Multi-objective optimization method and system for co2 injection for enhanced recovery and storage in gas reservoirs

Abstract

The present application belongs to the technical field of energy exploitation, and particularly relates to a multi-target optimization method and system for gas reservoir injection, extraction, production and storage. First, the control parameter variables for extraction and storage are obtained, and an improved sampling algorithm is called to sample and combine to generate sample data. Then, a CMG data file is called to modify in batches, and the well name and DATE command are extracted from the original dat file. The sample parameters are simulated, a new scheduling block is generated and inserted into the original file at the last DATE command, a new CMG file is generated, and the processed CSV is exported and merged. Next, a multi-output regression prediction model is constructed, and the model is trained with the CSV file. Then, all the control parameter variables are combined and exported, the trained model is called to predict, and the parameter full combination is obtained. Finally, a multi-target optimization model is constructed, the cumulative production and storage efficiency are maximized, and the Pareto optimal solution is exported as the result. The present application can solve the problem of low efficiency and inaccurate precision in the existing technology of storage and natural gas exploitation collaborative operation.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of energy exploitation, and particularly relates to a gas reservoir injection optimization method and system. BACKGROUND

[0002] Under the dual driving of the "double carbon" target and the upgrading of energy demand, Geological storage and natural gas exploitation are the core direction of low-carbon and efficient development in the energy field. Currently, CMG numerical simulator (such as GEM component model) is the core tool in this field, which can build a fine model covering reservoir geological characteristics and fluid phase change, simulate the exploitation efficiency and storage effect under different injection parameters, and provide support for engineering scheme design; at the same time, Although the flooding technology can improve the recovery rate of gas reservoirs and realize storage, the existing technical system has multiple shortcomings, and relies on manual completion of parameter optimization, numerical simulation and data processing of storage and natural gas exploitation, which is complicated and consumes a lot of manpower and material resources, and is low in efficiency, which is difficult to meet the actual production and environmental protection needs.

[0003] First of all, in the determination of engineering parameters, the existing method mainly relies on artificial experience to set parameters and determine parameters one by one, but this way is not only low in efficiency but also ignores the relationship between parameters, so it is difficult to ensure that the global optimal solution is obtained.

[0004] Then in the CMG data file processing link, the current method mainly relies on manual modification of ".dat" parameters file by file, which is complicated and easy to make mistakes, and the modification process has no standardized record and poor traceability, and often needs to check the results after the simulation is completed to find that there are parameter setting errors.

[0005] Finally, in the prediction of the development potential of the flooding gas reservoir, the traditional numerical simulation method relies on special and expensive component simulator, and the calculation efficiency is extremely low, which is difficult to quickly and accurately evaluate the actual development potential. At the same time, the current optimization target often focuses too much on economic benefits or maximum storage capacity, and seriously ignores the great threat of leakage to the storage effect, which makes the feasibility of the optimization result greatly discounted in actual engineering application. These defects jointly restrict the efficiency and accuracy of the collaborative operation, and innovative technology breakthrough is urgently needed. SUMMARY

[0006] The technical problem solved by the present application is to provide a gas reservoir injection optimization method and system to solve the problems of low efficiency and inaccurate accuracy in the existing storage and natural gas exploitation collaborative operation.

[0007] The application provides a basic scheme: a gas reservoir injection The multi-objective optimization method for extraction and storage includes:

[0008] S1: obtaining The control parameter variables of extraction and storage are sampled and combined by calling a preset sampling algorithm to generate sample data;

[0009] S2: calling a CMG data file batch modification tool to extract all well names and DATE commands from the original dat file, verifying the effectiveness of the target control well and positioning the last DATE command position; reading the control parameter variables from the sample data and performing injection-production ratio, Concentration calculation; generating a new scheduling block based on the CMG syntax rules, inserting the new scheduling block into the last DATE command position of the original dat file, generating a new complete CMG data file, and then importing the CMG software for simulation, exporting a CSV file from the result file of the CMG running and performing merging processing;

[0010] S3: constructing a multi-output regression prediction model, generating a training set, a validation set and a test set after preprocessing the exported merged CSV file, inputting the training set into the multi-output regression prediction model for training, verifying and testing the trained multi-output regression prediction model through the validation set and the test set, and obtaining the trained multi-output regression prediction model until the training is completed;

[0011] S4: extracting All control parameter variables of extraction and storage are combined and exported as file data, and the trained multi-output regression prediction model is called to predict the file data to obtain All combinations of control parameter variables in extraction and storage;

[0012] S5: calling a multi-objective optimization algorithm to construct a multi-objective optimization model, and inputting All combinations of control parameter variables in extraction and storage into the multi-objective optimization model, selecting maximum Cumulative production and maximum Storage efficiency according to the optimization target, exporting the Pareto optimal solution as The multi-objective optimization result of extraction and storage.

[0013] Further, the S1 includes:

[0014] S1-1: dividing the obtained Control parameter variables of extraction and storage into equal-probability subintervals according to their variable types, wherein the variable types include discrete variables and continuous variables;

[0015] S1-2: Assign a sampling quota to each variable according to the total number of samples, and establish a sampling index list for each variable;

[0016] S1-3: Randomly shuffle the sampling index list of each variable, randomly select a position from the random index list of each variable, extract the corresponding division point value to form a sampling combination, and continue until the preset sample size is reached;

[0017] S1-4: Call the improved lion optimization algorithm to extract the optimal boundary reinforcement combination of the extreme value of each variable, and after repetitive judgment with the sampling combination generated in S1-3, combine the sampling combination to generate the final variable sampling combination result.

[0018] Further, the S1-4 calling the improved lion optimization algorithm to extract the optimal boundary reinforcement combination of the extreme value of each variable is specifically:

[0019] Obtain the extreme value of each control parameter variable , and construct a boundary target vector;

[0020] Initialize the lion population and the ant population, and add the boundary target vector to the lion population; According to the proximity boundary and the engineering feasible target, the fitness function is constructed, and the expression is:

[0021]

[0022]

[0023]

[0024]

[0025] Wherein, is the boundary guiding term, is the engineering feasibility constraint term, is the ant individual, is the kth boundary target vector, is the Hamming distance, and the engineering constraint is based on the physical law of gas reservoir, is the penalty function; and are weight coefficients adjusted according to optimization requirements, and are greater than 0 and add up to 1; the engineering feasibility constraint term other indicates that x has an item that does not meet all engineering constraints;

[0026] Adjust the motion direction of the ant through the boundary attraction factor, and execute the boundary guiding type walking, and the expression is: ​​

[0027]

[0028]

[0029] in, Indicates in In the iteration, the ant's new position on the i-th control parameter variable, Indicates the first In the iteration, the ant's current position on the i-th control parameter variable is... Indicates the first The step size of the random walk in the iteration. For the first Generated in generation iteration Uniformly distributed random numbers Let i be the current value of the ant in the t-th generation on the i-th variable. For the boundary attraction term, Let be the i-th extreme value of the boundary target vector that the ant individual is currently closest to; The attraction intensity coefficient, ; For individual ants The boundary target vector corresponding to the current minimum Hamming distance. Let x be the ant's current position and the optimal boundary target vector. Hamming distance, is the distance decay factor, which indicates that the closer an ant is to the boundary, the stronger the attraction, and the farther away it is, the weaker the attraction.

