Method, equipment and medium for predicting geological storage quantity of carbon dioxide

Abstract

The invention discloses a method and device for predicting the geological storage amount of carbon dioxide and a medium, and relates to the technical field of carbon dioxide capture, utilization and burying. The method comprises the steps that whether a to-be-predicted oil and gas reservoir and a control oil and gas reservoir are in the same block or not is judged; if yes, inputting the geological data, the production dynamic data and the injection parameter combination scheme of the to-be-predicted oil and gas reservoir into the optimal control oil and gas reservoir carbon dioxide geological sequestration quantity prediction model to determine the whole-area accumulated carbon dioxide geological sequestration quantity of the to-be-predicted oil and gas reservoir; if not, geological data, production data and an injection parameter value range of the to-be-predicted oil and gas reservoir are obtained, migration training is conducted on the optimal control oil and gas reservoir carbon dioxide geological sequestration quantity prediction model, and a to-be-predicted oil and gas reservoir carbon dioxide geological sequestration quantity prediction model is obtained; and determining the whole-area accumulated carbon dioxide geological sequestration amount of the to-be-predicted oil and gas reservoir. According to the invention, the precision and efficiency of carbon dioxide geological sequestration quantity prediction are improved.

Description

Technical Field

[0001] The present application relates to the technical field of carbon dioxide capture, utilization and storage, and in particular to a method, equipment and medium for predicting the geological storage amount of carbon dioxide. Background Art

[0002] After a period of development, due to the limitations of the technology and economic conditions at the time, some crude oil cannot be extracted from oil and gas reservoirs, thus losing their mining value. These oil and gas reservoirs are often called abandoned oil and gas reservoirs. The geological storage of carbon dioxide through abandoned oil and gas reservoirs can, on the one hand, make full use of the existing oil and gas reservoir exploration and development data, well sites and injection equipment, saving investment and engineering time; on the other hand, the industrial carbon dioxide emissions can be stored in abandoned oil and gas reservoirs, which can effectively reduce carbon dioxide emissions. Accurately predicting the amount of geological storage of carbon dioxide is of great significance for the screening of abandoned oil and gas reservoirs and the evaluation of storage sites.

[0003] The commonly used method for predicting the geological storage of carbon dioxide is the numerical simulation method, that is, for the target block, the geological parameters and injection parameters are set according to the field experience of the oil field, and the numerical simulation of the reservoir is carried out to obtain the geological storage of carbon dioxide. This method requires a large amount of data, but much of the data is difficult to obtain at the oil field site; and the numerical simulation needs to be repeated for each block and each set of injection parameters, which is very time-consuming and has poor operability. Summary of the invention

[0004] The purpose of this application is to provide a method, device and medium for predicting the geological storage amount of carbon dioxide, which can improve the accuracy and efficiency of predicting the geological storage amount of carbon dioxide.

[0005] To achieve the above objectives, this application provides the following solutions: In a first aspect, the present application provides a method for predicting the amount of geological storage of carbon dioxide, comprising: Determine whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are in the same block; If so, the geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted are input into the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model to obtain the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted; the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model is input with the geological data, production dynamic data and injection parameter combination scheme of the reference oil and gas reservoir, and the cumulative carbon dioxide geological storage capacity of the entire area of ​​the reference oil and gas reservoir is output, and the hyperparameters in the multi-feature embedded carbon dioxide geological storage prediction model are optimized by using an optimization algorithm, and the multi-feature embedded carbon dioxide geological storage prediction model is trained multiple times to obtain the model; If not, then obtain the geological data, production performance data and injection parameter value range of the oil and gas reservoir to be predicted; Based on the geological data and production performance data of the oil and gas reservoirs to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoirs to be predicted is constructed; Sampling once within the range of CO2 injection parameter values ​​for the oil and gas reservoir to be predicted, and obtaining multiple groups of CO2 injection parameter combination schemes for the oil and gas reservoir to be predicted; the CO2 injection parameters include but are not limited to: CO2 injection amount, CO2 injection rate, CO2 injection time, CO2 injection pressure, CO2 injection slug size, water injection / gas injection ratio, and well soaking time; Input multiple groups of injection parameter combinations of the oil and gas reservoirs to be predicted into the numerical simulation model of carbon dioxide storage in the oil and gas reservoirs to be predicted, and obtain the cumulative carbon dioxide geological storage volume of the whole area corresponding to different injection parameter combinations of the oil and gas reservoirs to be predicted; Taking the geological data and production dynamic data of the oil and gas reservoir to be predicted and the injection parameter combination scheme of the oil and gas reservoir to be predicted as input, and taking the cumulative carbon dioxide geological storage volume of the whole area corresponding to the injection parameter combination scheme of the oil and gas reservoir to be predicted as output, the optimal control oil and gas reservoir carbon dioxide geological storage volume prediction model is transferred and trained to obtain the carbon dioxide geological storage volume prediction model of the oil and gas reservoir to be predicted; The geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted are input into the prediction model of carbon dioxide geological storage capacity of the oil and gas reservoir to be predicted, and the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted is obtained.

[0006] Optionally, the multi-feature embedded carbon dioxide geological storage capacity prediction model includes a physical feature layer and a multi-feature embedded prediction layer connected in sequence.

[0007] Optionally, the physical feature layer includes: a geological data feature sequence extraction branch, a production dynamic data feature sequence extraction branch, a parameter feature sequence extraction branch, a trend feature sequence extraction branch and a splicing layer; the output end of the production dynamic data feature sequence extraction branch, the output end of the geological data feature sequence extraction branch, the output end of the parameter feature sequence extraction branch and the output end of the trend feature sequence extraction branch are all connected to the input end of the splicing layer; The geological data feature sequence extraction branch includes a geological encoder; the input of the geological data feature sequence extraction branch is the geological data of the oil and gas reservoir to be predicted; The production dynamic data feature sequence extraction branch includes a production encoder; the input of the production dynamic data feature sequence extraction branch is the production dynamic data of the oil and gas reservoir to be predicted; The parameter feature sequence extraction branch includes a value encoder and a position encoder connected in parallel; the input of the parameter feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted; The trend feature sequence extraction branch includes a trend encoder; the input of the trend feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted; The production dynamic data feature sequence extraction branch and the geological data feature sequence extraction branch respectively extract the overall characteristics of all parameters in the production dynamic data and geological data, the parameter feature sequence extraction branch extracts the individual characteristics of any parameter in the production dynamic data and geological data, and the trend feature sequence extraction branch extracts the trend characteristics of any parameter in the production dynamic data and geological data; the splicing layer is spliced ​​by the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch in a horizontal splicing manner, and the output end of the splicing layer is connected to the multi-feature embedding prediction layer; the horizontal splicing method aligns the heads of the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch, and splices the aligned extraction branches in sequence from left to right to obtain a splicing layer.

