Method, device and equipment for determining fracturing operation parameters

Abstract

The application discloses a method, device and equipment for determining fracturing construction parameters, wherein the method comprises the following steps: acquiring historical data of a plurality of historical production wells; performing data processing on the historical data, and determining parameters affecting productivity in the historical data as productivity main control parameters; performing data expansion on the historical data based on the productivity main control parameters to obtain expanded data; training a machine learning model based on the expanded data to obtain a productivity determination model taking data corresponding to the productivity main control parameters as input and taking production data as output; acquiring target geological data corresponding to a target production well, constructing a parameter optimization model based on the productivity determination model and the target geological data, and performing optimization solving on the parameter optimization model to obtain target fracturing construction parameters corresponding to the target production well. Through the method, the target fracturing construction parameters suitable for a target region can be quickly, accurately and reliably obtained, and the production capacity of the production well is maximized.

Description

Technical Field

[0001] This application relates to the field of oil and gas technology, and in particular to a method, apparatus and equipment for determining fracturing construction parameters. Background Technology

[0002] With the development of oil and gas resource development, unconventional oil and gas resources, such as deep coalbed methane and shale gas, have become an important direction for energy exploration. Unconventional oil reservoirs generally have characteristics such as low permeability, strong heterogeneity, and complex stress distribution, requiring fracturing to form an effective fracture network to improve production capacity. Currently, the optimization of fracturing operation parameters mainly relies on theoretical models and empirical formulas, as well as artificial intelligence (AI) algorithms. Compared with methods relying on theoretical models and empirical formulas, such as fracture propagation simulation and fracture network geometry optimization based on rock mechanics assumptions, optimizing fracturing operation parameters through AI algorithms can guide fracturing scheme design and the entire life cycle production management of coalbed methane wells. This can provide a scientific basis and technical support for realizing integrated geological engineering and cost-effective development of coalbed methane, while also promoting the construction of smart oil and gas fields.

[0003] Current AI algorithms generally optimize fracturing parameters by introducing machine learning techniques, but there is still a problem that the selection of model input parameters depends on human experience, making it impossible to accurately screen the main factors controlling production capacity; moreover, the algorithm's generalization ability is insufficient, making it difficult to adapt to the complex geological conditions of deep coal seams, and unable to achieve accurate and reliable optimization of fracturing construction parameters.

[0004] There is currently no effective solution to the problem of low accuracy and reliability of fracturing parameters in the aforementioned unconventional oil reservoirs. Summary of the Invention

[0005] The purpose of this specification is to provide a method, apparatus, and equipment for determining fracturing operation parameters, in order to solve the problem of low accuracy and reliability of fracturing operation parameters in unconventional oil reservoirs.

[0006] To solve the above-mentioned technical problems, the first aspect of this specification provides a method for determining fracturing construction parameters, including:

[0007] Acquire historical data from multiple historical production wells, including historical geological data, historical engineering data, and historical production data;

[0008] The historical data is processed to determine the parameters that affect production capacity from among the multiple parameters corresponding to the historical data, which are then used as the main production capacity control parameters.

[0009] Based on the aforementioned capacity control parameters, the historical data is augmented to obtain augmented data.

[0010] The pre-built machine learning model is trained based on the expanded data to obtain a capacity determination model that takes the data corresponding to the main capacity control parameters as input and the production data as output.

[0011] Obtain the target geological data corresponding to the target production well, construct a parameter optimization model based on the production capacity determination model and the target geological data, and optimize and solve the parameter optimization model to obtain the target fracturing construction parameters corresponding to the target production well.

[0012] In some embodiments of this specification, the historical data is processed to determine the parameters affecting production capacity among multiple parameters corresponding to the historical data as the main production capacity control parameters, including:

[0013] Construct at least two feature selection models, with different feature selection algorithms used for different feature selection models;

[0014] The historical data is input into each feature selection model to obtain the output results of each feature selection model;

[0015] Construct an evaluation matrix based on the outputs of at least two feature selection models;

[0016] Based on the evaluation matrix, at least one parameter is determined from multiple parameters as the main control parameter for production capacity.

[0017] In some embodiments of this specification, the at least two feature selection models include at least: a grey relational analysis model, a principal component coefficient calculation model, and a machine learning model.

[0018] In some embodiments of this specification, the first dimension of the evaluation matrix corresponds to multiple parameters, and the second dimension corresponds to at least two feature selection models;

[0019] Accordingly, based on the evaluation matrix, at least one parameter is determined from multiple parameters as the main control parameter for production capacity, including:

[0020] The evaluation matrix is ​​normalized.

[0021] Based on the normalized evaluation matrix, the information entropy corresponding to each feature selection model in the evaluation matrix is ​​calculated.

[0022] The weights of each feature selection model are determined based on the information entropy corresponding to each feature selection model.

[0023] Based on the weights corresponding to each feature selection model and the normalized evaluation matrix, the comprehensive score of each parameter is determined.

[0024] Based on the comprehensive score of each parameter and the preset feature selection conditions, at least one parameter is selected from multiple parameters as the main control parameter for production capacity.

[0025] In some embodiments of this specification, the historical data is augmented based on the main capacity control parameters to obtain augmented data, including:

[0026] The target data is obtained by performing dimensionality reduction processing on the historical data based on the aforementioned capacity control parameters.

[0027] Determine the statistical characteristics and parameter dependency characteristics of the target data;

[0028] Based on the statistical features and the parameter dependence features, synthetic data similar to the target data is generated, and the synthetic data and the target data are used as augmented data.

[0029] In some embodiments of this specification, a pre-built machine learning model is trained based on the expanded data to obtain a capacity determination model that takes data corresponding to the main capacity control parameters as input and production data as output, including:

[0030] A machine learning model is constructed, which includes multiple sub-models that take data corresponding to the main control parameters of production capacity as input and production data as output. Different sub-models adopt different types of machine learning algorithms.