[0030] After each iteration, based on the boundary guiding term Update the antlion trap radius, mark the individual with the best fitness in the population as an elite antlion, and force the retention of at least one individual that meets the criteria. An elite individual with an extreme value for a single variable;

[0031] After the maximum number of iterations is reached, the best individuals are extracted from the final elite antlion population as the boundary reinforcement combination, and a repeatability judgment is performed on the sampled combination generated in S1-3. Finally, the final sample data of the random sampled combination and the deduplicated boundary reinforcement combination are output.

[0032] Furthermore, S2 includes:

[0033] S2-1: Read the contents of the original dat file, extract all well names and DATE commands using regular expressions, verify the existence of wells with control parameter variables by using the extracted well names, and locate the position of the last DATE command in the original dat file;

[0034] S2-2: Read the regulation parameter variables from the sample data and perform... Concentration calculation, injection-production ratio calculation, and generation of a new schedule block according to CMG syntax rules combined with all monthly DATE commands of the injection well starting date and ending date, PRODUCER and OPERATE rate control statements of the production well, INJECTOR of the injection well, INCOMP fluid concentration configuration and corresponding OPERATE statements;

[0035] S2-3: Truncate the original dat file to the last DATE command, insert the generated new schedule block, and splice the original dat file to combine the subsequent content into a complete new file, and then generate a unique file name according to the preset file naming rule to batch generate CMG data files, and save them to a specified directory.

[0036] Further, the S2 further comprises:

[0037] S2-4: Import the batch-modified generated CMG data file into the CMG software for running simulation, and then export the running simulation results as a CSV file, and perform data cleaning processing on the exported CSV file, including calling a preset regular expression matching format to match and convert the file name to a standard format and process invalid values, filtering data from the CSV file according to business rules, and calculating the core indicators of storage efficiency; wherein the storage efficiency expression is:

[0038]

[0039] wherein, represents storage efficiency, represents the cumulative injection molar number, represents the cumulative output molar number; if the injection is 0, then the storage efficiency is 0;

[0040] S2-5: Merge the CSV files after data cleaning processing into a single file, retain the table header of the first file, and only append data rows to the subsequent CSV files, to generate a merged CSV file containing all samples.

[0041] Further, the S3 comprises:

[0042] S3-1: Construct a multi-output regression prediction model based on an improved stacked ensemble algorithm;

[0043] S3-2: After preprocessing the merged CSV file, divide it into a training set, a validation set, and a test set according to a ratio of 7:1.5:1.5;

[0044] S3-3: Input the training set into the multi-output regression prediction model for training, specifically:

[0045] Based on the input training set, the first... Feature vector of each sample Input into two base models:

[0046] Random Forest model expression:

[0047]

[0048] XGBoost model expression:

[0049]

[0050] Among them, for the training set Feature vector of each sample Random Forest Model Output predicted value vector XGBoost model It will output a vector of predicted values. , In the random forest model Forecast of cumulative output, In the random forest model Forecast of cumulative output In the random forest model Predicted value of mole fraction; In the XGBoost model Forecast of cumulative output In the XGBoost model Forecast of cumulative output, In the XGBoost model Predicted value of mole fraction;

[0051] The prediction results of the two base models are concatenated into a meta-feature vector. The expression is:

[0052]

[0053] Metamodel Meta-features As input, output the final predicted value. The expression is:

[0054]

[0055] in, These are the final predicted results for the three dependent variables; express The final forecast of cumulative production, express The final forecast of cumulative production, express The final prediction result of the mole fraction;

[0056] The meta-model is optimized by minimizing the loss function on the training set, expressed as:

[0057]

[0058] in, For the first The true value of the t-th dependent variable in a sample, where n is the total number of samples. For the first The predicted value of the t-th dependent variable in a sample. Representing the metamodel;

[0059] To impose a business scope constraint on the prediction results, the expression is:

[0060]

[0061] in Let t represent the minimum value of the t-th dependent variable. Let t be the maximum value of the dependent variable. For the first The predicted value of the t-th dependent variable in a sample;

[0062] S3-4: Based on the training results, call the preset indicators for evaluation, and perform validation and testing based on the validation set and test set until a multi-output regression prediction model that meets the evaluation indicators is output.

[0063] Furthermore, S5 includes:

[0064] S5-1: Based on the complete combination of input control parameter variables, extract the injection rate, injection-to-enzyme ratio, and other parameters based on the date index. concentration, Gas mole fraction;

[0065] S5-2: Call The algorithm constructs a multi-objective optimization model, calling the variable influence coefficient formula to evaluate injection rate, injection-to-enzyme ratio, and other variables. concentration, Influence coefficient analysis of gas mole fraction;

[0066] S5-3: Injection rate and injection-production ratio based on influence coefficient analysis concentration, In a multi-objective optimization model, the gas mole fraction is used as the optimization target, and the output of the optimization objective is to maximize the gas mole fraction. Cumulative production and maximization Storage efficiency.

[0067] Further, the variable influence coefficient formula in S5-2 influences the injection rate, injection-production ratio, Concentration, Gas mole fraction to perform influence coefficient analysis specifically as follows:

[0068] Injection rate influence coefficient analysis:

[0069]

[0070] Wherein, represents the injection rate influence coefficient, is the injection rate weight, is the injection rate mean value, is the injection rate standard deviation;

[0071] Injection-production ratio influence coefficient analysis:

[0072]

[0073] Wherein, represents the injection-production ratio influence coefficient, is the injection-production ratio weight, is the injection-production ratio mean value, is the injection-production ratio standard deviation;

[0074] Concentration influence coefficient analysis includes respectively on Production analysis and Sequestration analysis:

[0075] Concentration on Production analysis:

[0076]

[0077] Concentration on Sequestration analysis:

[0078]

[0079] Wherein, represents the influence coefficient of Concentration on Production, represents the influence coefficient of Concentration on Sequestration, is the concentration weight, is the concentration mean value, is the concentration standard deviation, is the concentration standard deviation.The mean value of the concentration, For The standard deviation of the concentration;

[0080] Gas mole fraction influence coefficient analysis:

[0081]

[0082] Wherein, Indicates The influence coefficient of the gas mole fraction, For The mean value of the mole fraction, The mole fraction standard deviation.

