generator generates new matching data according to the input noise vector and condition vector and outputs it;

[0125] Step S240: Integrate the historical geological information and historical matching data with the new geological information and new matching data to obtain first geological information and first matching data;

[0126] This step generates an adversarial network model through training conditions, which can generate new ratio data under new geological information conditions, providing more possibilities for proppant ratio and more data samples for the training of subsequent performance indicator prediction models;

[0127] Step S300: using geological information, historical mix ratio data and performance index data from historically exploited oil fields to train a performance index prediction model;

[0128] The performance index prediction model is a deep learning model that predicts the performance index data of the petroleum proppant produced under the ratio data based on the geological information of the oil field and the ratio data of the petroleum proppant corresponding to the geology;

[0129] Preferably, the performance indicator prediction model uses a neural network model as an initial model;

[0130] The specific method of using geological information, historical proportioning data and performance index data in historically produced oil fields to train a performance index prediction model is as follows:

[0131] Step S310: using geological information and historical proportion data of historically exploited oil fields as feature data, and using performance index data as feature labels corresponding to the feature data;

[0132] Step S320: Selecting a neural network model as an initial model, taking geological information, historical proportion data of petroleum proppants corresponding to the geological information, and performance index data of the historical proportion data under the geological information as a set of training samples;

[0133] Step S330: using geological information of historically exploited oil fields and historical proportioning data as input data for training a performance indicator prediction model, and using the performance indicator data as output data for training the performance indicator prediction model;

[0134] Step S340: taking accurate prediction of the performance indicator data in each group of training samples as the prediction target, taking converging the loss function Loss between the real performance indicator data and the predicted performance indicator data as the training target of the performance indicator prediction model, and training to obtain the performance indicator prediction model;

[0135] The calculation formula of the loss function Loss is: ;

[0136] Where a is the number of training samples, i is a variable, i=1, 2, ..., a, represents the actual performance index data in the i-th training sample, Represents the performance indicator data predicted by the i-th training sample;

[0137] This step uses historical data to train a performance index prediction model, which can predict the performance index of proppant application based on the ratio data and geological information, providing a basis for optimizing the ratio. A neural network model is used as the performance index prediction model. Compared with the traditional regression method, the neural network model can better fit complex nonlinear relationships and improve the accuracy of prediction.

[0138] Step S400: taking the first geological information and the first ratio data as input data, and using the performance index prediction model to predict and obtain first performance index data;

[0139] The specific method of using the first geological information and the first ratio data as input data and using the performance index prediction model to predict the first performance index data is:

[0140] The first geological information and the first ratio data are standardized, the trained performance indicator prediction model is loaded, and the standardized first geological information and the first ratio information are input into the performance indicator prediction model as input data, and the performance indicator prediction model outputs the corresponding first performance indicator data;

[0141] Step S500: setting a standard performance index. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by using an optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant.

[0142] The optimal matching strategy includes a first adjustment strategy and a second adjustment strategy;

[0143] The standard performance indicators are set according to the actual production needs and the experience of professionals;

[0144] Figure 2 A schematic diagram of the optimal ratio strategy for the automatic optimization method for the petroleum proppant raw material ratio provided in this application;

[0145] The method of setting the standard performance index, if the first performance index data is less than the standard performance index, then adjusting the first ratio data by using the optimal ratio strategy to obtain the optimal ratio data, and if the first performance index data is greater than or equal to the standard performance index, then using the first ratio data as the optimal ratio data to produce the proppant is as follows:

[0146] Step S510: define the objective function , used to evaluate the first performance indicator data of the first ratio data, Indicates the first ratio data;

[0147] The optimization goal of the objective function is to maximize the value of the objective function;

[0148] Step S520: defining constraint conditions, wherein the constraint conditions include performance index constraint, ratio constraint and sum constraint;

[0149] The expression of the performance indicator constraint is: , where T is the preset standard performance index;

[0150] The expression of the ratio constraint is: ,λ=1,2,...,m;

[0151] The expression of the sum constraint is: ;

[0152] Step S530: Calculate the difference ratio ΔQ between the first performance indicator data and the standard performance indicator. When the difference ratio ΔQ is less than or equal to the threshold value θ, the difference between the first performance indicator data and the standard performance indicator is determined to be a slight difference; when the difference ratio ΔQ is greater than the threshold value θ, the difference between the first performance indicator data and the standard performance indicator is determined to be a significant difference.