[0031] The augmented data is divided into a training set and a test set;

[0032] Multiple sub-models are trained based on the training set to obtain multiple capacity determination sub-models;

[0033] Based on the test set and the model accuracy of each capacity determination sub-model, a capacity determination model is selected from the multiple capacity determination sub-models.

[0034] In some embodiments of this specification, a parameter optimization model is constructed based on the production capacity determination model and the target geological data, and the parameter optimization model is optimized and solved to obtain the target fracturing construction parameters corresponding to the target production well, including:

[0035] Using the production capacity determination model as the objective function and the target geological data as the constraint, and with the goal of maximizing production capacity, a parameter optimization model including the objective function and the constraint is constructed.

[0036] Determine the value range corresponding to the fracturing operation parameters, and initialize the fracturing operation parameters based on the value range to obtain initial fracturing operation data;

[0037] Input the initial fracturing operation data into the objective function to calculate the production capacity calculation result corresponding to the initial fracturing operation data;

[0038] Based on the production capacity calculation results and the constraints, the initial fracturing construction data is iteratively optimized until the preset optimization conditions are met. The value of the fracturing construction parameter corresponding to the largest production capacity calculation result during the iterative optimization process is taken as the target fracturing construction parameter.

[0039] The second aspect of this specification provides a device for determining fracturing operation parameters, comprising:

[0040] The acquisition module is used to acquire historical data from multiple historical production wells, including historical geological data, historical engineering data, and historical production data.

[0041] The processing module is used to process the historical data and determine the parameters that affect production capacity among the multiple parameters corresponding to the historical data as the main production capacity control parameters.

[0042] An expansion module is used to expand the historical data based on the main control parameters of the production capacity to obtain expanded data;

[0043] The training module is used to train a pre-built machine learning model based on the expanded data to obtain a capacity determination model that takes the data corresponding to the main control parameters of capacity as input and the production data as output.

[0044] The optimization module is used to acquire the target geological data corresponding to the target production well, construct a parameter optimization model based on the production capacity determination model and the target geological data, and optimize and solve the parameter optimization model to obtain the target fracturing construction parameters corresponding to the target production well.

[0045] A third aspect of this specification provides an electronic device, comprising: a memory and a processor, the processor and the memory being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to implement the steps of the method described in the first aspect.

[0046] A fourth aspect of this specification provides a computer-readable storage medium storing computer program instructions that, when executed by a processor, implement the steps of the method described in the first aspect.

[0047] The method, apparatus, and equipment for determining fracturing construction parameters provided in the embodiments of this specification involve: acquiring historical data from multiple historical production wells, including historical geological data, historical engineering data, and historical production data; processing the historical data to determine the parameters affecting production capacity among the multiple parameters corresponding to the historical data as the main production capacity control parameters; expanding the historical data based on the main production capacity control parameters to obtain expanded data; training a machine learning model based on the expanded data to obtain a production capacity determination model with the data corresponding to the main production capacity control parameters as input and production data as output; acquiring the target geological data corresponding to the target production well; constructing a parameter optimization model based on the production capacity determination model and the target geological data; and optimizing and solving the parameter optimization model to obtain the target fracturing construction parameters corresponding to the target production well. By analyzing historical data of the target area using the above method, the main control parameters affecting production capacity can be determined. The historical data can be expanded for training subsequent machine learning models, improving the prediction accuracy of the production capacity determination model. Based on the trained production capacity calculation model and historical geological data, the fracturing construction parameters can be optimized. This allows for the rapid, accurate, and reliable acquisition of target fracturing construction parameters suitable for the target area, maximizing the production capacity of production wells. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 The diagram shown is a schematic representation of a method for determining fracturing construction parameters provided in an embodiment of this specification.

[0050] Figure 2 The diagram shown is a schematic representation of an optimization method for deep coal seam fracturing construction parameters based on physical constraints and data-driven approaches, as provided in an embodiment of this specification.

[0051] Figure 3 The diagram shown is a schematic representation of the comprehensive score of multiple parameters provided in the embodiments of this specification.

[0052] Figure 4 The diagram shown is a schematic of a fracturing parameter optimization method provided in an embodiment of this specification.

[0053] Figure 5 The diagram shown is a schematic of a device for determining fracturing construction parameters provided in an embodiment of this specification.

[0054] Figure 6The diagram shown is a schematic of an electronic device provided in an embodiment of this specification. Detailed Implementation

[0055] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0056] The method for determining fracturing construction parameters provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0057] Figure 1 The diagram illustrates a method for determining fracturing construction parameters provided in an embodiment of this specification. While this specification provides method operation steps or apparatus structures as shown in the following embodiments or figures, based on conventional or non-inventive methods, the method or apparatus may include more or fewer operation steps or module units, either combined or integrated. In steps or structures where there is no logically necessary causal relationship, the execution order of these steps or the module structure of the apparatus is not limited to the execution order or module structure shown in the embodiments or figures of this specification. When the method or module structure is applied in actual devices, servers, or terminal products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or figures (e.g., in a parallel processor or multi-threaded processing environment, or even in a distributed processing or server cluster implementation environment). Figure 1 As shown, the method may include:

[0058] S101: Obtain historical data from multiple historical production wells, including historical geological data, historical engineering data, and historical production data.

[0059] It is understandable that historical geological parameters can be used to characterize the geological features of the area where historical production wells are located, historical engineering data can characterize the engineering data used during fracturing operations on historical production wells, and historical production data can include production data reflecting the production capacity of historical production wells after fracturing operations are completed. Specifically, historical geological data can be obtained by processing seismic and well logging data of the area where historical production wells are located. For example, the geological parameters corresponding to historical geological data can include formation lithology, coal seam thickness, coal quality characteristics, roof and floor lithology, porosity, permeability, etc.; the fracturing operation parameters corresponding to historical engineering data can include sand addition scale, displacement, whether temporary plugging is required, fracturing technology, etc.; the production parameters corresponding to historical production data can include cumulative gas production, average daily gas production, peak gas production, etc. within the previous preset time period.