[0083] Further, the optimization target output in S5-3 is to maximize The cumulative yield and maximize The storage efficiency is specifically:

[0084] The optimization target is to maximize The cumulative yield expression is:

[0085]

[0086] The optimization target maximizes The storage efficiency expression is:

[0087]

[0088] Wherein, The final calculated Cumulative yield optimization target value, Indicates The reference value of the cumulative yield, The final calculated Storage efficiency optimization target value, For The storage efficiency reference value, Avoid the result being 0 or negative.

[0089] A gas reservoir injection Multi-objective optimization system for extraction and storage, applying the above-mentioned gas reservoir injection Multi-objective optimization method for extraction and storage, comprising a sample data generation module, a CMG processing module, a prediction model generation training module, a variable combination module, and a multi-objective optimization module, wherein:

[0090] The sample data generation module is used to obtain Extraction and storage control parameter variables, call a preset sampling algorithm to sample and combine the control parameter variables, and generate sample data;

[0091] CMG processing module: for calling CMG data file batch modification tool, extracting all well names and DATE commands from the original dat file, verifying the validity of the target control well and locating the last DATE command position; reading the control parameter variables from the sample data, and carrying out injection-production ratio, Concentration calculation; based on the CMG syntax rules, a new scheduling block is generated, the new scheduling block is inserted into the last DATE command position of the original dat file, a new complete CMG data file is generated, and then imported into the CMG software for simulation, and a CSV file is exported from the result file of the CMG running and merged;

[0092] Prediction model generation training module: for constructing a multi-output regression prediction model, preprocessing the exported merged CSV file to generate a training set, a validation set and a test set, inputting the training set into the multi-output regression prediction model for training, and verifying and testing the trained multi-output regression prediction model through the validation set and the test set, until a trained multi-output regression prediction model is obtained;

[0093] Variable combination module: for extracting all control parameter variables of extraction and storage, combining variables and exporting as file data, calling the trained multi-output regression prediction model to predict the file data, and obtaining all combinations of control parameter variables in extraction and storage;

[0094] Multi-objective optimization module: for calling a multi-objective optimization algorithm to construct a multi-objective optimization model, inputting all combinations of control parameter variables in extraction and storage into the multi-objective optimization model, selecting maximum cumulative production and maximum storage efficiency according to the optimization target, exporting Pareto optimal solution as multi-objective optimization result of extraction and storage.

[0095] The principle and advantages of the present application are that the technical scheme of the present application constructs a multi-objective optimization framework by fusing experimental design, numerical simulation, machine learning and intelligent optimization algorithm; specifically, first, the improved ant lion optimization algorithm is used to generate a control parameter combination sample by combining sampling processing, and the gas reservoir development dynamic data is obtained by combining CMG numerical simulation, to provide a high-quality data set for model training; then, the multi-output regression prediction model constructed based on the improved stacking algorithm learns the mapping relationship between the parameters and the development indexes (such as production, storage efficiency), to replace the time-consuming numerical simulation process; finally, the NSGA-II algorithm is used to search for Pareto optimal solution in the output space of the prediction model, to realize cumulative production and The core of the synergistic optimization of storage efficiency is to solve the complexity and coupling of multi-objective decision-making of gas reservoir development through "data-driven modeling + intelligent optimization search".

[0096] The application has the advantages that the gas reservoir injection The optimization efficiency and decision-making scientificity of extraction and storage are improved. The combination of CMG numerical simulation and improved ant lion optimization algorithm not only ensures the physical reliability of data, but also greatly reduces the calculation cost of the optimization process; the NSGA-II algorithm can effectively mine the nonlinear relationship between the control parameters and the double objectives, and the generated Pareto optimal solution set presents the optimal scheme under different objective weights, providing flexible selection space for engineering practice. Compared with the traditional single-objective optimization or experience decision-making, the method can improve the optimization accuracy under the same calculation resources, and simultaneously realize The synergistic improvement of recovery rate and storage rate, which has significant economic and environmental benefits. BRIEF DESCRIPTION OF DRAWINGS

[0097] Figure 1 The flowchart of the embodiment of the application is shown in the figure;

[0098] Figure 2 The function block diagram of the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0099] The following will be further described in detail through specific embodiments:

[0100] The embodiment is basically as shown in the figure: a multi-objective optimization method for gas reservoir injection Figure 1 extraction and storage, comprising:

[0101] S1: obtaining extraction and storage control parameter variables, calling a preset sampling algorithm to perform sampling processing and combination on the control parameter variables, and generating sample data; wherein S1 comprises:

[0102] S1-1: dividing the obtained extraction and storage control parameter variables into equal-probability subintervals according to their variable types, wherein the variable types include discrete variables and continuous variables;

[0103] S1-2: distributing a sampling quota to all division points of each variable according to the total sample number, and establishing a sampling index list of each variable;

[0104] S1-3: randomly shuffling the sampling index list of each variable, randomly selecting a position from the random index list of each variable, extracting the corresponding division point value to form a sampling combination, and repeating until the preset sample size is reached;

[0105] ​S1-4: Call the improved ant lion optimization algorithm to extract the optimal boundary enhanced combination of the extreme value of each variable, and after repeatability judgment with the sampling combination generated in S1-3, combine the sampling combination to generate the final variable sampling combination result.

[0106] In this embodiment, the control parameter variables include injection rate, concentration, injection-production ratio, production rate, for example, for the injection rate, the variable type is continuous variable, the value range is 100000-300000, and the step size is 10000; The variable type of the concentration is discrete variable, the numerical value is 0.6, 0.7, 0.8, 0.9, and 1.0; the variable type of the injection-production ratio is discrete variable, the numerical value is 0.6, 0.7, 0.8, 0.9, and 1.0; the variable type of the production rate is continuous, and the numerical value is determined according to the total number of samples.

[0107] The numerical values of each control parameter variable given above are only used as examples, and also to better represent the numerical value examples of different variable types in the control parameter variable;

[0108] Subsequently, the sample data is generated, in order to better comprehensively cover the value range of the control parameter variable, the sampling process first divides each variable into equal probability subintervals according to its value range, and different variable types are divided into different subintervals, specifically:

[0109] For discrete variables, no additional division is performed, and the discrete value set is directly used as the division result, for example The concentration is a discrete variable, and the numerical value is 0.6, 0.7, 0.8, 0.9, and 1.0, so it is directly divided according to the comma separation;

[0110] For continuous variables, linear division is used, for example, according to the set number of divisions (at least 2), the np.linspace function is used to generate equally spaced division points in The interval, when the minimum value is 0, the maximum value is 10, and the number of divisions is 5, then Such a division result is generated.

[0111] After the variable division is completed, the sampling quota is allocated to each variable according to the total number of samples, and the calculation method is to divide the total number of samples by the division points to obtain the quota, and if it cannot be divided, the remainder is randomly allocated to the first few division points.

[0112] Subsequently, the sampling index list of each variable is randomly shuffled, and the sampling index list contains the quota number of each division point, so as to ensure that the sampling position of each division point is randomly distributed, and to avoid systematic errors.