[0008] Optionally, the multi-feature embedding prediction layer includes: The branch at time t, the branch at time t+1, the first fully connected layer and the activation function are connected in sequence; The input end of the branch at time t and the input end of the branch at time t+1 are both connected to the output end of the physical feature layer; The input end of the branch at time t and the input end of the branch at time t+1 are also used to input a combination scheme of injection parameters of the oil and gas reservoir to be predicted.

[0009] Optionally, the branch at time t includes: a first multi-head attention layer, a first residual layer, a first regularization layer, a second fully connected layer, a second residual layer, and a second regularization layer connected in sequence; The branch at time t+1 includes: a second multi-head attention layer, a third residual layer, a third regularization layer, a third fully connected layer, a fourth residual layer and a fourth regularization layer connected in sequence; The input end of the first multi-head attention layer and the input end of the first regularization layer are used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t in the physical feature sequence; the physical feature sequence is obtained by splicing the production dynamic data feature sequence, the geological data feature sequence, the parameter feature sequence and the trend feature sequence; The output end of the first regularization layer is also connected to the input end of the second regularization layer; The output end of the second regularization layer is also connected to the input end of the third regularization layer; The input end of the second multi-head attention layer is used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t+1 in the physical feature sequence; The output end of the third regularization layer is also connected to the input end of the fourth regularization layer; The output end of the fourth regularization layer is also connected to the input end of the first fully connected layer. Optionally, before determining whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are the same block, the method further includes: Obtain geological data, production performance data and injection parameter value ranges of reference oil and gas reservoirs; Based on the geological data and production performance data of the reference oil and gas reservoirs, a numerical simulation model for CO2 storage in the reference oil and gas reservoirs is constructed; Secondary sampling is performed within the range of CO2 injection parameter values ​​to obtain multiple groups of CO2 injection parameter combination schemes for control oil and gas reservoirs; Inputting a plurality of control oil and gas reservoir injection parameter combination schemes into the control oil and gas reservoir carbon dioxide storage numerical simulation model respectively, and obtaining the cumulative carbon dioxide geological storage volume of the whole area corresponding to different control oil and gas reservoir injection parameter combination schemes; A small sample data set was constructed by taking the injection parameter combination scheme of the control oil and gas reservoir as input and the cumulative carbon dioxide geological storage volume in the whole area corresponding to different injection parameter combination schemes of the control oil and gas reservoir as output. The small sample data set is expanded using the small perturbation data to obtain an expanded sample data set; the expansion processing includes but is not limited to the small perturbation data that conforms to the Gaussian distribution, uniform distribution, and Laplace distribution; the small sample data set and the expanded sample data set are spliced ​​in a vertical splicing manner to obtain a training data set; the vertical splicing manner aligns the left sides of the expanded sample data set and the small sample data set, and the aligned data sets are stacked in order from top to bottom to obtain the training data set; Taking the geological data and production dynamic data of the reference oil and gas reservoirs and the injection parameter combination scheme of the reference oil and gas reservoirs in the training data set as input, and taking the cumulative carbon dioxide geological storage volume of the whole area in the training data set as output, the multi-feature embedded carbon dioxide geological storage volume prediction model is trained multiple times and the hyperparameters in the multi-feature embedded carbon dioxide geological storage volume prediction model are optimized by using an optimization algorithm, and the evaluation index of the reference oil and gas reservoir carbon dioxide geological storage volume prediction model obtained by each training is determined, and the optimal reference oil and gas reservoir carbon dioxide geological storage volume prediction model is obtained based on the evaluation index; the evaluation index includes but is not limited to the perturbation fit goodness, the storage process error under small perturbation and the average absolute error of geological storage volume under small perturbation; The perturbation goodness of fit is: ; The error of the sealing process under small disturbance is: ; The average absolute error of geological storage under small disturbance is: ; Where CCSR is the perturbation fit goodness, CCSME is the storage process error under small disturbance, CCSMA is the mean absolute error of geological storage under small disturbance, N is the number of sample sets in the training data set, i is the sample set number, j is the sample sequence number under a certain sample set, and N is the number of samples in the training data set. j is the total number of samples in the i-th sample set, is the prediction result of carbon dioxide geological storage at the jth sample in the i-th sample set at time t, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample of the i-th sample set at time t, is the arithmetic mean of the numerical simulation results of the geological storage of carbon dioxide of the jth sample in the ith sample set within the carbon dioxide injection time, num is the total number of samples, Total is the carbon dioxide injection time, t is the single time step of the carbon dioxide injection time, is the distribution function corresponding to the small perturbation data of the i-th sample set, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample Total time of the i-th sample set, It is the prediction result of the geological storage amount of carbon dioxide at the jth sample Total time in the i-th sample set.

[0010] Optionally, based on the geological data and production performance data of the oil and gas reservoir to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoir to be predicted is constructed, including: Based on the geological data of the oil and gas reservoir to be predicted, a geological model of the oil and gas reservoir to be predicted is constructed using geological modeling software; Import the geological model of the oil and gas reservoir to be predicted into the oil reservoir numerical simulation software, use the production performance data to perform historical matching on the production performance of the oil and gas reservoir to be predicted, and obtain the development numerical simulation model of the oil and gas reservoir to be predicted; Based on the numerical simulation model of the development of the oil and gas reservoir to be predicted, a numerical simulation of carbon dioxide storage in the oil and gas reservoir to be predicted is performed to obtain a numerical simulation model of carbon dioxide storage in the oil and gas reservoir to be predicted.

[0011] Optionally, the number of samplings for the primary sampling is 5-10% of the number of samplings for the secondary sampling.

[0012] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned method for predicting the geological storage amount of carbon dioxide.

[0013] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for predicting the geological storage amount of carbon dioxide.

[0014] According to the specific embodiments provided in this application, this application discloses the following technical effects: The present application provides a method, device and medium for predicting the geological storage amount of carbon dioxide. A numerical simulation model for carbon dioxide storage is established through geological data and production dynamic data. Different injection parameter combination schemes are composed through a sampling method. Numerical simulation is performed based on the different injection parameter combination schemes to obtain the cumulative geological storage amount of carbon dioxide in the entire area. The geological data, production dynamic data and the cumulative geological storage amount of carbon dioxide in the entire area are used to train a multi-feature embedded carbon dioxide geological storage prediction model. The trained multi-feature embedded carbon dioxide geological storage prediction model is combined with transfer learning to predict the cumulative geological storage amount of carbon dioxide in the entire area of ​​the oil and gas reservoir to be predicted.