[0153] The difference ratio is a value that distinguishes the degree of difference between the first performance indicator data and the standard performance indicator, which is the ratio of the difference between the first performance indicator data and the standard performance indicator to the standard performance indicator, and the calculation formula is: ;

[0154] The selection of the threshold value θ depends on the specific production and the need for adjustment of the proportioning data;

[0155] Assume that, through expert opinions and historical data analysis, it is determined that when the first performance indicator data is less than or equal to 5% of the standard performance indicator, it is considered a slight gap, and when the first performance indicator data exceeds 5% of the standard performance indicator, it is considered a significant gap, so the threshold θ is set to 0.05;

[0156] Step S540: determining whether the gap between the first performance indicator data and the standard performance indicator is a small gap, if so, adjusting the first ratio data using the first adjustment strategy, if not, proceeding to step S550;

[0157] The first adjustment strategy is a linear adjustment strategy, and the expression is: ;

[0158] in, is the adjusted first ratio data, is the adjustment factor, is the difference ratio of the first performance indicator;

[0159] The adjustment factor Used to control the adjustment range and set according to actual production needs;

[0160] The difference ratio of the first performance indicator Used to dynamically adjust the ratio data;

[0161] Step S550: When the gap between the first performance indicator data and the standard performance indicator is significant, the first ratio data is adjusted using the second adjustment strategy;

[0162] The second adjustment strategy is a nonlinear adjustment strategy, and the expression is: ;

[0163] Among them, α is the nonlinear adjustment coefficient, which is used to control the sensitivity of the adjustment and is set according to actual production needs;

[0164] Step S560: The adjusted first ratio data and the first geological information are predicted using the performance index prediction model, and the predicted value is used to update the first performance index data, and the new first performance index data is compared with the standard performance index. If the new first performance index data is less than the standard performance index, return to step S540 to continue to adjust, and if the new first performance index data is greater than or equal to the standard performance index, the corresponding first ratio data is used as the optimal ratio data to produce proppant;

[0165] When the first performance index data is greater than or equal to the standard performance index, the first ratio data corresponding to the first performance index data at this time meets the standard performance index required by the petroleum proppant, so the first ratio data corresponding to the first performance index data is used as the optimal ratio data to produce the petroleum proppant;

[0166] This step adopts different ratio adjustment strategies based on the comparison between the predicted performance index data and the standard performance index to ensure that the proppant ratio reaches the required performance level. Two different adjustment strategies are proposed, including linear adjustment and nonlinear adjustment. The adjustment strategy is dynamically selected according to the actual situation. Compared with the traditional fixed adjustment method, it is more flexible and intelligent. At the same time, the concepts of difference ratio and threshold are introduced to dynamically judge the degree of insufficient performance indicators, thereby improving the accuracy and efficiency of ratio adjustment.

[0167] Step S600: obtaining current actual geological information, generating preliminary matching data, using the preliminary matching data and the actual geological information to obtain predicted performance index data, and adjusting the preliminary matching data according to the optimal matching strategy until the standard performance index is met;

[0168] The specific method of obtaining the current actual geological information, generating preliminary matching data, using the preliminary matching data and the actual geological information to obtain predicted performance index data, and adjusting the preliminary matching data according to the optimal matching strategy until the standard performance index is met is:

[0169] Step S610: using geological exploration equipment to collect actual geological information of the currently exploited oil field, and using a conditional generative adversarial network model to generate preliminary matching data;

[0170] Step S620: Integrate the preliminary ratio data and the actual geological information, input them into the performance index prediction model as input data, and output the predicted performance index data;

[0171] Step S630: If the predicted performance index data does not meet the standard performance index, the preliminary matching data is adjusted according to the optimal matching strategy to obtain the optimal matching data that ultimately meets the standard performance index.

[0172] Example 2

[0173] like Figure 3 As shown, it is a timing diagram of the method for automatically optimizing the ratio of petroleum proppant raw materials provided in this application, including:

[0174] First, data obtained from historical oilfield production are collected, including geological information, historical mix ratio data, and performance index data;

[0175] The collected geological information and historical ratio data are used as training samples to train the conditional generative adversarial network model;

[0176] Acquire new geological information in the exploitation of new oil fields, use the trained conditional generative adversarial network model, take the new geological information as input, generate new matching data, integrate historical data and new data, and obtain first geological information and first matching data;

[0177] The collected geological information, historical ratio data and performance index data are used as training data to train the performance index prediction model;

[0178] Taking the first geological information and the first ratio data as input, inputting into the performance index prediction model, and obtaining the first performance index data;

[0179] Compare the first performance index data with the standard performance index, formulate an optimal ratio strategy to adjust the first ratio data to obtain the optimal ratio data;

[0180] Collect real-time geological information data and use the conditional generative adversarial network model to generate preliminary matching data;

[0181] The performance indicator prediction model is used to predict the performance indicator data with preliminary ratio data and real-time geological information as input data. The preliminary ratio data is adjusted according to the optimal ratio strategy until the performance indicator data meets the standard performance indicator. The final ratio data is used as the optimal ratio data to produce petroleum proppants.