[0113] Then a position is randomly selected from the random index list of each variable, and the corresponding division point value is extracted to form a sample combination. Here, the combination is stored in a set to ensure uniqueness. If duplicate combinations are generated, resampling is performed until the specified sample size is reached.

[0114] After generating the random sample combination, the effectiveness of the sample data cannot be ensured at this time, so boundary reinforcement combinations are added. The improved lion-ant optimization algorithm is used to generate boundary reinforcement combinations, and the specific steps are as follows:

[0115] The extreme value of each control parameter variable is obtained , and a boundary target vector is constructed. For discrete variables, the minimum / maximum value in the value set can be directly selected. For continuous variables, the extreme value is determined according to the physical constraints and engineering upper limit of the gas reservoir;

[0116] Initialize the lion population and ant population, and add the boundary target vector to the lion population;

[0117] According to the near boundary and engineering feasible target, the fitness function is constructed, and the expression is as follows:

[0118]

[0119]

[0120]

[0121] Among them, is the boundary guiding term, is the ant individual, is the kth boundary target vector, is the Hamming distance, is the engineering feasibility constraint term, which is based on the physical laws of gas reservoirs and is converted into an algorithm that can be recognized by engineering experience and physical laws, is the penalty function, which is used to filter parameter combinations that are not feasible in engineering, such as wellbore leakage, pressure imbalance, etc. The more serious the violation of the constraint, the greater the penalty value; and are weight coefficients adjusted according to optimization requirements, and are both greater than 0 and add up to 1, for example, if the boundary is prioritized, take 0.6-0.8, and if the engineering feasibility is prioritized, take 0.7-0.9; the engineering feasibility constraint term Other representations in this application indicate that if x does not satisfy any of the engineering constraints; in this application, Hamming distance is introduced into the fitness function, where Hamming distance is used to measure the relationship between individual x and the boundary target vector. The degree of difference; the smaller the difference, the closer the individual is to the target boundary. The specific expression process is as follows:

[0122] Define dimension matching function :

[0123]

[0124] Indicates the precision threshold;

[0125] Individual x and the boundary target vector of the kth class Hamming distance:

[0126]

[0127] ,when This indicates that the individual perfectly matches the boundary target vector. This indicates a complete mismatch. This represents the total number of variables.

[0128] The ant's movement direction is adjusted by a boundary attraction factor to perform boundary-guided walking, expressed as follows:

[0129]

[0130]

[0131] in, Indicates in In the iteration, the ant's new position on the i-th control parameter variable, Indicates the first In the iteration, the ant's current position on the i-th control parameter variable is... Indicates the first The step size of the random walk in the iteration. For the first Generated in generation iteration Uniformly distributed random numbers Let i be the current value of the ant in the t-th generation on the i-th variable. For the boundary attraction term, Let be the i-th extreme value of the boundary target vector that the ant individual is currently closest to; The attraction intensity coefficient, This is used to avoid over-guided local optima; For individual ants The boundary target vector corresponding to the current minimum Hamming distance. Hamming distance between the current position x of the ant and the optimal boundary target vector is a distance attenuation factor, which represents that the closer the ant individual is to the boundary, the stronger the attraction is, and the farther away the weaker the attraction is;

[0132] After each iteration, the ant lion trap radius is updated based on the boundary guiding term

[0133]

[0134] wherein, is the initial trap radius, is the current ant lion of the tth generation, is the ant lion population of the tth generation; thus, when the boundary guiding term is smaller, the trap radius is smaller, and it is easier to capture ants near the boundary, thereby strengthening the local search of the boundary area;

[0135] After each iteration, the individual with the best fitness in the population is marked as an elite ant lion, specifically the top 5%-10% individuals with the smallest fitness values, and at least one elite individual that satisfies 1 variable extremum is forced to be retained, ensuring that the algorithm does not deviate from the boundary strengthening goal.

[0136] After the number of iterations reaches the maximum number of iterations, the optimal individual is extracted from the final elite ant lion population as the boundary strengthening combination, and in the present application, the optimal individual satisfies the following two conditions:

[0137] Condition 1: contains variable extremum; and condition 2: , indicating that the project is feasible.

[0138] Finally, a repeatability judgment is performed on the sample combination generated in S1-3, and the final sample data of the random sample combination and the de-duplicated boundary strengthening combination is output.

[0139] Thus, in addition to the sample combination obtained by random sampling, the present application also ensures the addition of the boundary strengthening combination, ensuring that the sample data covers the entire region of the parameter variable.

[0140] S2: Call the CMG data file batch modification tool to extract all well names and DATE commands from the original dat file, verify the effectiveness of the target control well and locate the position of the last DATE command; read the control parameter variables from the sample data, and perform injection-production ratio, ​​Concentration calculation; based on the CMG syntax rules to generate a new scheduling block, insert the new scheduling block into the last DATE command position of the original dat file, generate a new complete CMG data file, import the CMG software for simulation, export the CSV file in the CMG running result file and perform merging processing; wherein S2 comprises:

[0141] S2-1: reading the original dat file content, extracting all well names and DATE commands through regular expressions, verifying whether the well exists through the extracted well name to obtain the control parameter variable, and positioning to the last DATE command position of the original dat file;

[0142] S2-2: reading the control parameter variable in the sample data, performing Concentration calculation, injection-production ratio calculation, and generating a new scheduling block according to the CMG syntax rules combined with the injection well start date, end date, all monthly DATE commands, the PRODUCER and OPERATE rate control statements of the production well, the INJECTOR, INCOMP fluid concentration configuration and corresponding OPERATE statement of the injection well;

[0143] S2-3: truncating the original dat file to the last DATE command, inserting the generated new scheduling block, and then splicing the subsequent content of the original dat file to form a complete new file, generating a unique file name according to the preset file naming rule to batch generate CMG data files, and saving to a specified directory;

[0144] S2-4: importing the batch-modified generated CMG data file into the CMG software for running simulation, exporting the running simulation result as a CSV file, and performing data cleaning processing on the exported CSV file, including calling a preset regular expression matching format to match and convert the file name to a standard format and process invalid values, filtering data from the CSV file according to business rules, and calculating the core indicators of storage efficiency; wherein The storage efficiency expression is:

[0145]

[0146] Among them, represents storage efficiency, represents the cumulative injection molar number, represents the cumulative output molar number; if the injection is 0, then the storage efficiency is 0;

[0147] S2-5: Merge the CVS files after data cleaning into a single file, keep the table header of the first file, and only append data rows to the subsequent CSV files, to generate a merged CVS file containing all samples.