[0060] In some embodiments of this specification, multiple historical production wells can be production wells completed in the same region or in different regions. It is understood that different regions have different geological characteristics, meaning their geological data characteristics differ, and the expanded data obtained based on these regions will also differ. Therefore, the production capacity determination model trained on these regions is more suitable for optimizing fracturing parameters for production wells in regions with similar geological characteristics. Furthermore, when multiple historical production wells are completed in the same region, the subsequently trained production capacity determination model can be used to optimize fracturing parameters for target production wells within that region, resulting in more accurate and reliable optimization results. When multiple historical production wells are completed in different regions, the subsequently trained production capacity determination model can be used to accurately and reliably optimize fracturing parameters for target production wells in multiple different regions.

[0061] S102: Perform data processing on the historical data to determine the parameter that affects production capacity among the multiple parameters corresponding to the historical data as the main control parameter for production capacity.

[0062] It is understandable that the capacity control parameters are determined by analyzing historical data to identify the correlation between each parameter and the corresponding capacity parameters in historical production data. The determined correlation can be used as the degree of influence of each parameter on capacity. Based on this determined correlation, parameters whose correlation meets preset conditions can be selected as the capacity control parameters from among multiple parameters. These preset conditions may include the parameter's correlation being among the top N (N is a positive integer) parameters, or the correlation meeting a preset threshold, etc. This specification does not impose any restrictions on these conditions.

[0063] S103: Based on the aforementioned capacity control parameters, the historical data is augmented to obtain augmented data.

[0064] It is understandable that after determining the key production capacity control parameters, historical data can be filtered based on these parameters. Furthermore, by analyzing the characteristics of the filtered historical data, similar data can be generated. This generated data can then be merged with the filtered historical data, quickly yielding a large dataset representing the geological, engineering, and production processes of historical production wells. This expands the historical data base, providing a sufficient data foundation for training subsequent machine learning models, thereby improving the prediction accuracy of the production capacity determination model. Moreover, expanding historical data based on key production capacity control factors allows for the omission of data corresponding to parameters that have little or no impact on production capacity during feature analysis, reducing data processing workload. Simultaneously, it enables the rapid and accurate learning of relationships between data in historical data and the generation of simulation data with high feature similarity to historical data, addressing issues such as insufficient data volume, data privacy restrictions, or limited data sharing.

[0065] S104: Based on the expanded data, the pre-built machine learning model is trained to obtain a capacity determination model that takes the data corresponding to the main control parameters of capacity as input and the production data as output.

[0066] It is understood that machine learning models may include, but are not limited to, linear regression models, decision tree regression models, random forest regression models, CatBoost (symmetric gradient boosting decision tree) regression models, etc. Pre-built machine learning models may include at least one of the above models, and productivity determination models trained based on augmented data may include at least one of the above models.

[0067] S105: Obtain the target geological data corresponding to the target production well, construct a parameter optimization model based on the production capacity determination model and the target geological data, and optimize and solve the parameter optimization model to obtain the target fracturing construction parameters corresponding to the target production well.

[0068] It is understandable that the target production well and historical production wells can be located in the same area or in different areas. Furthermore, to improve the accuracy of parameter optimization, the area where the target production well is located can have similar geological features to the area where the historical production well is located.

[0069] In the embodiments described in this specification, by analyzing historical data of the target area, the main control parameters affecting production capacity can be determined, and the historical data can be expanded for subsequent training of machine learning models to improve the prediction accuracy of the production capacity determination model. Then, based on the production capacity calculation model obtained from the training and historical geological data, the fracturing construction parameters can be optimized. The target fracturing construction parameters applicable to the target area can be obtained quickly, accurately, and reliably, maximizing the production capacity of the production well.

[0070] It is understood that the execution entity of each step in the methods provided in the embodiments of this specification can be an electronic device, which refers to an electronic device with data computing, processing, and storage capabilities. This electronic device can be a terminal such as a personal computer (PC), tablet computer, smartphone, wearable device, or intelligent robot; it can also be a server. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0071] In some embodiments of this specification, data processing of the historical data to determine the parameter affecting production capacity among multiple parameters corresponding to the historical data as the main control parameter for production capacity may include: constructing at least two feature selection models, with different feature selection algorithms for different feature selection models; inputting the historical data into each feature selection model to obtain the output results of each feature selection model; constructing an evaluation matrix based on the output results of at least two feature selection models; and determining at least one parameter among multiple parameters as the main control parameter for production capacity based on the evaluation matrix.

[0072] It is understandable that feature selection models can be used to select parameters that meet certain conditions from multiple parameters, thereby achieving dimensionality reduction of data. Specifically, feature selection models can select features based on the correlation between multiple parameters and the target parameter and / or their importance to the target parameter. For example, a feature selection model can calculate the correlation between multiple parameters and the target parameter (i.e., the capacity parameter corresponding to historical production data), and can use the calculated correlation as the output of the corresponding model; a feature selection model can also analyze the degree of influence of each parameter on the capacity parameter, obtain values ​​representing the degree of influence of each parameter on the capacity parameter as the importance of each parameter to the capacity parameter, and can use the analyzed importance as the output of the model.

[0073] Furthermore, the feature selection algorithms employed by the feature selection model may include statistical algorithms such as grey relational analysis and principal component analysis, as well as machine learning algorithms for feature importance analysis, etc., which are not limited in this specification. Furthermore, the at least two feature selection models may include at least two or more of the following: a grey relational analysis calculation model, a principal component coefficient calculation model, and a machine learning model. Of course, it is understood that the at least two feature selection models may also employ more or fewer models beyond those mentioned above, which are not limited in this specification.