[0182] Example 3

[0183] like Figure 4 As shown, the petroleum proppant raw material ratio automatic optimization system provided in this application includes:

[0184] The data collection module is used to collect the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data of the petroleum proppant corresponding to the geology, and the performance index data;

[0185] A generative adversarial network module is used to train a conditional generative adversarial network model using geological information and historical matching data, and obtain first geological information and first matching data using the conditional generative adversarial network model;

[0186] The performance indicator model training module is used to train the performance indicator prediction model using geological information, historical proportion data and performance indicator data in the historical oil field;

[0187] A performance indicator prediction module is used to use the first geological information and the first ratio data as input data and predict the first performance indicator data using the performance indicator prediction model;

[0188] The optimal ratio data acquisition module is used to set the standard performance index. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by the optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant.

[0189] The practical application module is used to obtain the current actual geological information, generate preliminary proportioning data, use the preliminary proportioning data and the actual geological information to obtain the predicted performance index data, and adjust the preliminary proportioning data according to the optimal proportioning strategy until the standard performance index is met.

[0190] Example 4

[0191] Figure 5 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 5 As shown, according to another aspect of the present application, an electronic device is also provided. The electronic device may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are run by one or more processors, the above-mentioned method for automatically optimizing the ratio of petroleum proppant raw materials can be executed.

[0192] The method or system according to the embodiment of the present application can also be used Figure 5 The electronic device architecture shown in FIG. Figure 5As shown, the electronic device may include a bus, one or more CPUs, a read-only memory (ROM), a random access memory (RAM), a communication port connected to a network, an input / output component, a hard disk, etc. A storage device in the electronic device, such as a ROM or a hard disk, may store the method for automatically optimizing the ratio of petroleum proppant raw materials provided in the present application. The method for automatically optimizing the raw material ratio of petroleum proppants may, for example, include: collecting geological information of each oil field during the historical exploitation of the oil field, historical ratio data and performance index data of the petroleum proppants corresponding to the geology; using the geological information and historical ratio data to train a conditional generative adversarial network model, and using the conditional generative adversarial network model to obtain first geological information and first ratio data; using the geological information, historical ratio data and performance index data in the historical exploitation of the oil field to train a performance index prediction model; using the first geological information and the first ratio data as input data, and using the performance index prediction model to predict the first performance index data; setting a standard performance index, if the first performance index data is less than the standard performance index, then using the optimal ratio strategy to adjust the first ratio data to obtain the optimal ratio data, if the first performance index data is greater than or equal to the standard performance index, then using the first ratio data as the optimal ratio data to produce the proppant; obtaining the current actual geological information, generating preliminary ratio data, using the preliminary ratio data and the actual geological information to obtain the predicted performance index data, and adjusting the preliminary ratio data according to the optimal ratio strategy until the standard performance index is met. Furthermore, the electronic device may also include a user interface. Of course, Figure 5 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 5 One or more components of an electronic device are shown.

[0193] Example 5

[0194] Figure 6 Schematic diagram of a readable storage medium structure provided by an embodiment of the present application. Figure 6 As shown, it is a readable storage medium according to an embodiment of the present application. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by the processor, the automatic optimization method of the petroleum proppant raw material ratio according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache). Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0195] In addition, according to the implementation mode of the present application, the process described in the above reference flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores machine-readable instructions, the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided in the present application, such as: collecting geological information of each oil field during the historical exploitation of the oil field, historical proportioning data and performance index data of the petroleum proppant corresponding to the geology; using the geological information and historical proportioning data to train a conditional generative adversarial network model, and using the conditional generative adversarial network model to obtain the first geological information and the first proportioning data; using the geological information, historical proportioning data and performance index data in the historical exploitation of the oil field to train the performance index prediction The method comprises the following steps: a) using the first geological information and the first ratio data as input data, and using the performance index prediction model to predict the first performance index data; b) setting a standard performance index, and if the first performance index data is less than the standard performance index, then using the optimal ratio strategy to adjust the first ratio data to obtain the optimal ratio data; and c) using the first performance index data as the optimal ratio data to produce the proppant; and e) obtaining the current actual geological information, generating preliminary ratio data, and using the preliminary ratio data and the actual geological information to obtain the predicted performance index data, and adjusting the preliminary ratio data according to the optimal ratio strategy until the standard performance index is met. When the computer program is executed by the central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0196] The methods, apparatuses, and devices of the present application may be implemented in many ways. For example, the methods, apparatuses, and devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified. In addition, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, which include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers recording media storing programs for executing the method according to the present application.