[0148] In this embodiment, the generated sample data is batch modified through the CMG data file. Specifically, the batch modification process includes the following three stages:

[0149] File parsing stage: first read the original dat file content, extract all well names and DATE commands through regular expressions; the extracted well names are used to verify whether the specified target optimized production wells and injection wells exist, and locate the position of the last DATE command in the original dat file, which is used as the insertion point of the subsequent new scheduling block; wherein:

[0150] Regular expression for extracting well names: ^\s*WELL\s+(['\"])([^'\"]+)\1\s*$

[0151] ^: Match the beginning of the string (or each line) (when combined with the re.MULTILINE flag, match the beginning of each line).

[0152] \s*: Match 0 or more whitespace characters (spaces, tabs, etc.).

[0153] WELL: Exactly match the string "WELL" (case-sensitive).

[0154] \s+: Match 1 or more whitespace characters.

[0155] (['\"]) : Match a single quote'or double quote ", and capture it as the 1st group (for subsequent reference).

[0156] ([^'\"]+): Match 1 or more non-quote characters ([^'\"] represents any character except'and "), captured as the 2nd group (i.e. well name).

[0157] \1: Reference the content matched by the 1st group (ensure that the well name has consistent quotes, e.g. if the front is a single quote, this also matches a single quote).

[0158] \s*: Match 0 or more whitespace characters.

[0159] $: Match the end of the string (or each line) (when combined with the re.MULTILINE flag, match the end of each line).

[0160] Regular expression for extracting DATE commands: ^\s*DATE\s+(\d+)\s+(\d+)\s+

[0161] ^: Matches the beginning of each line (in combination with the re.MULTILINE flag).

[0162] \s*: Matches zero or more whitespace characters.

[0163] DATE: Matches the string "DATE" exactly (case sensitive).

[0164] \s+: Matches one or more whitespace characters.

[0165] (\d+): Matches one or more digits (years), capturing as the first group.

[0166] \s+: Matches one or more whitespace characters.

[0167] (\d+): Matches one or more digits (months), capturing as the second group.

[0168] \s+: Matches one or more whitespace characters.

[0169] Parameter processing and content generation stage: Obtain production rate, injection rate, etc. from sample data. Concentration and other parameters are used for Concentration calculation and injection-production ratio calculation, among which, Concentration simulation calculation is done by subtracting 1 Concentration was obtained, and the injection-production ratio was calculated using simulation (injection rate). 2) (Production rate) 3) Once obtained, a new scheduling block is generated according to the CMG syntax rules, such as all monthly DATE commands from the start date (July 2024) to the end date (September 2025), PRODUCER and OPERATE rate control statements for production wells, INJECTOR and INCOMP fluid concentration configurations for injection wells (e.g., Concentration and The concentration) and the corresponding OPERATE statement generate a new scheduling block; in this application, Concentration calculations are used for generating new scheduling blocks, and injection-to-production ratios are used for file naming in the subsequent file reorganization phase.

[0170] File reassembly phase: The original .dat file is truncated to the last DATE command, a newly generated scheduling block is inserted, and the remaining content of the original .dat file is then concatenated to form a complete new file, and the reassembly is performed according to the following criteria: time range + injection rate + injection-to-collection ratio + ... The "concentration" rule generates unique filenames, and the sampling ratio in the filename is replaced with "p" to replace the decimal point, for example, 1.5 is replaced with 1p5. Concentrations are presented as integer percentages, e.g. 0.3 is replaced by 30, which converts these discrete parameters into identifiable file characteristics and compliant CMG command parameters, enabling batch generation.

[0171] After importing the batch-modified generated CMG data files into the CMG software for running simulation, exporting the running simulation results as CSV files, and performing parameter extraction and cleaning processing and merging processing on the CSV files, specifically, during processing, first extract key parameters such as injection rate, injection-production ratio, concentration, parse the numerical value by regular expression matching format, for example, regular expression such as `CMG_(\d+)-(\d+)_(\d+)([a-z]*)_(\d+p\d+)-( )?(\d+)`;then read the data in the CSV file, convert the numerical column to standard format and process invalid values, and then filter the data according to business rules, for example, keep records after July 2024 and mole fraction 0.4, and calculate core indicators, which are storage efficiency, the expression is:

[0172]

[0173] wherein, represents storage efficiency, represents the cumulative injection mole number, represents the cumulative output mole number; if the injection is 0, then the storage efficiency is 0;

[0174] Finally, perform merging processing to generate a merged file containing all samples, facilitating subsequent batch analysis.

[0175] S3: Construct a multi-output regression prediction model, pre-process the exported merged CSV file to generate a training set, a validation set and a test set, input the training set into the multi-output regression prediction model for training, and verify and test the trained multi-output regression prediction model through the validation set and the test set until a trained multi-output regression prediction model is obtained; wherein, S3 comprises:

[0176] S3-1: Construct a multi-output regression prediction model based on an improved stacked ensemble algorithm;

[0177] S3-2: After pre-processing the merged CVS file, divide it into a training set, a validation set and a test set according to 7:1.5:1.5.

[0178] S3-3: input the training set into the multi-output regression prediction model for training, specifically:

[0179] According to the feature vector of the i-th sample in the input training set , input into two base models: Random forest model expression:

[0180]

[0181] XGBoost model expression:

[0182]

[0183] Where, for the feature vector of the i-th sample in the training set

[0184] , the random forest model outputs the predicted value vector , and the XGBoost model outputs the predicted value vector , represents the predicted value of the cumulative yield in the random forest model, represents the predicted value of the cumulative yield in the random forest model, represents the predicted value of the mole fraction in the random forest model; represents the predicted value of the cumulative yield in the XGBoost model, represents the predicted value of the cumulative yield in the XGBoost model, represents the predicted value of the mole fraction in the XGBoost model; Concatenate the predicted results of the two base models into the meta-feature vector , the expression is:

[0185] The meta-model inputs the meta-feature and outputs the final predicted value

[0186] , the expression is:

[0187] where,

[0188]

[0189] ​​​​​​​​These are the final predicted results for the three dependent variables; express The final forecast of cumulative production, express The final forecast of cumulative production, express The final prediction result of the mole fraction;

[0190] The meta-model is optimized by minimizing the loss function on the training set, expressed as:

[0191]

[0192] in, For the first The true value of the t-th dependent variable in a sample, where n is the total number of samples. For the first The predicted value of the t-th dependent variable in a sample. Representing the metamodel;

[0193] To impose a business scope constraint on the prediction results, the expression is:

[0194]

[0195] in Let t represent the minimum value of the t-th dependent variable. Let t be the maximum value of the dependent variable. For the first The predicted value of the t-th dependent variable in a sample;

[0196] S3-4: Based on the training results, call the preset indicators for evaluation, and perform validation and testing based on the validation set and test set until a multi-output regression prediction model that meets the evaluation indicators is output.

[0197] In this embodiment, for the multi-output regression prediction model, the stacked ensemble algorithm is improved to enhance the accuracy of the prediction results. In addition, multi-output regression prediction models including neural networks, random forests, Transformers, and XGBoost can also be used.