[0015] Using the multi-feature embedded carbon dioxide geological storage prediction model trained by this application, the cumulative carbon dioxide geological storage prediction of other oil and gas reservoirs can be achieved efficiently, quickly and accurately through transfer learning. For a given oil and gas reservoir, if the existing method is used to predict the geological storage of carbon dioxide, it is necessary to rebuild the numerical simulation model and perform numerical simulation to obtain it. It not only requires a large amount of data, but also takes time and effort. When the target oil and gas reservoir or injection parameters change, it is necessary to re-carry out numerical simulation, so the model has poor portability. However, using the method provided by this application, it is only necessary to add 5% to 10% of new samples to the model obtained by training the existing oil and gas reservoir for retraining, and a better prediction effect can be obtained, and the trained model can directly replace the numerical simulation for carbon dioxide storage prediction, and no numerical simulation is required. Therefore, the generalization ability of this application is excellent, and the computational efficiency is significantly improved.

[0016] The present application is based on the above-mentioned method and system for predicting the geological storage volume of carbon dioxide. The geological data and production dynamic data features of abandoned oil and gas reservoirs are extracted through multiple encoders such as geological encoders, production encoders and trend encoders, and embedded into the carbon dioxide geological storage volume prediction model as a feature embedding layer, which retains the original information while giving it physical meaning. The hyperparameters in the multi-feature embedded carbon dioxide address storage volume prediction model are optimized through a hyperparameter optimization method, so that when the cumulative carbon dioxide geological storage volume prediction task of the entire area is carried out for different target oil reservoirs, the model prediction accuracy can still be guaranteed. Therefore, the present application is highly robust. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 This is a flow chart of a method for predicting the amount of geological storage of carbon dioxide in one embodiment of the present application; Figure 2 This is a schematic diagram of a method for predicting the amount of geological storage of carbon dioxide in one embodiment of the present application; Figure 3 This is a schematic diagram of a physical characteristic layer in an embodiment of the present application; Figure 4 This is a schematic diagram of a multi-feature embedding prediction layer in an embodiment of the present application; Figure 5 This is a 45° intersection diagram of the predicted value and the actual value of the geological storage amount of carbon dioxide in one embodiment of the present application; Figure 6 This is a schematic diagram of a device for predicting the geological storage amount of carbon dioxide in one embodiment of the present application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0021] In an exemplary embodiment, Figure 1 As shown, a method for predicting the amount of geological storage of carbon dioxide is provided, comprising: Step 101: Determine whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are in the same block.

[0022] Step 102: If yes, the geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted are input into the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model to obtain the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted. The optimal reference oil and gas reservoir carbon dioxide geological storage prediction model is obtained by taking the geological data, production dynamic data and injection parameter combination scheme of the reference oil and gas reservoir as input and the cumulative carbon dioxide geological storage capacity of the entire area of ​​the reference oil and gas reservoir as output, and optimizing the hyper parameters in the multi-feature embedded carbon dioxide geological storage prediction model by using an optimization algorithm, and training the multi-feature embedded carbon dioxide geological storage prediction model multiple times.

[0023] Step 103: If not, then obtain the geological data, production performance data and injection parameter value range of the oil and gas reservoir to be predicted.

[0024] Step 104: Based on the geological data and production performance data of the oil and gas reservoir to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoir to be predicted is constructed.

[0025] Step 105: Sampling is performed once within the range of carbon dioxide injection parameter values ​​of the oil and gas reservoir to be predicted to obtain multiple groups of carbon dioxide injection parameter combination schemes for the oil and gas reservoir to be predicted; the carbon dioxide injection parameters include but are not limited to: carbon dioxide injection amount, carbon dioxide injection rate, carbon dioxide injection time, carbon dioxide injection pressure, injected carbon dioxide segment plug size, water injection / gas injection ratio and well soaking time.

[0026] Step 106: Input multiple groups of injection parameter combinations of the oil and gas reservoirs to be predicted into the CO2 storage numerical simulation model of the oil and gas reservoirs to be predicted, and obtain the cumulative CO2 geological storage capacity of the whole area corresponding to different injection parameter combinations of the oil and gas reservoirs to be predicted.

[0027] Step 107: Taking the geological data and production dynamic data of the oil and gas reservoir to be predicted and the injection parameter combination scheme of the oil and gas reservoir to be predicted as input, and taking the cumulative carbon dioxide geological storage capacity in the whole area corresponding to the injection parameter combination scheme of the oil and gas reservoir to be predicted as output, migration training is performed on the carbon dioxide geological storage capacity prediction model of the optimal control oil and gas reservoir to obtain the carbon dioxide geological storage capacity prediction model of the oil and gas reservoir to be predicted.

[0028] Step 108: Input the geological data, production performance data and injection parameter combination scheme of the oil and gas reservoir to be predicted into the prediction model of carbon dioxide geological storage capacity of the oil and gas reservoir to be predicted, and obtain the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted.

[0029] The multi-feature embedded carbon dioxide geological storage prediction model includes a physical feature layer and a multi-feature embedded prediction layer connected in sequence. Figure 3 The physical feature layer includes: a geological data feature sequence extraction branch, a production dynamic data feature sequence extraction branch, a parameter feature sequence extraction branch, a trend feature sequence extraction branch and a splicing layer; the output end of the production dynamic data feature sequence extraction branch, the output end of the geological data feature sequence extraction branch, the output end of the parameter feature sequence extraction branch and the output end of the trend feature sequence extraction branch are all connected to the input end of the splicing layer.

[0030] The geological data feature sequence extraction branch includes a geological encoder; the input of the geological data feature sequence extraction branch is the geological data of the oil and gas reservoir to be predicted.

[0031] The production dynamic data feature sequence extraction branch includes a production encoder. The input of the production dynamic data feature sequence extraction branch is the production dynamic data of the oil and gas reservoir to be predicted.

[0032] The parameter feature sequence extraction branch includes a parallel value encoder and a position encoder. The input of the parameter feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted.

[0033] The trend feature sequence extraction branch includes a trend encoder. The input of the trend feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted. The production dynamic data feature sequence extraction branch and the geological data feature sequence extraction branch extract the overall characteristics of all parameters in the production dynamic data and geological data respectively, the parameter feature sequence extraction branch extracts the individual characteristics of any parameter in the production dynamic data and geological data, and the trend feature sequence extraction branch extracts the trend characteristics of any parameter in the production dynamic data and geological data. The splicing layer is spliced ​​by the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch in a horizontal splicing manner, and the output end of the splicing layer is connected to the multi-feature embedding prediction layer. The horizontal splicing method aligns the heads of the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch, and splices the aligned extraction branches in order from left to right to obtain the splicing layer.