[0074] It is understood that the evaluation matrix is ​​constructed based on multiple parameters corresponding to historical geological data and / or historical engineering data, as well as the output results of each model. The number of rows and columns of the evaluation matrix can be related to the number of multiple parameters and the number of feature selection models. For example, if historical geological data and / or historical engineering data correspond to 9 parameters, and 3 feature selection models are constructed, then the evaluation matrix constructed based on the historical data and the output results of the feature selection models can be a 9×3 or 3×9 matrix. That is, each row in the evaluation matrix can correspond to one parameter, and each column can correspond to one feature selection model. Therefore, the data elements in the evaluation matrix can characterize the correlation between different parameters output by each feature selection model and the productivity parameter, or their importance to the productivity parameter. For example, the element in the i-th row and j-th column of the evaluation matrix can characterize the correlation between the i-th parameter among the multiple parameters calculated using the j-th feature selection model and the productivity parameter, or their importance to the productivity parameter. Alternatively, each column in the evaluation matrix can correspond to one parameter, and each row can correspond to one feature selection model; this specification does not impose any restrictions on this.

[0075] Of course, it is understandable that before performing feature selection on multiple parameters based on at least two feature selection models, a preliminary screening of historical geological data and / or historical engineering data can be conducted to filter out some data that has little or no impact on historical production data, in order to improve the data processing efficiency of the feature selection model.

[0076] In the embodiments of this specification, when selecting multiple parameters to obtain the main control parameters of production capacity, a variety of feature selection algorithms combined with the evaluation matrix method can be used to screen out more accurate and reliable parameters affecting production capacity, providing a foundation for subsequent data expansion and training of machine learning models.

[0077] In some embodiments of this specification, the first dimension of the evaluation matrix corresponds to multiple parameters, and the second dimension corresponds to at least two feature selection models. Accordingly, determining at least one parameter as the main control parameter for production capacity based on the evaluation matrix can include: normalizing the evaluation matrix; calculating the information entropy corresponding to each feature selection model in the evaluation matrix based on the normalized evaluation matrix; determining the weight of each feature selection model based on the information entropy; determining the comprehensive score of each parameter based on the weight of each feature selection model and the normalized evaluation matrix; and selecting at least one parameter as the main control parameter for production capacity based on the comprehensive score of each parameter and preset feature selection conditions. The preset feature selection conditions can be that the comprehensive score meets a preset threshold, or that the ranking result based on the comprehensive score meets a preset ranking threshold.

[0078] Specifically, the comprehensive score can be obtained by weighting the output results of each parameter in the normalized evaluation matrix in different feature selection models based on the weights corresponding to each feature selection model.

[0079] In some embodiments of this specification, data augmentation of the historical data based on the capacity control parameters to obtain augmented data may include: performing dimensionality reduction processing on the historical data based on the capacity control parameters to obtain target data; determining the statistical characteristics and parameter dependency characteristics of the target data; generating synthetic data similar to the target data based on the statistical characteristics and the parameter dependency characteristics, and using the synthetic data and the target data as augmented data.

[0080] Furthermore, when determining the statistical characteristics and parameter dependence characteristics of the target data, they can be calculated using statistical methods or analyzed using artificial intelligence models such as machine learning models and neural network models. When generating synthetic data, conventional numerical simulation methods or artificial intelligence models can be used. This specification does not impose any limitations on this.

[0081] In some embodiments of this specification, training a pre-built machine learning model based on the expanded data to obtain a capacity determination model that takes data corresponding to the main capacity control parameters as input and production data as output may include: constructing a machine learning model comprising multiple sub-models that take data corresponding to the main capacity control parameters as input and production data as output, with different sub-models employing different types of machine learning algorithms; dividing the expanded data to obtain a training set and a test set; training the multiple sub-models separately based on the training set to obtain multiple capacity determination sub-models; and selecting a capacity determination model from the multiple capacity determination sub-models based on the test set and the model accuracy of each capacity determination sub-model.

[0082] Specifically, the multiple sub-models can include at least linear regression, decision tree regression, random forest regression, and CatBoost regression models. Different sub-models are adapted to different data characteristics. Consequently, for the same training set, the prediction accuracy of each sub-model trained for capacity determination will also differ. By training each sub-model using the same training set, and then combining the test set and the accuracy of each trained sub-model, the model suitable for the expanded data can be selected as the capacity determination model from among the multiple sub-models. The ratio of the training set to the test set can be determined based on actual application needs, for example, 8:2; this specification does not impose any restrictions on this.

[0083] In some embodiments of this specification, the extended data may correspond to multiple regions. Due to differences in geological features and other characteristics, the geological, engineering, and production data of different regions will have different data characteristics. The extended data can be classified, and multiple sub-models can be trained based on the classified extended data to determine a capacity determination model suitable for predicting each category of data. Specifically, the process of training a pre-built machine learning model based on the extended data may include: clustering the extended data and dividing it into multiple datasets based on the clustering results, each dataset corresponding to a data category; further dividing each dataset to obtain a training subset and a test subset; constructing a machine learning model including multiple sub-models, with different sub-models employing different types of machine learning algorithms; for any dataset, training each sub-model using the corresponding training subset, and selecting the sub-model suitable for capacity prediction for that dataset from the trained sub-models based on the test subset and the accuracy of each sub-model; and encapsulating and combining the obtained capacity determination sub-models corresponding to the multiple datasets to obtain the final capacity determination model.

[0084] Furthermore, when optimizing the fracturing parameters of the target well in the subsequent process, the target geological data of the target well and the fracturing construction data used in the current iteration optimization can be classified first to determine their corresponding data categories. Then, based on the data category, the applicable production capacity determination sub-model can be determined, and the determined production capacity determination sub-model can be used as the objective function of the current iteration cycle.