[0198] For the training metrics of the multi-output regression prediction model, evaluation metrics such as MSE and RMSE are used for evaluation until the multi-output regression prediction model with the best prediction results is obtained.

[0199] S4: Extract All control parameters for extraction and storage are combined and exported as file data. A trained multi-output regression prediction model is then used to predict the file data. The regulatory parameter variables in extraction and storage are fully combined;

[0200] In the embodiment, the variable full combination example is: the date range is from 2024-08-01 to 2034-01-01, and the step size is 1 month; the injection rate is set to have a starting value of 100000, an ending value of 300000, and a step size of 10000; the injection-production ratio is set to have a starting value of 0.6, an ending value of 1.0, and a step size of 0.1; The concentration is input as 0.5, 0.6, 0.7, 0.8, 0.9, and 1; after full combination of the above variables, 71820 combination data are obtained; the combination data are exported as a variable combination table; the optimal multi-output regression prediction model is used for prediction; after obtaining the prediction result, data cleaning processing is performed to process invalid dates, correct unreasonable prediction values, and obtain the prediction result after data cleaning.

[0201] S5: calling An algorithm is used to construct a multi-objective optimization model, and The regulatory parameter variables in extraction and storage are fully combined and input into the multi-objective optimization model, and the optimization target is selected to maximize cumulative production and maximize storage efficiency. The multi-objective optimization result of extraction and storage; wherein, S5 includes:

[0202] S5-1: according to the full combination of the input regulatory parameter variables, injection rate, injection-production ratio, concentration, gas mole fraction based on date index are extracted;

[0203] S5-2: calling An algorithm is used to construct a multi-objective optimization model, and the variable influence coefficient formula is used to analyze the influence coefficient of injection rate, injection-production ratio, concentration, gas mole fraction;

[0204] S5-3: based on the injection rate, injection-production ratio, concentration, gas mole fraction after influence coefficient analysis, as the optimization object, input into the multi-objective optimization model, and the optimization target output is to maximize cumulative production and maximize storage efficiency.

[0205] In S5-2, the variable influence coefficient formula is used to analyze the influence coefficient of injection rate, injection-production ratio, concentration, gas mole fraction specifically as follows:

[0206] Injection rate impact coefficient analysis:

[0207]

[0208] wherein, represents the injection rate impact coefficient, is the injection rate weight, is the injection rate mean, is the injection rate standard deviation;

[0209] Injection-production ratio impact coefficient analysis:

[0210]

[0211] wherein, represents the injection-production ratio impact coefficient, is the injection-production ratio weight, is the injection-production ratio mean, is the injection-production ratio standard deviation;

[0212] Concentration impact coefficient analysis includes separately for production analysis and sequestration analysis:

[0213] Concentration on production analysis:

[0214]

[0215] Concentration on sequestration analysis:

[0216]

[0217] wherein, represents the concentration impact coefficient on production, represents the concentration impact coefficient on sequestration, is the concentration weight, is the concentration mean, is the concentration standard deviation; Gas mole fraction impact coefficient analysis:

[0218]

[0219]

[0220] wherein,​​​ represents the influence coefficient of the gas mole fraction, is the mean value of the mole fraction, is the standard deviation of the mole fraction.

[0221] The optimization objective in S5-3 is to maximize cumulative production and maximize The storage efficiency is specifically:

[0222] The optimization objective is to maximize The cumulative production expression is:

[0223]

[0224] The optimization objective is to maximize The storage efficiency expression is:

[0225]

[0226] wherein, is the final calculated cumulative production optimization objective value, represents the reference value of the cumulative production, is the final calculated storage efficiency optimization objective value, is the reference value of the storage efficiency, to avoid the result being 0 or negative.

[0227] Therefore, the technical scheme of the present application:

[0228] 1. Can realize full-process automatic operation, integrate functions such as sampling, CMG data file batch modification, CSV data processing and merging, multiple output regression prediction modeling and NSGA-II multi-objective optimization, and does not require a large amount of manual intervention from parameter design to optimization decision, greatly reducing labor costs, significantly improving work efficiency, reducing human operation errors, and ensuring result accuracy;

[0229] 2. Supports multi-objective collaborative optimization, generates Pareto optimal frontier solutions through the NSGA-II algorithm, and can find a balanced solution between mutually conflicting goals such as storage efficiency and methane cumulative production, solve the problem of balancing multiple objectives that traditional technologies cannot balance, and provide scientific parameter optimization basis for actual production;

[0230] ​3. Integrates multiple advanced algorithms and models, covers improved sampling algorithm (ensures uniform and boundary coverage of parameter samples), multiple multi-output regression prediction models (accurately captures data time series characteristics and complex nonlinear relationships, improves prediction accuracy), suitable for complex reservoir environment under multivariate scene, provides diversified technical selection for different needs, enhances system applicability and reliability;

[0231] 4. Has perfect data processing capability, processes invalid data through CSV data cleaning tool, converts numerical type and calculates storage efficiency, combines multi-output regression model result verification mechanism (such as correcting non-negative yield, limiting mole fraction range), guarantees data quality and prediction result rationality, provides high-quality data support for subsequent optimization decision;

[0232] The system can realize the coordinated optimization of gas reservoir injection storage and natural gas extraction, improve the efficiency of natural gas extraction, enhance the storage effect, provide strong technical support for the sustainable development of energy industry, and help realize the win-win of energy extraction and environmental protection.

[0233] As Figure 2 shown, in another embodiment of the present embodiment, a multi-objective optimization system for gas reservoir injection production and storage is also included, which is applied to the above-mentioned multi-objective optimization method for gas reservoir injection production and storage, comprising:

[0234] It includes a sample data generation module, a CMG processing module, a prediction model generation and training module, a variable combination module, and a multi-objective optimization module, wherein:

[0235] The sample data generation module is used to obtain the regulation and control parameter variables of production and storage, call the preset sampling algorithm to perform sampling processing and combination on the regulation and control parameter variables, and generate sample data;

[0236] The CMG processing module is used to call the CMG data file batch modification tool, extract all well names and DATE commands from the original dat file, verify the effectiveness of the target regulation and control well and locate the position of the last DATE command; read the regulation and control parameter variables from the sample data, and perform injection-production ratio, concentration calculation; then generate a new scheduling block based on the CMG syntax rules, insert the new scheduling block into the last DATE command position of the original dat file, generate a new complete CMG data file, import it into the CMG software for simulation, export the CSV file from the result file of the CMG running and perform merging processing;

[0237] The prediction model generation training module is configured to build a multi-output regression prediction model, generate a training set, a verification set and a test set after preprocessing the exported merged CSV file, input the training set into the multi-output regression prediction model for training, and verify and test the trained multi-output regression prediction model through the verification set and the test set until a trained multi-output regression prediction model is obtained.