[0034] like Figure 4 , the multi-feature embedding prediction layer includes: a sequentially connected branch at time t, a branch at time t+1, the first fully connected layer, and an activation function.

[0035] The input end of the branch at time t and the input end of the branch at time t+1 are both connected to the output end of the physical feature layer.

[0036] The input end of the branch at time t and the input end of the branch at time t+1 are also used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted.

[0037] The branch at time t includes: the first multi-head attention layer, the first residual layer, the first regularization layer, the second fully connected layer, the second residual layer and the second regularization layer connected in sequence.

[0038] The branch at time t+1 includes: the second multi-head attention layer, the third residual layer, the third regularization layer, the third fully connected layer, the fourth residual layer and the fourth regularization layer connected in sequence.

[0039] The input end of the first multi-head attention layer and the input end of the first regularization layer are used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t in the physical feature sequence. The physical feature sequence is obtained by splicing the production dynamic data feature sequence, geological data feature sequence, parameter feature sequence and trend feature sequence.

[0040] The output of the first regularization layer is also connected to the input of the second regularization layer.

[0041] The output of the second regularization layer is also connected to the input of the third regularization layer.

[0042] The input end of the second multi-head attention layer is used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t+1 in the physical feature sequence.

[0043] The output terminal of the third regularization layer is also connected to the input terminal of the fourth regularization layer.

[0044] The output of the fourth regularization layer is also connected to the input of the first fully connected layer.

[0045] Before step 101, the method further includes: Obtain geological data, production performance data and injection parameter value ranges of the reference oil and gas reservoirs.

[0046] Based on the geological data and production performance data of the reference oil and gas reservoirs, a numerical simulation model for carbon dioxide storage in the reference oil and gas reservoirs is constructed.

[0047] Secondary sampling is carried out within the range of CO2 injection parameter values ​​to obtain multiple groups of CO2 injection parameter combination schemes for control oil and gas reservoirs. The number of samplings for primary sampling is 5-10% of the number of samplings for secondary sampling.

[0048] Multiple groups of control oil and gas reservoir injection parameter combination schemes were input into the numerical simulation model of carbon dioxide storage in the control oil and gas reservoirs, and the cumulative carbon dioxide geological storage capacity of the whole area corresponding to different control oil and gas reservoir injection parameter combination schemes was obtained.

[0049] A small sample data set was constructed with the injection parameter combination scheme of the control oil and gas reservoirs as input and the cumulative carbon dioxide geological storage capacity in the whole area corresponding to different injection parameter combination schemes of the control oil and gas reservoirs as output.

[0050] The small sample data set is expanded using the slightly disturbed data to obtain an expanded sample data set. The expansion processing includes but is not limited to slightly disturbed data that conforms to Gaussian distribution, uniform distribution, and Laplace distribution. The small sample data set and the expanded sample data set are spliced ​​in a vertical splicing manner to obtain a training data set. The vertical splicing method aligns the left sides of the expanded sample data set and the small sample data set, and the aligned data sets are stacked in order from top to bottom to obtain the training data set. Taking the geological data and production dynamic data of the reference oil and gas reservoirs, as well as the injection parameter combination scheme of the reference oil and gas reservoirs in the training data set as input, and the cumulative carbon dioxide geological storage in the whole area in the training data set as output, the multi-feature embedded carbon dioxide geological storage prediction model is trained multiple times and the optimization algorithm is used to optimize the hyperparameters in the multi-feature embedded carbon dioxide geological storage prediction model, and the evaluation index of the reference oil and gas reservoir carbon dioxide geological storage prediction model obtained by each training is determined, and the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model is obtained based on the evaluation index. Evaluation indicators include but are not limited to disturbance fitting goodness, storage process error under slight disturbance and average absolute error of geological storage volume under slight disturbance.

[0051] The perturbation goodness of fit is: .

[0052] The error of the sealing process under small disturbance is: .

[0053] The average absolute error of geological storage under small disturbance is: .

[0054] Where CCSR is the perturbation fit goodness, CCSME is the storage process error under small disturbance, CCSMA is the mean absolute error of geological storage under small disturbance, N is the number of sample sets in the training data set, i is the sample set number, j is the sample sequence number under a certain sample set, and N is the number of samples in the training data set. j is the total number of samples in the i-th sample set, is the prediction result of carbon dioxide geological storage at the jth sample in the i-th sample set at time t, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample of the i-th sample set at time t, is the arithmetic mean of the numerical simulation results of the geological storage of carbon dioxide of the jth sample in the ith sample set within the carbon dioxide injection time, num is the total number of samples, Total is the carbon dioxide injection time, t is the single time step of the carbon dioxide injection time, is the distribution function corresponding to the small perturbation data of the i-th sample set, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample Total time of the i-th sample set, It is the prediction result of the geological storage amount of carbon dioxide at the jth sample Total time in the i-th sample set.

[0055] Based on the geological data and production dynamic data of the oil and gas reservoir to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoir to be predicted is constructed, including: constructing a geological model of the oil and gas reservoir to be predicted using geological modeling software according to the geological data of the oil and gas reservoir to be predicted. Importing the geological model of the oil and gas reservoir to be predicted into the reservoir numerical simulation software, performing historical fitting on the production dynamics of the oil and gas reservoir to be predicted using the production dynamic data, and obtaining a numerical simulation model for the development of the oil and gas reservoir to be predicted. Performing numerical simulation of carbon dioxide storage in the oil and gas reservoir to be predicted based on the numerical simulation model for the development of the oil and gas reservoir to be predicted, and obtaining a numerical simulation model for carbon dioxide storage in the oil and gas reservoir to be predicted.

[0056] Based on the range of CO2 injection parameters of the predicted oil and gas reservoirs, a small sample data set for training the multi-feature embedded CO2 geological storage prediction model is constructed using a sampling method. A small sample data set refers to a data set that uses the geological data and production dynamic data of the predicted oil and gas reservoirs, the CO2 injection parameter combination scheme obtained by the sampling method according to the injection parameter value range as input data, and the cumulative CO2 geological storage in the entire area obtained by numerical simulation using the CO2 storage numerical simulation model of the predicted oil and gas reservoirs as output data. The size of the small sample data set varies according to the range of injection parameters.