[0085] In some embodiments of this specification, a parameter optimization model is constructed based on the production capacity determination model and the target geological data, and the parameter optimization model is optimized and solved to obtain the target fracturing construction parameters corresponding to the target production well. This may include: using the production capacity determination model as the objective function and the target geological data as the constraint condition, and taking the maximum production capacity as the objective, constructing a parameter optimization model including the objective function and the constraint condition; determining the value range corresponding to the fracturing construction parameters, and initializing the fracturing construction parameters based on the value range to obtain initial fracturing construction data; inputting the initial fracturing construction data into the objective function to calculate the production capacity calculation result corresponding to the initial fracturing construction data; iteratively optimizing the initial fracturing construction data based on the production capacity calculation result and the constraint condition until a preset optimization condition is met, and taking the value of the fracturing construction parameter corresponding to the largest production capacity calculation result during the iterative optimization process as the target fracturing construction parameter.

[0086] Specifically, when iteratively optimizing initial fracturing operation data, at least one of the following optimization algorithms can be used: particle swarm optimization, whale optimization, simulated annealing, etc. When selecting multiple optimization algorithms for iterative optimization, the weights of each algorithm can be determined by considering constraints, fracturing operation data, characteristics of various data types in the production capacity calculation results, and the characteristics between data points. Then, based on the determined weights, multiple optimization algorithms are fused to determine the target optimization algorithm suitable for the current application scenario. Iterative optimization of the initial fracturing operation data is then performed based on the target optimization algorithm to obtain the target fracturing operation parameters. Furthermore, for each iteration, the weights of each optimization algorithm can be dynamically adjusted to achieve adaptive dynamic adjustment of the target optimization algorithm throughout the entire iterative optimization process, improving the accuracy and reliability of fracturing operation parameter optimization.

[0087] In some embodiments of this specification, the target geological data of the target well can be used as a physical constraint. Physical restrictions are added during the optimization process to ensure the rationality of the optimization results. The production capacity determination model is used as the objective function to optimize the fracturing operation parameters of the target well. The specific optimization process may include: based on particle swarm optimization (PSO), whale optimization, and simulated annealing algorithms, dynamically adjusting the weight ratio between PSO and whale optimization algorithms and incorporating the perturbation strategy of simulated annealing, an adaptive dynamic optimization intelligent optimization algorithm is obtained. Then, a fracturing operation parameter optimization model is established based on this adaptive dynamic optimization intelligent optimization algorithm. The trained production capacity determination model is used as the objective function of the optimization model, with maximizing production capacity as the objective. By fixing the input target geological parameters as physical constraints, the range of values ​​to be optimized for the fracturing operation parameters is adjusted. Initial fracturing operation data is generated by randomly sampling within the parameter space corresponding to the range of values ​​to be optimized. Subsequent iterative optimization processes can then be performed based on this initial fracturing operation data and the parameter space. During the iterative optimization process, the objective function value (i.e., the production capacity calculation result) of each set of fracturing operation data (initial fracturing operation data calculated in the first iteration) can be calculated, and the fracturing operation parameters of the target well can be optimized based on the objective function value of each set of fracturing operation data. During the optimization process, physical constraints are added to ensure the rationality of the optimized parameter results in the actual application scenario, and the best fracturing operation data of the current period is retained. This optimization step is repeated until the termination condition is met or the maximum number of iterations is reached, at which point the iterative optimization stops. The output optimal production capacity calculation result and its corresponding fracturing operation data are used as the target fracturing operation parameters under the maximum production capacity.

[0088] In the embodiments of this specification, the above-mentioned method for determining fracturing construction parameters is used. By making full use of existing actual production data and combining artificial intelligence technology, the fracturing construction parameters are optimized based on different reservoir geology and different oil and gas properties, with oil and gas production capacity as the evaluation condition. This can break through the limitations of traditional models, realize the intelligent integration of geological, engineering and production capacity data, thereby accurately guide on-site construction, maximize the development benefits of deep coalbed methane wells, realize the rapid and intelligent design of hydraulic fracturing construction schemes, and maximize the production capacity of target wells.

[0089] This specification also provides an embodiment of a method for optimizing deep coal seam fracturing construction parameters based on physical constraints and data-driven approaches. (Reference) Figure 2 As shown, the above method may include:

[0090] S201: Collect geological, engineering, and production capacity data of deep coal seam production wells within the study area.

[0091] Specifically, the study involves collecting and organizing on-site seismic and well logging data, as well as reservoir geological parameters for the study area, such as stratigraphic lithology, coal seam thickness, coal quality characteristics, roof and floor lithology, porosity, and permeability. It also includes collecting fracturing engineering parameters from production wells constructed using fracturing techniques, including proppant addition scale, displacement, whether temporary plugging was used, and fracturing processes. Production data is selected based on parameters reflecting the gas production capacity of the wells, such as cumulative gas production over the past three months, average daily gas production, and peak gas production.

[0092] S202: Based on geological, engineering and production capacity data, conduct data processing and analysis to determine the main control parameters affecting production capacity.

[0093] The factors influencing deep coalbed methane production capacity are complex and multifaceted. Deep coal seam production capacity is affected by a combination of geological and engineering parameters. Different geological blocks have different geological conditions, and the degree of influence of the main controlling parameters affecting production capacity also varies. By processing and analyzing the above data, more accurate analytical results can be obtained.

[0094] Specifically, the collected geological-engineering-production data can be preprocessed, and outliers and missing values ​​can be handled using correlation coefficients and box plots to obtain a high-quality geological-engineering-production dataset. Furthermore, grey relational analysis, principal component coefficient analysis, and machine learning models (such as random forest models and gradient boosting tree models) can be established among the geological-engineering-production parameters. The machine learning model can perform feature importance analysis. Further, the analysis results corresponding to different parameters obtained by the above methods can be used to form a multi-method evaluation matrix through fuzzy mathematics, and the evaluation results of different parameters in the evaluation matrix can be weighted according to the entropy weight method to obtain the final comprehensive score of the geological and engineering parameters. In the embodiments of this specification, the comprehensive scores of each parameter can be obtained as follows: Figure 3 As shown, the horizontal axis represents various parameters, including gas content, GSI (Geological Strength Index) value, sand content, sand addition intensity, flowback rate, average sand ratio, pre-treatment liquid ratio, density, and vertical thickness. The vertical axis represents the fuzzy comprehensive evaluation (i.e., comprehensive score) corresponding to each parameter. Geological engineering parameters can be sorted from highest to lowest score, and the main controlling factors affecting production capacity, i.e., the main production capacity control parameters, can be determined based on the sorting results.