[0238] The variable combination module is configured to extract all the control parameter variables of extraction and storage, combine the variables and export the combined variables as file data, call the trained multi-output regression prediction model to predict the file data, and obtain all combinations of the control parameter variables in extraction and storage.

[0239] The multi-objective optimization module is configured to call a multi-objective optimization algorithm to build a multi-objective optimization model, input all combinations of the control parameter variables in extraction and storage into the multi-objective optimization model, and select maximum cumulative production and maximum storage efficiency according to the optimization target, export a Pareto optimal solution as the multi-objective optimization result of extraction and storage.

[0240] The above is only an embodiment of the present application, and common knowledge such as specific structures and properties in the scheme is not described in detail here. The person skilled in the art knows all the common technical knowledge in the field of the present application before the filing date or the priority date, can know all the prior art in the field, and has the ability to apply conventional experimental means before that date. The person skilled in the art can improve and implement the present scheme based on their own ability under the guidance of the present application, and some typical known structures or known methods should not be an obstacle to the implementation of the present application by the person skilled in the art. It should be noted that for those skilled in the art, without departing from the structure of the present application, a number of modifications and improvements can be made, which should also be considered as the protection scope of the present application, and these will not affect the effect and practicality of the present application. The protection scope of the present application should be subject to the content of its claims, and the specific implementation mode and the like in the specification can be used to explain the content of the claims.​​

Claims

1. A gas reservoir injection A multi-objective optimization method for extraction and storage, characterized by: include: S1: Obtain The control parameter variables for extraction and storage are sampled and combined using a preset sampling algorithm to generate sample data. S2: Use the CMG data file batch modification tool to extract all well names and DATE commands from the original .dat file, verify the effectiveness of the target control wells, and locate the position of the last DATE command; read the control parameter variables from the sample data and perform injection-production ratio, Concentration calculation; then, based on the CMG syntax rules, a new scheduling block is generated and inserted into the last DATE command position of the original dat file. After generating a new complete CMG data file, it is imported into the CMG software for simulation. The CSV file is exported from the CMG running result file and merged. S3: Construct a multi-output regression prediction model. After preprocessing the exported merged CSV file, generate a training set, a validation set, and a test set. Input the training set into the multi-output regression prediction model for training. Validate and test the trained multi-output regression prediction model using the validation set and the test set until the trained multi-output regression prediction model is obtained. S4: Extract All control parameters for extraction and storage are combined and exported as file data. A trained multi-output regression prediction model is then used to predict the file data. All combinations of control parameters and variables in extraction and storage; S5: Use a multi-objective optimization algorithm to construct a multi-objective optimization model. The entire combination of control parameters and variables in extraction and storage is input into a multi-objective optimization model, and the maximization parameter is selected based on the optimization objective. Cumulative output and maximization Storage efficiency, deriving the Pareto optimal solution, as Multi-objective optimization results for extraction and storage; S1 includes: S1-1: For the acquisition The control parameters for extraction and storage are divided into equally probable sub-intervals according to their variable types, including discrete variables and continuous variables. S1-2: Allocate sampling quotas to all split points of each variable according to the total number of samples, and establish a sampling index list for each variable; S1-3: Randomly shuffle the sampling index list of each variable, randomly select a position from the random index list of each variable, extract the corresponding split point value to form a sampling combination, until the preset sample size is reached; S1-4: Use the improved Antlion optimization algorithm to extract the optimal boundary reinforcement combination of the extreme values ​​of each variable, and after performing a repeatability check on the sampling combination generated in S1-3, combine the sampling combination to generate the final variable sampling combination result.

2. A gas reservoir injection system according to claim 1 A multi-objective optimization method for extraction and storage, characterized by: The optimal boundary reinforcement combination for extracting the extreme values ​​of each variable in S1-4 by calling the improved antlion optimization algorithm is as follows: Obtain each control parameter variable extreme values and build Class boundary target vector; Initialize the antlion colony and the ant colony, and... The class boundary target vector is added to the antlion population; The fitness function is constructed based on proximity to the boundary and engineering feasibility objectives, and its expression is: in, For boundary-leading terms, This is a constraint on the feasibility of the project. For individual ants, Let k be the boundary target vector of the kth class. The distance is defined as Hamming distance, and the engineering constraints are preset based on the physical laws of gas reservoirs. For the penalty function; and The weighting coefficients are adjusted according to optimization requirements. and All are greater than 0, and their sum is 1; Project feasibility constraints Other representations in the text indicate that if x has one constraint that does not satisfy all engineering constraints; The ant's movement direction is adjusted by a boundary attraction factor to perform boundary-guided walking, expressed as follows: in, Indicates in In the iteration, the ant's new position on the i-th control parameter variable, Indicates the first In the iteration, the ant's current position on the i-th control parameter variable is... Indicates the first The step size of the random walk in the iteration. For the first Generated in generation iteration Uniformly distributed random numbers Let i be the current value of the ant in the t-th generation on the i-th variable. For the boundary attraction term, Let be the i-th extreme value of the boundary target vector that the ant individual is currently closest to; The attraction intensity coefficient, ; For individual ants The boundary target vector corresponding to the current minimum Hamming distance. Let x be the ant's current position and the optimal boundary target vector. Hamming distance, is the distance decay factor, which indicates that the closer an ant is to the boundary, the stronger the attraction, and the farther away it is, the weaker the attraction. After each iteration, based on the boundary guiding term Update the antlion trap radius, mark the individual with the best fitness in the population as an elite antlion, and force the retention of at least one individual that meets the criteria. An elite individual with an extreme value for a single variable; After the maximum number of iterations is reached, the best individuals are extracted from the final elite antlion population as the boundary reinforcement combination, and a repeatability judgment is performed on the sampled combination generated in S1-3. Finally, the final sample data of the random sampled combination and the deduplicated boundary reinforcement combination are output.

3. A gas reservoir injection system according to claim 1 A multi-objective optimization method for extraction and storage, characterized by: S2 includes: S2-1: Call the CMG data file batch modification tool, read the contents of the original dat file, extract all well names and DATE commands through regular expressions, verify the existence of wells with control parameter variables by using the extracted well names, and locate the position of the last DATE command in the original dat file; S2-2: Read the regulation parameter variables from the sample data and perform... Concentration calculation, injection-production ratio calculation, and according to CMG syntax rules, combine all monthly DATE commands for the start and end dates of injection wells, PRODUCER and OPERATE rate control statements for production wells, INJECTOR and INCOMP fluid concentration configurations for injection wells, and corresponding OPERATE statements to generate a new scheduling block; S2-3: Truncate the original dat file to the last DATE command, insert the generated new scheduling block, then concatenate the remaining content of the original dat file to form a complete new file, and generate unique filenames according to the preset file naming rules to batch generate CMG data files and save them to the specified directory.