[0197] In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0198] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. Automatic optimization method for raw material ratio of petroleum proppant, It is characterized in that include: Collect geological information of each oil field during historical oil field exploitation, historical proportion data and performance index data of petroleum proppants corresponding to the geology; Using geological information and historical matching data to train a conditional generative adversarial network model, and using the conditional generative adversarial network model to obtain first geological information and first matching data; The performance index prediction model is obtained by training the geological information, historical proportion data and performance index data of the historically exploited oil fields; The first geological information and the first ratio data are used as input data, and the first performance indicator data is predicted by using the performance indicator prediction model; A standard performance index is set. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by using an optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant. The optimal matching strategy includes a first adjustment strategy and a second adjustment strategy; Obtain current actual geological information, generate preliminary ratio data, use the preliminary ratio data and actual geological information to obtain predicted performance index data, and adjust the preliminary ratio data according to the optimal ratio strategy until the standard performance index is met; The method of setting the standard performance index, if the first performance index data is less than the standard performance index, then adjusting the first ratio data by using the optimal ratio strategy to obtain the optimal ratio data, and if the first performance index data is greater than or equal to the standard performance index, then using the first ratio data as the optimal ratio data to produce the proppant is as follows: Step S510: define the objective function , used to evaluate the first performance indicator data of the first ratio data, Indicates the first ratio data; Step S520: defining constraint conditions, wherein the constraint conditions include performance index constraint, ratio constraint and sum constraint; Step S530: Calculate the difference ratio ΔQ between the first performance indicator data and the standard performance indicator. When the difference ratio ΔQ is less than or equal to a threshold value θ, determine the difference between the first performance indicator data and the standard performance indicator as a small difference. When the difference ratio ΔQ is greater than a threshold value θ, the gap between the first performance indicator data and the standard performance indicator is determined to be a significant gap; Step S540: determining whether the gap between the first performance indicator data and the standard performance indicator is a small gap, if so, adjusting the first ratio data using the first adjustment strategy, if not, proceeding to step S550; Step S550: When the gap between the first performance indicator data and the standard performance indicator is significant, the first ratio data is adjusted using the second adjustment strategy; Step S560: The adjusted first ratio data and the first geological information are predicted using a performance indicator prediction model. The predicted value is used to update the first performance indicator data. The new first performance indicator data is compared with the standard performance indicator. If the new first performance indicator data is less than the standard performance indicator, return to step S540 to continue adjusting. If the new first performance indicator data is greater than or equal to the standard performance indicator, the corresponding first ratio data is used as the optimal ratio data to produce proppant.

2. The method for automatically optimizing the ratio of petroleum proppant raw materials according to claim 1, It is characterized in that The specific method of collecting the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data and performance index data of the petroleum proppant corresponding to the geology is as follows: During the historical oil field exploitation process, geological exploration equipment was used to collect geological information on each oil field; Access the database to obtain all proppant components of historically produced petroleum proppants and obtain historical ratio data , obtain the performance index data corresponding to each proppant ; Among them, λ=1, 2, ..., m, and the sum of the proportions of all the ratio data is 1, that is, ; Among them, δ=1, 2, ..., g, and the performance index data contains a total of g performance indicators .

3. The method for automatically optimizing the ratio of petroleum proppant raw materials according to claim 2, It is characterized in that The specific method of using geological information and historical matching data to train a conditional generative adversarial network model and using the conditional generative adversarial network model to obtain the first geological information and the first matching data is: The conditional generative adversarial network model cGAN is trained using existing geological information and historical matching data; Access to new geological information collected in new oil field development; Using the trained conditional generative adversarial network model, the new geological information is used as input to generate new proportion data of petroleum proppants corresponding to the new geological information conditions; The geological information and historical proportion data of historical time are integrated with the new geological information and new proportion data to obtain the first geological information and the first proportion data.