[0095] S203: Establishing simulation extended data based on the main control factors of production capacity and the SDV-GCM method.

[0096] Specifically, based on the main production capacity control parameters obtained from S202, the SDV-GCM (Synthetic Data Vault-Gaussian Copula Model) model method is used to generate similar synthetic data while preserving the statistical characteristics and variable dependencies of the original data. This can solve problems such as insufficient data volume, limited data privacy, or limited data sharing. SDV is a Python toolkit for generating high-quality synthetic data. It is used to generate similar synthetic data while preserving the statistical characteristics and variable dependencies of the original data, thus addressing issues such as insufficient data volume, limited data privacy, or limited data sharing. By augmenting the field data using the GCM (Gaussian Copula Model) model in the SDV method, highly similar simulation data containing the relationships between actual data can be obtained.

[0097] S204: Establish a machine learning capacity prediction agent model and train the model based on expanded data to obtain a capacity intelligent agent model.

[0098] The capacity intelligent proxy model (i.e., the capacity determination model mentioned earlier) takes capacity control parameters as input and outputs capacity data. Specifically, various machine learning regression proxy models can be established, and the expanded data obtained in S203 can be used to train these proxy models, resulting in a capacity intelligent proxy model with higher training accuracy. These various machine learning regression proxy models can include, but are not limited to, linear regression models, decision tree regression models, random forest regression models, and CatBoost regression models. The quality of the expanded data affects the capacity prediction accuracy of the machine learning model. Using the high-quality expanded data obtained earlier for machine model training yields a capacity intelligent proxy model with higher training accuracy. By encapsulating and fusing multiple machine learning models into a capacity intelligent proxy model, and combining the accuracy and prediction results of each model, adaptive optimization of the optimal intelligent proxy model under different data conditions can be achieved.

[0099] S205: Combine actual field data with the proxy model to verify the effectiveness of the model.

[0100] Specifically, the effectiveness and rationality of the intelligent production capacity proxy model can be ensured by predicting and fitting actual field data using a trained intelligent production capacity proxy model. For example, the trained intelligent production capacity proxy model can be combined with actual field data, and the mean squared error (MSE) can be used to evaluate the fitting accuracy of the intelligent production capacity proxy model to the field data, thereby verifying that the established intelligent production capacity proxy model has high practical physical significance.

[0101] S206: Use the geological data of the fracturing section of the deep coal seam production well as the physical constraints, and use the production capacity prediction proxy model as the objective function to optimize the fracturing construction parameters of the deep coal seam.

[0102] Specifically, the high-precision intelligent agent model for production capacity trained in S205 is used as the objective function of the optimization algorithm. Based on field experience and analysis of field data, the geological and engineering parameters to be optimized in the corresponding study area are obtained as constraints. The fracturing parameters are optimized with the goal of maximizing production capacity, and the fracturing parameter scheme corresponding to the maximum production capacity is obtained.

[0103] refer to Figure 4 As shown in some embodiments of this specification, the optimization method for the above-mentioned 206 fracturing construction parameters may include:

[0104] S401: Determine the range of values ​​for the fracturing construction parameters to be optimized, and use the geological data of the fracturing section of the deep coal seam production well as a physical constraint;

[0105] S402: Initialize the fracturing parameters for deep coal seams;

[0106] S403: Calculate the fitness value of fracturing construction parameters for deep coal seams based on the intelligent agent model of production capacity;

[0107] S404: Optimize the fracturing parameters of deep coal seams based on the fitness value of the fracturing operation parameters. During the optimization process, control the addition of physical constraints to ensure the rationality of the optimized parameter results in a practical physical sense.

[0108] S405: When the global optimum or the maximum number of iterations is reached, the fracturing parameters of the deep coal seam are optimized. The output of the optimal gas production and the corresponding values ​​of various parameters are taken as the optimal combination of fracturing parameters under the maximum gas production.

[0109] During the optimization process, intelligent optimization algorithms such as particle swarm optimization, whale optimization, and simulated annealing can be used. By dynamically adjusting the weight ratio between particle swarm optimization and whale optimization, and incorporating the perturbation strategy of simulated annealing, an adaptive dynamic optimization intelligent optimization algorithm can be obtained. Then, an optimization model for fracturing construction parameters can be established based on this adaptive dynamic optimization intelligent optimization algorithm.

[0110] Based on the method for determining fracturing operation parameters described above, one or more embodiments of this specification also provide a device for determining fracturing operation parameters. The device may include an apparatus (including a distributed system), software (application), module, plug-in, server, client, etc., using the method described in the embodiments of this specification, combined with necessary implementation hardware. Based on the same innovative concept, the devices in one or more embodiments provided in this specification are as described in the following embodiments. Since the implementation schemes and methods for solving the problem by the device are similar, the implementation of the specific device in the embodiments of this specification can refer to the implementation of the foregoing method, 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 device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated. Figure 5 The diagram shown is a schematic representation of a device for determining fracturing operation parameters provided in an embodiment of this application. Figure 5 As shown, the device 500 for determining fracturing operation parameters may include:

[0111] The acquisition module 501 is used to acquire historical data of multiple historical production wells, including historical geological data, historical engineering data, and historical production data;

[0112] Processing module 502 is used to process the historical data and determine the parameter that affects the production capacity among the multiple parameters corresponding to the historical data as the main control parameter of production capacity.

[0113] The expansion module 503 is used to expand the historical data based on the main control parameters of the production capacity to obtain expanded data;

[0114] Training module 504 is used to train a pre-built machine learning model based on the expanded data to obtain a capacity determination model with the data corresponding to the main control parameters of capacity as input and production data as output.