[0057] Furthermore, the steps of determining the small sample data set include: S12. Within the range of the CO2 injection parameters, a sampling method is used to perform a sampling, and each sampling forms a set of CO2 injection parameter combination schemes. The sampling methods include but are not limited to importance sampling, Latin hypercube sampling, Markov chain Monte Carlo sampling, random sampling, and stratified sampling. The injection parameters include but are not limited to: CO2 injection amount, CO2 injection rate, CO2 injection time, CO2 injection pressure, CO2 injection slug size, water injection / gas injection ratio, and well soaking time.

[0058] S13. According to the set sampling times, a CO2 injection parameter sample set is formed, and the CO2 storage numerical simulation model of the predicted oil and gas reservoir is called to perform batch numerical simulation. The CO2 injection parameters are set according to the CO2 injection parameter sample set for simulation to obtain the cumulative CO2 geological storage capacity of the entire area.

[0059] S14. A small sample data set DataSet-1 for training a multi-feature embedded carbon dioxide geological storage prediction model is constructed using each set of carbon dioxide injection parameters and geological data, production dynamic data, and the cumulative carbon dioxide geological storage capacity in the entire region obtained by numerical simulation.

[0060] The constructed small sample data set is expanded using a data augmentation method to obtain a sample data set for training a multi-feature embedded carbon dioxide geological storage prediction model. Specifically, the constructed small sample data set is expanded. First, a set of slight disturbance data is generated. Then, the set of slight disturbance data is added to the input carbon dioxide injection parameters and geological parameters. Further, slight disturbance data is added to each set of carbon dioxide injection parameters in S12. The order of magnitude of the slight disturbance data is much smaller than 2 orders of magnitude of the original data.

[0061] Data augmentation methods include but are not limited to slightly disturbed data that conform to Gaussian distribution, uniform distribution, and Laplace distribution. Mathematical reasoning formulas can be expressed as: (10) in, is the data after adding a small disturbance, is the small sample data constructed in S14, It is expressed in three different ways. When i=1, it is slightly perturbed data that conforms to Gaussian distribution, when i=2, it is slightly perturbed data that conforms to uniform distribution, and when i=3, it is slightly perturbed data that conforms to Laplace distribution. That is, after generating slightly perturbed data, add it to the small sample data to get a new sample set.

[0062] S21. Add small perturbation data that conforms to Gaussian distribution. It follows a Gaussian distribution Random small perturbation data with mean 0 and variance A random value of . The probability density function of is as follows: (11) After generating random small perturbation data that obeys Gaussian distribution, add it to the small sample data constructed in S13 to generate a class of sample set DataSet-2.

[0063] S22. Add small disturbance data that conforms to uniform distribution. It is uniformly distributed Random small perturbation data, that is, the value is A random value between . The probability density function of is as follows: where parameter a will be automatically adjusted according to the specific data size rather than a fixed value. For example, if in a set of carbon dioxide injection parameters, the carbon dioxide injection pressure is on the order of 10 1 , then the parameter a can be -0.05-0.05.

[0064] (12) After generating random small perturbation data that obeys uniform distribution, add it to the small sample data constructed in S13 to generate the second category sample set DataSet-3.

[0065] S23. Add small perturbation data that conforms to the Laplace distribution. is a random small perturbation data that obeys the Laplace distribution. The probability density function of is as follows: (13) In the formula, μ is the mean parameter of the Laplace distribution, and β is the scale parameter of the Laplace distribution. After generating random small perturbation data, add it to the small sample data constructed in S13 to generate three types of sample sets DataSet-4.

[0066] S24. Combine the small sample data set DataSet-1 constructed in S14, the first-class sample set DataSet-2 constructed in S21 with added small perturbation data that conforms to Gaussian distribution, the second-class sample set DataSet-3 constructed in S22 with added small perturbation data that conforms to uniform distribution, and the third-class sample set DataSet-4 constructed in S23 with added small perturbation data that conforms to Laplace distribution, and merge DataSet-1, DataSet-2, DataSet-3 and DataSet-4 along the vertical splicing method of the sample sets to construct a sample data set for training a multi-feature embedded carbon dioxide geological storage prediction model.

[0067] S31. Establish a multi-feature embedded carbon dioxide geological storage prediction model, which consists of a multi-head attention layer, a residual layer, a fully connected layer and a regularization layer. The setting of each layer is based on the prediction task of the cumulative carbon dioxide geological storage in the entire area of ​​the oil and gas reservoir to be predicted proposed in this application.

[0068] S311. Specifically, the multi-head attention layer allows the prediction model to simultaneously focus on information at different positions of the multi-dimensional input data, which is used to describe the weighted average of the multi-dimensional input data.

[0069] (14) In the formula, Query is the query matrix, Key is the keyword matrix, Value is the value matrix, head i is the ith attention head, that is, there are h attention heads. Attention (*) is a general representation of a single attention head, which contains three parameters: Query, Key, and Value. Softmax (*) is a function. After being processed by this function, the sum of the sequence data is 1. O is the weight matrix, W Query , W Key and WValue are learnable parameters, is the input sequence, d is the dimension of the weight matrix, and T is the length of the sequence. Multihead (*) is a multi-head attention layer.

[0070] Specifically, the input sequence With learnable parameters W Query , W Key and W Value After data mapping, we get Query, Key, and Value. We use softmax (*) to calculate the attention of a single attention head. We use Concat to concatenate h attention heads, and then multiply them with the matrix to calculate the multi-head attention of the multi-head attention.

[0071] S312. Specifically, the residual layer mainly adds a structure directly from input to output between layers, and adds the structure to the output between layers.

[0072] Between layers, including but not limited to multi-head attention layer and regularization layer, regularization layer and fully connected layer, multi-head attention layer and fully connected layer.

[0073] S313. Specifically, the fully connected layer flattens the multi-dimensional data processed by other layers, that is, converts it into a long vector.

[0074] S314. Specifically, the regularization layer can transform each feature of the multi-dimensional input data into sequence data with a mean of 0 and a variance of 1.

[0075] S315. Specifically, an encoder is constructed based on the multi-head attention layer of S311, the residual layer of S312, and the regularization layer of S313, and the encoders are connected using the residual layer of S312.

[0076] S316. Specifically, a decoder is constructed based on the multi-head attention layer of S311, the residual layer of S312, and the regularization layer of S313, and the decoders are connected using the residual layer of S312.

[0077] S317. Specifically, the multi-feature embedded carbon dioxide geological storage prediction model proposed in this application is constructed according to the encoder of S315, the decoder of S316, the softmax activation function and the linear output layer, so that it can meet the prediction task of the cumulative carbon dioxide geological storage volume of the entire oil and gas reservoir to be predicted proposed in this application.

[0078] S32. Define the input data and output data of the multi-feature embedded CO2 geological storage prediction model, convert the geological data and production dynamic data of the oil and gas reservoirs to be predicted into the physical feature layer embedded prediction model, use the physical feature layer and injection parameters as input data, and use the accumulated CO2 geological storage in the whole area as output data.