4. A gas reservoir injection system according to claim 3 A multi-objective optimization method for extraction and storage, characterized by: S2 further includes: S2-4: After importing the batch-modified CMG data files into the CMG software for simulation, export the simulation results as a CSV file. Perform data cleaning on the exported CSV file, including using preset regular expression matching formats to match filenames and convert them to a standard format, handling invalid values, filtering data from the CSV file according to business rules, and calculating... The core metrics of storage efficiency; among them The storage efficiency expression is: in, express Storage efficiency, Indicates cumulative injection Number of moles Indicates cumulative output Number of moles; if injected If it is 0, then Storage efficiency is 0; S2-5: Merge the CVS files after data cleaning into a single file, retaining the header of the first file, and appending only data rows to subsequent CSV files to generate a merged CVS file containing all samples.

5. A gas reservoir injection system according to claim 1 A multi-objective optimization method for extraction and storage, characterized by: S3 includes: S3-1: Constructing a multi-output regression prediction model based on an improved stacked ensemble algorithm; S3-2: After preprocessing the merged CVS files, divide them into training, validation, and test sets in a ratio of 7:1.5:1.5; S3-3: Input the training set into the multi-output regression prediction model for training, specifically: Based on the input training set, the first... Feature vector of each sample Input into two base models: Random Forest model expression: XGBoost model expression: Among them, for the training set Feature vector of each sample Random Forest Model Output predicted value vector XGBoost model It will output a vector of predicted values. , In the random forest model Forecast of cumulative output, In the random forest model Forecast of cumulative output, In the random forest model Predicted value of mole fraction; In the XGBoost model Forecast of cumulative output, In the XGBoost model Forecast of cumulative output, In the XGBoost model Predicted value of mole fraction; The prediction results of the two base models are concatenated into a meta-feature vector. The expression is: Metamodel Meta-features As input, output the final predicted value. The expression is: in, These are the final predicted results for the three dependent variables; express The final forecast of cumulative production, express The final forecast of cumulative production, express The final prediction result of the mole fraction; The meta-model is optimized by minimizing the loss function on the training set, expressed as: in, For the first The true value of the t-th dependent variable in a sample, where n is the total number of samples. For the first The predicted value of the t-th dependent variable in a sample. Representing the metamodel; To impose a business scope constraint on the prediction results, the expression is: in Let t represent the minimum value of the t-th dependent variable. Let t be the maximum value of the dependent variable. For the first The predicted value of the t-th dependent variable in a sample. This represents the forecast value after adjustment for business scope constraints; S3-4: Based on the training results, call the preset indicators for evaluation, and perform validation and testing based on the validation set and test set until a multi-output regression prediction model that meets the evaluation indicators is output.

6. A gas reservoir injection system according to claim 1 A multi-objective optimization method for extraction and storage, characterized by: S5 includes: S5-1: Based on the complete combination of input control parameter variables, extract the injection rate, injection-to-enzyme ratio, and other parameters based on the date index. concentration, Gas mole fraction; S5-2: Call The algorithm constructs a multi-objective optimization model, calling the variable influence coefficient formula to evaluate injection rate, injection-to-enzyme ratio, and other variables. concentration, Influence coefficient analysis of gas mole fraction; S5-3: Injection rate and injection-production ratio based on influence coefficient analysis concentration, In a multi-objective optimization model, the gas mole fraction is used as the optimization target, and the output of the optimization objective is to maximize the gas mole fraction. Cumulative output and maximization Storage efficiency.

7. A gas reservoir injection system according to claim 6 A multi-objective optimization method for extraction and storage, characterized by: The formula for the influence coefficient of variables in S5-2 is used to apply to injection rate, injection-to-enzyme ratio, concentration, The influence coefficient analysis of gas mole fraction is as follows: Analysis of the influence coefficient of injection rate: in, Indicates the influence coefficient of injection rate. As the injection rate weight, The average injection rate. The standard deviation of the injection rate; Analysis of the influence coefficient of injection-production ratio: in, This represents the influence coefficient of the injection-production ratio. For the injection-production ratio weighting, The injection-production ratio is the average. The standard deviation of the injection-production ratio; Concentration influence coefficient analysis includes respectively for Production analysis and Sequestration Analysis: Concentration Production Analysis: Concentration Sequestration Analysis: in, express Concentration The impact coefficient of output express Concentration The impact coefficient of sealing for Concentration weight, for The average concentration, for Standard deviation of concentration; Analysis of the influence coefficient of gas mole fraction: in, express The influence coefficient of gas mole fraction for Mean of mole fraction denoted as the standard deviation of the mole fraction.

8. A gas reservoir injection system according to claim 7 A multi-objective optimization method for extraction and storage, characterized by: The optimization objective output in S5-3 is to maximize... Cumulative output and maximization Storage efficiency is specifically as follows: The optimization objective is to maximize The cumulative output expression is: Maximize the optimization objective The storage efficiency expression is: in, For the final calculation Cumulative production optimization target value, express The baseline value for cumulative production. For the final calculation Storage efficiency optimization target value for Storage efficiency benchmark, Avoid results that are 0 or negative.

9. A gas reservoir injection A multi-objective optimization system for extraction and storage, applying the gas reservoir injection method described in any one of claims 1-8 above. A multi-objective optimization method for extraction and storage, characterized by: It includes a sample data generation module, a CMG processing module, a prediction model generation and training module, a variable combination module, and a multi-objective optimization module, among which: Sample data generation module: used to acquire The control parameter variables for extraction and storage are sampled and combined using a preset sampling algorithm to generate sample data. CMG processing module: Used to call the CMG data file batch modification tool, extract all well names and DATE commands from the original .dat file, verify the effectiveness of the target control wells, and locate the position of the last DATE command; read control parameter variables from sample data, and perform injection-production ratio, Concentration calculation; then, based on the CMG syntax rules, a new scheduling block is generated and inserted into the last DATE command position of the original dat file. After generating a new complete CMG data file, it is imported into the CMG software for simulation. The CSV file is exported from the CMG running result file and merged. Prediction Model Generation and Training Module: This module is used to build a multi-output regression prediction model. It preprocesses the exported merged CSV file to generate training, validation, and test sets. The training set is then input into the multi-output regression prediction model for training. The training multi-output regression prediction model is validated and tested using the validation and test sets until the trained multi-output regression prediction model is obtained. Variable combination module: used for extraction All control parameters for extraction and storage are combined and exported as file data. A trained multi-output regression prediction model is then used to predict the file data. All combinations of control parameters and variables in extraction and storage; Multi-objective optimization module: Used to call multi-objective optimization algorithms to build multi-objective optimization models. The entire combination of control parameters and variables in extraction and storage is input into a multi-objective optimization model, and the maximization parameter is selected based on the optimization objective. Cumulative output and maximization Storage efficiency, deriving the Pareto optimal solution, as The results of multi-objective optimization of extraction and storage.

Images