[0115] The optimization module 505 is used to acquire the target geological data corresponding to the target production well, construct a parameter optimization model based on the production capacity determination model and the target geological data, and optimize and solve the parameter optimization model to obtain the target fracturing construction parameters corresponding to the target production well.

[0116] In some embodiments of this specification, the processing module 502 may be specifically used to: construct at least two feature selection models, with different feature selection algorithms for different feature selection models; input the historical data into each feature selection model respectively to obtain the output results of each feature selection model; construct an evaluation matrix based on the output results of at least two feature selection models; and determine at least one parameter as the main control parameter for production capacity based on the evaluation matrix among multiple parameters.

[0117] In some embodiments of this specification, the at least two feature selection models may include at least: a grey relational analysis model, a principal component coefficient calculation model, and a machine learning model.

[0118] In some embodiments of this specification, the first dimension of the evaluation matrix corresponds to multiple parameters, and the second dimension corresponds to at least two feature selection models. Correspondingly, when the processing module 502 determines at least one parameter as the main control parameter for production capacity based on the evaluation matrix among multiple parameters, it can specifically be used to: normalize the evaluation matrix; calculate the information entropy corresponding to each feature selection model in the evaluation matrix based on the normalized evaluation matrix; determine the weight corresponding to each feature selection model based on the information entropy corresponding to each feature selection model; determine the comprehensive score of each parameter based on the weight corresponding to each feature selection model and the normalized evaluation matrix; and select at least one parameter as the main control parameter for production capacity from among multiple parameters based on the comprehensive score of each parameter and preset feature selection conditions.

[0119] In some embodiments of this specification, the expansion module 503 may be specifically used to: perform dimensionality reduction processing on the historical data based on the main control parameters of the production capacity to obtain target data; determine the statistical characteristics and parameter dependency characteristics of the target data; generate synthetic data similar to the target data based on the statistical characteristics and the parameter dependency characteristics, and use the synthetic data and the target data as expansion data.

[0120] In some embodiments of this specification, the training module 504 may specifically be used to: construct a machine learning model comprising multiple sub-models that take data corresponding to the main control parameters of production capacity as input and production data as output, wherein different sub-models employ different types of machine learning algorithms; divide the expanded data to obtain a training set and a test set; train the multiple sub-models based on the training set to obtain multiple production capacity determination sub-models; and select a production capacity determination model from the multiple production capacity determination sub-models based on the test set and the model accuracy of each production capacity determination sub-model.

[0121] In some embodiments of this specification, the optimization module 505 may specifically be used to: construct a parameter optimization model including the objective function and the constraint conditions, with the production capacity determination model as the objective function and the target geological data as the constraint conditions, and with the maximum production capacity as the objective; determine the value range corresponding to the fracturing construction parameters, and initialize the fracturing construction parameters based on the value range to obtain initial fracturing construction data; input the initial fracturing construction data into the objective function to calculate the production capacity calculation result corresponding to the initial fracturing construction data; iteratively optimize the initial fracturing construction data based on the production capacity calculation result and the constraint conditions until the preset optimization conditions are met, and take the value of the fracturing construction parameter corresponding to the largest production capacity calculation result during the iterative optimization process as the target fracturing construction parameter.

[0122] The descriptions and functions of the above modules can be understood by referring to the section on methods for determining fracturing construction parameters, and will not be repeated here.

[0123] This application also provides an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 601 and a memory 602, wherein the processor 601 and the memory 602 may be connected via a bus or other means. Figure 6 Taking the example of a connection between China and Israel via a bus.

[0124] Processor 601 may be a central processing unit (CPU). Processor 601 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof.

[0125] The memory 602, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for determining fracturing construction parameters in this embodiment of the invention (e.g., Figure 5 The processor 601 comprises an acquisition module 501, a processing module 502, an expansion module 503, a training module 504, and an optimization module 505. The processor 601 executes various functional applications and data processing by running non-transitory software programs, instructions, and modules stored in the memory 602, thereby implementing the method for determining fracturing construction parameters in the above-described method embodiments.

[0126] The memory 602 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor 601, etc. Furthermore, the memory 602 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 602 may optionally include memory remotely located relative to the processor 601, and these remote memories may be connected to the processor 601 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0127] The one or more modules are stored in the memory 602, and when executed by the processor 601, they perform the following method for determining fracturing construction parameters:

[0128] Historical data from multiple historical production wells is acquired, including historical geological data, historical engineering data, and historical production data. This historical data is then processed to determine the parameters affecting production capacity among multiple parameters corresponding to the historical data, which are then used as the main production capacity control parameters. Based on these main production capacity control parameters, the historical data is augmented to obtain augmented data. A pre-constructed machine learning model is trained using the augmented data to obtain a production capacity determination model that takes the data corresponding to the main production capacity control parameters as input and the production data as output. Target geological data corresponding to the target production well is acquired. Based on the production capacity determination model and the target geological data, a parameter optimization model is constructed, and the parameter optimization model is optimized and solved to obtain the target fracturing operation parameters corresponding to the target production well.

[0129] The specific details of the aforementioned electronic device can be understood by referring to the relevant descriptions and effects in the above method embodiments, and will not be repeated here.

[0130] This specification also provides a computer storage medium storing computer program instructions, which, when executed, implement the steps of the method for determining the fracturing construction parameters described above.

[0131] This specification also provides a computer program product, which includes a computer program that, when executed, implements the steps of the method for determining the fracturing construction parameters described above.

[0132] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0133] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, please refer to each other. The focus of each embodiment is to describe the differences from other embodiments.

[0134] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0135] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0136] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute certain parts of the methods of various embodiments of this application.