[0079] The physical feature layer is embedded into the multi-feature embedded CO2 geological storage prediction model, including: S321. Read the sample data set constructed in S24, and extract the geological data and production dynamic data of the oil and gas reservoir to be predicted in the sample data set.

[0080] S322. Use a geological encoder to encode the geological data of the oil and gas reservoir to be predicted, extract the characteristics of the geological data, and obtain a geological data characteristic sequence.

[0081] S323. Use a production encoder to encode the production dynamic data of the oil and gas reservoir to be predicted, extract the characteristics of the production dynamic data, and obtain a production dynamic data characteristic sequence.

[0082] S324. Utilize a trend encoder to encode trend features in the geological data of the oil and gas reservoirs to be predicted and the production dynamic data, and extract trend feature sequences.

[0083] S325. Use the value encoder and the position encoder to encode the sequence features in the geological data of the oil and gas reservoir to be predicted and the production dynamic data, and extract the parameter feature sequence.

[0084] Specifically, the encoding methods corresponding to the geological encoder, production encoder, trend encoder, value encoder and position encoder include but are not limited to: serial number encoding, average encoding, Helmert encoding, binary encoding, one-hot encoding, label encoding and frequency encoding.

[0085] S326. Construct a physical characteristic layer based on geological data characteristics, production dynamic data characteristics, trend characteristics and sequence characteristics, and use the physical characteristic layer and injection parameters as inputs of a multi-feature embedded carbon dioxide geological storage prediction model.

[0086] S33. According to the defined input data and output data, train the multi-feature embedded CO2 geological storage prediction model, and optimize the hyperparameters in the multi-feature embedded CO2 geological storage prediction model through an optimization algorithm. Hyperparameter optimization methods include but are not limited to covariance matrix adaptive evolutionary algorithm, grid search, random search, Bayesian optimization, gradient descent method, etc. Based on S33 training of multi-feature embedded CO2 geological storage prediction model, the preset number of training times is 1000 times, and the optimal model and corresponding hyperparameters of multi-feature embedded CO2 geological storage prediction model with the best prediction effect are retained according to the model evaluation indicators. Model evaluation indicators include but are not limited to CCSR, CCSME and CCSMA.

[0087] The specific calculation methods of CCSR, CCSME and CCSMA are as follows: (36) (37) (38) Specifically, CCSR observes the overall prediction performance of the multi-feature embedded carbon dioxide geological storage prediction model, CCSME observes whether the multi-feature embedded carbon dioxide geological storage prediction model converges, and CCSMA observes whether the multi-feature embedded carbon dioxide geological storage prediction model stops training.

[0088] Call the multi-feature embedded carbon dioxide geological storage prediction model with the best prediction effect retained by S33, input the geological data, production dynamic data and injection parameter combination plan of the oil and gas reservoir to be predicted, and output the cumulative carbon dioxide geological storage of the entire area of ​​the oil and gas reservoir to be predicted. Including: S41. Determine whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are in the same block.

[0089] S42. If yes, then input the geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted, and output the cumulative geological storage capacity of carbon dioxide in the entire area of ​​the oil and gas reservoir to be predicted.

[0090] S43. If not, obtain the geological data, production dynamic data and injection parameter value range of the oil and gas reservoir to be predicted, build the geological model and CO2 storage numerical simulation model of the oil and gas reservoir to be predicted, use the importance sampling method, set 5% of the original sampling times in S12 to perform secondary sampling on the CO2 injection parameters, and perform numerical simulation on each set of injection parameters of the oil and gas reservoir to be predicted to obtain the cumulative CO2 geological storage capacity of the whole area. Add small data disturbances to the CO2 injection parameters, geological parameters and the cumulative CO2 geological storage capacity of the whole area and construct a sample set.

[0091] The geological data and production dynamic data of the reservoir to be optimized are used as the physical feature layer to embed the prediction model, the physical feature layer and injection parameters are used as input data, and the cumulative carbon dioxide geological storage in the whole area is used as output data. The multi-feature embedded carbon dioxide geological storage prediction model after migration training is used to predict the cumulative carbon dioxide geological storage in the whole area of ​​the oil and gas reservoir to be predicted.

[0092] The present embodiment is verified by using a specific example below. Figure 2 , the method of this embodiment includes: Step 1: Based on the geological data and production dynamic data of the oil and gas reservoir to be predicted, a geological model corresponding to the actual reservoir is established, and a numerical simulation model for carbon dioxide storage in the oil and gas reservoir to be predicted is constructed using numerical simulation software. The parameters used are shown in Table 1.

[0093] Table 1 Basic parameters of numerical simulation model for CO2 geological storage

[0094] Step 2: Based on the numerical simulation model of carbon dioxide storage in the oil and gas reservoir to be predicted, a small sample data set is constructed to train the multi-feature embedded carbon dioxide geological storage prediction model. The injection parameter value range is shown in Table 2, and some injection parameter combination schemes formed by importance sampling are shown in Table 3. A small sample data set refers to a sampling of less than 200 times.

[0095] Table 2 Carbon dioxide injection parameter ranges

[0096] Table 3 Examples of some CO2 injection parameter combination schemes

[0097] Step 3: Use the data expansion method to expand the constructed small sample data set to obtain an expanded sample data set for training the multi-feature embedded carbon dioxide geological storage prediction model. Some data of the expanded sample data set are shown in Table 4.

[0098] Table 4 Schematic diagram of some data in the sample data set

[0099] Step 4, establish a multi-feature embedded carbon dioxide geological storage prediction model: embed the geological data and production dynamic data of the oil and gas reservoirs to be predicted into the prediction model as the physical feature layer, use the physical feature layer and the carbon dioxide injection parameter combination plan as input data, and use the cumulative carbon dioxide geological storage in the whole area as output data. Train the multi-feature embedded carbon dioxide geological storage prediction model, use the covariance matrix adaptive evolution strategy to optimize the hyperparameters in the multi-feature embedded carbon dioxide geological storage prediction model, and select the multi-feature embedded carbon dioxide geological storage prediction model with the best prediction effect to predict the cumulative carbon dioxide geological storage in the whole area of ​​the oil and gas reservoirs to be predicted. The 45° intersection diagram of the predicted value and the true value of carbon dioxide geological storage is shown in the figure below. Figure 5 As shown in Figure 2, the perturbation fit goodness is 96.86%.