4. The method for automatically optimizing the ratio of raw materials of petroleum proppants according to claim 3, It is characterized in that The specific method of using the existing geological information and historical matching data to train the conditional generative adversarial network model cGAN is: Taking geological information as feature data, taking historical proppant ratio data corresponding to the geological information as label data, and encoding the feature data to obtain a conditional vector; Design the generator structure of the conditional generative adversarial network model. The network structure of the generator consists of convolutional layers. Each layer uses the ReLU activation function. The last layer uses the sigmoid function as the activation function. The input of the generator is the condition vector C and the noise vector z, and the output is the historical ratio data of the proppant. Design the discriminator structure of the conditional generative adversarial network model. The network structure of the discriminator consists of convolutional layers. Each layer uses the LeakyReLU activation function. The last layer is a fully connected layer containing one neuron. The sigmoid function is used as the activation function. The input of the discriminator is the real historical ratio data or the ratio data predicted by the generator. The output of the discriminator is the probability that the input of the discriminator is the real historical ratio data. The training goal is to achieve convergence of the loss function, and a loss function change threshold is set. When the change of the loss function is less than the loss function change threshold, the training is stopped. The loss function includes the generator loss and the discriminator loss , the generator loss The calculation formula is: , the discriminator loss The calculation formula is: , where x is the real historical ratio data, z is the noise vector, C is the conditional vector, G is the generator, D is the discriminator, and m is the batch size.

5. The method for automatically optimizing the ratio of petroleum proppant raw materials according to claim 4, It is characterized in that The specific method of using geological information, historical proportioning data and performance index data in historically produced oil fields to train a performance index prediction model is as follows: The geological information and historical proportion data of the historically exploited oil fields are used as feature data, and the performance index data are used as feature labels of the corresponding feature data; Selecting a neural network model as an initial model, taking geological information, historical proportion data of petroleum proppants corresponding to the geological information, and performance index data of the historical proportion data under the geological information as a group of training samples; Using geological information and historical proportion data of historically exploited oil fields as input data for training a performance indicator prediction model, and using the performance indicator data as output data for training the performance indicator prediction model; The prediction target is to accurately predict the performance indicator data in each group of training samples, and the training target of the performance indicator prediction model is to converge the loss function Loss between the real performance indicator data and the predicted performance indicator data, and the performance indicator prediction model is obtained by training; The calculation formula of the loss function Loss is: ; Where a is the number of training samples, i is a variable, i=1, 2, ..., a, represents the actual performance index data in the i-th training sample, Represents the performance indicator data predicted by the i-th training sample.

6. The method for automatically optimizing the ratio of raw materials of petroleum proppant according to claim 5, It is characterized in that The first adjustment strategy is a linear adjustment strategy, and the expression is: ; in, is the adjusted first ratio data, is the adjustment factor, is the difference ratio of the first performance indicator.

7. The method for automatically optimizing the ratio of raw materials of petroleum proppant according to claim 6, It is characterized in that The second adjustment strategy is a nonlinear adjustment strategy, and the expression is: ; Among them, α is the nonlinear adjustment coefficient.

8. A system for automatically optimizing the ratio of raw materials of petroleum proppants, which is implemented based on the method for automatically optimizing the ratio of raw materials of petroleum proppants according to any one of claims 1 to 7. It is characterized in that include: The data collection module is used to collect the geological information of each oil field during the historical exploitation of the oil field, the historical proportion data of the petroleum proppant corresponding to the geology, and the performance index data; A generative adversarial network module is used to train a conditional generative adversarial network model using geological information and historical matching data, and obtain first geological information and first matching data using the conditional generative adversarial network model; The performance indicator model training module is used to train the performance indicator prediction model using geological information, historical proportion data and performance indicator data in the historical oil field; A performance indicator prediction module is used to use the first geological information and the first ratio data as input data and predict the first performance indicator data using the performance indicator prediction model; The optimal ratio data acquisition module is used to set the standard performance index. If the first performance index data is less than the standard performance index, the first ratio data is adjusted by the optimal ratio strategy to obtain the optimal ratio data. If the first performance index data is greater than or equal to the standard performance index, the first ratio data is used as the optimal ratio data to produce the proppant. The practical application module is used to obtain the current actual geological information, generate preliminary proportioning data, use the preliminary proportioning data and the actual geological information to obtain the predicted performance index data, and adjust the preliminary proportioning data according to the optimal proportioning strategy until the standard performance index is met.

9. An electronic device, It is characterized in that The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the method for automatically optimizing the ratio of petroleum proppant raw materials as claimed in any one of claims 1 to 7 are implemented.

10. A readable storage medium, It is characterized in that The readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the steps in the method for automatically optimizing the ratio of petroleum proppant raw materials as described in any one of claims 1-7.

Citation Information

Patent Citations

  • Preparation process of petroleum fracturing propping agent based on utilization of tailings

    CN113563866A

  • Proppant determination method and device, electronic equipment and storage medium

    CN114136845A

  • Sand paving test system and optimization method of proppant

    CN115559697A

  • Simulations for Hydraulic Fracturing Treatments and Methods of Fracturing Naturally Fractured Formation

    US20080183451A1

  • Calculation systems for predicting proppant embedding depth based on shale softening effect

    US20240133281A1