[0137] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0138] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0139] The above description is merely a preferred embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to the embodiments described herein by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

Claims

1. A method of determining fracturing job parameters, characterized in that, The method comprises the following steps: acquiring historical data of a plurality of historical production wells, wherein the historical data comprises historical geological data, historical engineering data and historical production data; performing data processing on the historical data, and determining parameters affecting productivity among parameters corresponding to the historical data as productivity main control parameters; performing data expansion on the historical data based on the productivity main control parameters to obtain expanded data; training a pre-constructed machine learning model based on the expanded data to obtain a productivity determination model taking data corresponding to the productivity main control parameters as input and taking production data as output; acquiring target geological data corresponding to a target production well, constructing a parameter optimization model based on the productivity determination model and the target geological data, and performing optimization solving on the parameter optimization model to obtain target fracturing construction parameters corresponding to the target production well; performing data processing on the historical data, and determining parameters affecting productivity among parameters corresponding to the historical data as productivity main control parameters, comprising: constructing at least two feature selection models, wherein different feature selection models adopt different feature selection algorithms; inputting the historical data into each feature selection model to obtain output results of each feature selection model; constructing an evaluation matrix based on the output results of the at least two feature selection models; data elements in the evaluation matrix represent the correlation between different parameters output by each feature selection model and the productivity parameter or the importance of the productivity parameter; performing weighted calculation on the data elements corresponding to each parameter in the evaluation matrix according to an entropy weight method to obtain a comprehensive score of each parameter; based on the comprehensive score of each parameter and a preset feature selection condition, at least one parameter is screened from the plurality of parameters as the productivity main control parameter.

2. The method of determining fracturing treatment parameters according to claim 1, characterized in that, The at least two feature selection models comprise a grey correlation degree calculation model, a principal component coefficient calculation model and a machine learning model.

3. The method of determining fracturing treatment parameters of claim 1, wherein, The first dimension of the evaluation matrix corresponds to the plurality of parameters, and the second dimension corresponds to the at least two feature selection models. Accordingly, the weighted calculation on the data elements corresponding to each parameter in the evaluation matrix according to the entropy weight method to obtain the comprehensive score of each parameter comprises: performing normalization processing on the evaluation matrix; calculating information entropy corresponding to each feature selection model in the evaluation matrix based on the normalized evaluation matrix; determining weights corresponding to each feature selection model based on the information entropy corresponding to each feature selection model; determining the comprehensive score of each parameter based on the weights corresponding to each feature selection model and the normalized evaluation matrix.

4. The method of determining fracturing treatment parameters of claim 1, wherein, Performing data expansion on the historical data based on the productivity main control parameters to obtain expanded data, comprising: performing dimension reduction processing on the historical data based on the productivity main control parameters to obtain target data; determining statistical features and parameter dependent features of the target data; generating synthetic data similar to the target data based on the statistical features and the parameter dependent features, and taking the synthetic data and the target data as expanded data.

5. The method of determining fracturing treatment parameters according to claim 1, wherein, Training a pre-constructed machine learning model based on the expanded data to obtain a productivity determination model taking data corresponding to the productivity main control parameters as input and taking production data as output, comprising: The machine learning model comprises a plurality of sub-models, each of which takes data corresponding to a capacity main control parameter as input and takes production data as output, and different sub-models adopt different types of machine learning algorithms; The augmented data is divided to obtain a training set and a test set; The plurality of sub-models are trained based on the training set to obtain a plurality of capacity determination sub-models; Based on the test set and the model accuracy of each capacity determination sub-model, a capacity determination model is selected from the plurality of capacity determination sub-models.

6. The method of determining fracturing treatment parameters according to claim 1, wherein, Based on the capacity determination model and the target geological data, a parameter optimization model is constructed, and the parameter optimization model is optimized and solved to obtain target fracturing construction parameters corresponding to the target production well, including: The capacity determination model is taken as the objective function, the target geological data is taken as the constraint condition, and the maximum capacity is taken as the target to construct a parameter optimization model comprising the objective function and the constraint condition; The value range of the fracturing construction parameter is determined, and the fracturing construction parameter is initialized based on the value range to obtain initial fracturing construction data; The initial fracturing construction data is input into the objective function to calculate the capacity calculation result corresponding to the initial fracturing construction data; Based on the capacity calculation result and the constraint condition, the initial fracturing construction data is iteratively optimized until a preset optimization condition is met, and the value of the fracturing construction parameter corresponding to the maximum capacity calculation result in the iterative optimization process is taken as the target fracturing construction parameter.

7. An apparatus for determining fracturing job parameters, the apparatus comprising: It includes: An acquisition module is configured to acquire historical data of a plurality of historical production wells, the historical data comprising historical geological data, historical engineering data and historical production data; A processing module is configured to determine parameters affecting capacity from a plurality of parameters corresponding to the historical data as capacity main control parameters by processing the historical data; An expansion module is configured to expand the historical data based on the capacity main control parameters to obtain expanded data; A training module is configured to train a pre-constructed machine learning model based on the expanded data to obtain a capacity determination model taking data corresponding to the capacity main control parameters as input and taking production data as output; An optimization module is configured to obtain target geological data corresponding to a target production well, construct a parameter optimization model based on the capacity determination model and the target geological data, and optimize and solve the parameter optimization model to obtain target fracturing construction parameters corresponding to the target production well; The processing module is specifically configured to: At least two feature selection models are constructed, and different feature selection models adopt different feature selection algorithms; The historical data is input into each feature selection model to obtain output results of each feature selection model; An evaluation matrix is constructed based on the output results of the at least two feature selection models; data elements in the evaluation matrix represent the correlation between different parameters output by each feature selection model and the capacity parameter or the importance of the capacity parameter; According to the entropy weight method, the data elements corresponding to each parameter in the evaluation matrix are weighted and calculated to obtain a comprehensive score of each parameter; Based on the comprehensive score of each parameter and a preset feature selection condition, at least one parameter is selected from the plurality of parameters as the capacity main control parameter.

8. An electronic device, comprising: The computer readable storage medium stores computer program instructions, and the computer program instructions are executed by the processor to implement the steps of the method in any one of claims 1 to 6. ​ 9. A computer-readable storage medium, characterized in that, ​

Images