[0100] Step 5: Call the best multi-feature embedded CO2 geological storage prediction model, input the geological data and production dynamic data of the oil and gas reservoir to be predicted, and output the cumulative CO2 geological storage of the entire area of ​​the oil and gas reservoir to be predicted. The cumulative CO2 geological storage of the entire area is 3098.53×10 4 m 3 .

[0101] like Figure 6 In an exemplary embodiment, a device for predicting the amount of geological storage of carbon dioxide is provided, comprising: Importance sampling module 202: used for constructing different injection parameter combinations by using importance sampling based on the value range of each injection parameter determined by the parameter loading module.

[0102] Numerical simulation module 303: used to carry out numerical simulation of oil reservoirs in the oil and gas reservoirs to be predicted, and use the parameter combination constructed by module 202 to carry out numerical simulation to obtain the cumulative geological storage capacity of carbon dioxide in the whole area.

[0103] Data expansion module 404: used to expand the data of the cumulative carbon dioxide geological storage capacity in the whole area obtained by the numerical simulation module according to the data expansion process, calculate the data under three different small disturbance conditions, and merge them with the results obtained by module 303 into a sample set.

[0104] Physical feature layer embedding module 505: used to extract geological data features, production dynamic data features, trend features and sequence features of the oil and gas reservoir geological data and production dynamic data to be predicted in the sample set, and construct a physical feature layer with the above features as the input of the multi-feature embedded carbon dioxide geological storage prediction model.

[0105] Model training module 606: used to train the multi-feature embedded carbon dioxide geological storage prediction model, the input data is the physical feature layer and injection parameters embedded in the geological data and production dynamic data, and the output data is the cumulative carbon dioxide geological storage in the whole area. At the same time, the covariance matrix adaptive evolution strategy is used to optimize the hyperparameters of the multi-feature embedded carbon dioxide geological storage prediction model, and the multi-feature embedded carbon dioxide geological storage prediction model with the best prediction effect is retained.

[0106] Prediction module 707: used to predict the cumulative carbon dioxide geological storage capacity of the entire oil and gas reservoir to be predicted, call the multi-feature embedded carbon dioxide geological storage capacity prediction model with the best prediction effect, input geological data, production dynamic data and injection parameter combination plan, and obtain the cumulative carbon dioxide geological storage capacity prediction data of the entire oil and gas reservoir to be predicted.

[0107] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for predicting the geological storage of carbon dioxide is implemented.

[0108] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0109] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0110] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0111] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0112] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0113] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for predicting the amount of geological storage of carbon dioxide, characterized in that: include: Determine whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are in the same block; If so, the geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted are input into the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model to obtain the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted; the optimal reference oil and gas reservoir carbon dioxide geological storage prediction model is input with the geological data, production dynamic data and injection parameter combination scheme of the reference oil and gas reservoir, and the cumulative carbon dioxide geological storage capacity of the entire area of ​​the reference oil and gas reservoir is output, and the hyperparameters in the multi-feature embedded carbon dioxide geological storage prediction model are optimized by using an optimization algorithm, and the multi-feature embedded carbon dioxide geological storage prediction model is trained multiple times to obtain the model; If not, then obtain the geological data, production performance data and injection parameter value range of the oil and gas reservoir to be predicted; Based on the geological data and production performance data of the oil and gas reservoirs to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoirs to be predicted is constructed; Sampling once within the range of CO2 injection parameter values ​​for the oil and gas reservoir to be predicted, and obtaining multiple groups of CO2 injection parameter combination schemes for the oil and gas reservoir to be predicted; the CO2 injection parameters include but are not limited to: CO2 injection amount, CO2 injection rate, CO2 injection time, CO2 injection pressure, CO2 injection slug size, water injection / gas injection ratio, and well soaking time; Input multiple groups of injection parameter combinations of the oil and gas reservoirs to be predicted into the numerical simulation model of carbon dioxide storage in the oil and gas reservoirs to be predicted, and obtain the cumulative carbon dioxide geological storage volume of the whole area corresponding to different injection parameter combinations of the oil and gas reservoirs to be predicted; Taking the geological data and production dynamic data of the oil and gas reservoir to be predicted and the injection parameter combination scheme of the oil and gas reservoir to be predicted as input, and taking the cumulative carbon dioxide geological storage volume of the whole area corresponding to the injection parameter combination scheme of the oil and gas reservoir to be predicted as output, the optimal control oil and gas reservoir carbon dioxide geological storage volume prediction model is transferred and trained to obtain the carbon dioxide geological storage volume prediction model of the oil and gas reservoir to be predicted; The geological data, production dynamic data and injection parameter combination scheme of the oil and gas reservoir to be predicted are input into the prediction model of carbon dioxide geological storage capacity of the oil and gas reservoir to be predicted, and the cumulative carbon dioxide geological storage capacity of the entire area of ​​the oil and gas reservoir to be predicted is obtained.

2. The method for predicting the amount of geological storage of carbon dioxide according to claim 1, characterized in that The multi-feature embedded carbon dioxide geological storage capacity prediction model includes a physical feature layer and a multi-feature embedded prediction layer connected in sequence.

3. The method for predicting the amount of geological storage of carbon dioxide according to claim 2, characterized in that: The physical feature layer includes: a geological data feature sequence extraction branch, a production dynamic data feature sequence extraction branch, a parameter feature sequence extraction branch, a trend feature sequence extraction branch and a splicing layer; the output end of the production dynamic data feature sequence extraction branch, the output end of the geological data feature sequence extraction branch, the output end of the parameter feature sequence extraction branch and the output end of the trend feature sequence extraction branch are all connected to the input end of the splicing layer; The geological data feature sequence extraction branch includes a geological encoder; the input of the geological data feature sequence extraction branch is the geological data of the oil and gas reservoir to be predicted; The production dynamic data feature sequence extraction branch includes a production encoder; the input of the production dynamic data feature sequence extraction branch is the production dynamic data of the oil and gas reservoir to be predicted; The parameter feature sequence extraction branch includes a value encoder and a position encoder connected in parallel; the input of the parameter feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted; The trend feature sequence extraction branch includes a trend encoder; the input of the trend feature sequence extraction branch is the geological data and production dynamic data of the oil and gas reservoir to be predicted; The production dynamic data feature sequence extraction branch and the geological data feature sequence extraction branch respectively extract the overall characteristics of all parameters in the production dynamic data and geological data, the parameter feature sequence extraction branch extracts the individual characteristics of any parameter in the production dynamic data and geological data, and the trend feature sequence extraction branch extracts the trend characteristics of any parameter in the production dynamic data and geological data; the splicing layer is spliced ​​by the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch in a horizontal splicing manner, and the output end of the splicing layer is connected to the multi-feature embedding prediction layer; the horizontal splicing method aligns the heads of the production dynamic data feature sequence extraction branch, the geological data feature sequence extraction branch, the parameter feature sequence extraction branch and the trend feature sequence extraction branch, and splices the aligned extraction branches in sequence from left to right to obtain a splicing layer.

4. The method for predicting the amount of geological storage of carbon dioxide according to claim 2, characterized in that: The multi-feature embedding prediction layer includes: The branch at time t, the branch at time t+1, the first fully connected layer and the activation function are connected in sequence; The input end of the branch at time t and the input end of the branch at time t+1 are both connected to the output end of the physical feature layer; The input end of the branch at time t and the input end of the branch at time t+1 are also used to input a combination scheme of injection parameters of the oil and gas reservoir to be predicted.

5. The method for predicting the amount of geological storage of carbon dioxide according to claim 4, characterized in that: The branch at time t includes: a first multi-head attention layer, a first residual layer, a first regularization layer, a second fully connected layer, a second residual layer, and a second regularization layer connected in sequence; The branch at time t+1 includes: a second multi-head attention layer, a third residual layer, a third regularization layer, a third fully connected layer, a fourth residual layer and a fourth regularization layer connected in sequence; The input end of the first multi-head attention layer and the input end of the first regularization layer are used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t in the physical feature sequence; the physical feature sequence is obtained by splicing the production dynamic data feature sequence, the geological data feature sequence, the parameter feature sequence and the trend feature sequence; The output end of the first regularization layer is also connected to the input end of the second regularization layer; The output end of the second regularization layer is also connected to the input end of the third regularization layer; The input end of the second multi-head attention layer is used to input the injection parameter combination scheme of the oil and gas reservoir to be predicted and the data at time t+1 in the physical feature sequence; The output end of the third regularization layer is also connected to the input end of the fourth regularization layer; The output end of the fourth regularization layer is also connected to the input end of the first fully connected layer.

6. The method for predicting the amount of geological storage of carbon dioxide according to claim 1, characterized in that: Before judging whether the oil and gas reservoir to be predicted and the reference oil and gas reservoir are in the same block, the following steps are also included: Obtain geological data, production performance data and injection parameter value ranges of reference oil and gas reservoirs; Based on the geological data and production performance data of the reference oil and gas reservoirs, a numerical simulation model for CO2 storage in the reference oil and gas reservoirs is constructed; Secondary sampling is performed within the range of CO2 injection parameter values ​​to obtain multiple groups of CO2 injection parameter combination schemes for control oil and gas reservoirs; Inputting a plurality of control oil and gas reservoir injection parameter combination schemes into the control oil and gas reservoir carbon dioxide storage numerical simulation model respectively, and obtaining the cumulative carbon dioxide geological storage volume of the whole area corresponding to different control oil and gas reservoir injection parameter combination schemes; A small sample data set was constructed by taking the injection parameter combination scheme of the control oil and gas reservoir as input and the cumulative carbon dioxide geological storage volume in the whole area corresponding to different injection parameter combination schemes of the control oil and gas reservoir as output. The small sample data set is expanded using the small perturbation data to obtain an expanded sample data set; the expansion processing includes but is not limited to the small perturbation data that conforms to the Gaussian distribution, uniform distribution, and Laplace distribution; the small sample data set and the expanded sample data set are spliced ​​in a vertical splicing manner to obtain a training data set; the vertical splicing manner aligns the left sides of the expanded sample data set and the small sample data set, and the aligned data sets are stacked in order from top to bottom to obtain the training data set; Taking the geological data and production dynamic data of the reference oil and gas reservoirs and the injection parameter combination scheme of the reference oil and gas reservoirs in the training data set as input, and taking the cumulative carbon dioxide geological storage volume of the whole area in the training data set as output, the multi-feature embedded carbon dioxide geological storage volume prediction model is trained multiple times and the hyperparameters in the multi-feature embedded carbon dioxide geological storage volume prediction model are optimized by using an optimization algorithm, and the evaluation index of the reference oil and gas reservoir carbon dioxide geological storage volume prediction model obtained by each training is determined, and the optimal reference oil and gas reservoir carbon dioxide geological storage volume prediction model is obtained based on the evaluation index; the evaluation index includes but is not limited to the perturbation fit goodness, the storage process error under small perturbation and the average absolute error of geological storage volume under small perturbation; The perturbation goodness of fit is: ; The error of the sealing process under small disturbance is: ； The average absolute error of geological storage under small disturbance is: ； Where CCSR is the perturbation fit goodness, CCSME is the storage process error under small disturbance, CCSMA is the mean absolute error of geological storage under small disturbance, N is the number of sample sets in the training data set, i is the sample set number, j is the sample sequence number under a certain sample set, and N is the number of samples in the training data set. j is the total number of samples in the i-th sample set, is the prediction result of carbon dioxide geological storage at the jth sample in the i-th sample set at time t, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample of the i-th sample set at time t, is the arithmetic mean of the numerical simulation results of the geological storage of carbon dioxide of the jth sample in the ith sample set within the carbon dioxide injection time, num is the total number of samples, Total is the carbon dioxide injection time, t is the single time step of the carbon dioxide injection time, is the distribution function corresponding to the small perturbation data of the i-th sample set, is the numerical simulation result of the geological storage of carbon dioxide at the jth sample Total time of the i-th sample set, It is the prediction result of the geological storage amount of carbon dioxide at the jth sample Total time in the i-th sample set.

7. The method for predicting the amount of geological storage of carbon dioxide according to claim 6, characterized in that: Based on the geological data and production performance data of the oil and gas reservoirs to be predicted, a numerical simulation model for carbon dioxide storage in the oil and gas reservoirs to be predicted is constructed, including: Based on the geological data of the oil and gas reservoir to be predicted, a geological model of the oil and gas reservoir to be predicted is constructed using geological modeling software; Import the geological model of the oil and gas reservoir to be predicted into the oil reservoir numerical simulation software, use the production performance data to perform historical matching on the production performance of the oil and gas reservoir to be predicted, and obtain the development numerical simulation model of the oil and gas reservoir to be predicted; Based on the numerical simulation model of the development of the oil and gas reservoir to be predicted, a numerical simulation of carbon dioxide storage in the oil and gas reservoir to be predicted is performed to obtain a numerical simulation model of carbon dioxide storage in the oil and gas reservoir to be predicted.

8. The method for predicting the amount of geological storage of carbon dioxide according to claim 7, characterized in that: The number of samplings for primary sampling is 5~10% of the number of samplings for secondary sampling.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for predicting the geological storage amount of carbon dioxide according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the geological storage amount of carbon dioxide described in any one of claims 1 to 8 is implemented.

Images