Methods, devices, equipment and media for predicting oil and gas field flow data

Abstract

This invention provides a method, apparatus, equipment, and medium for predicting oil and gas field flow data. The oil and gas field flow correction method includes the following steps: S1, acquiring and preprocessing wellhead data of oil and gas wells; S2, identifying influencing factors of wellhead flow data by analyzing the correlation between the wellhead data and wellhead flow data; S3, identifying influencing factors of wellhead flow data by analyzing the correlation between the wellhead data and wellhead flow data; S4, establishing a hybrid virtual metering model of oil and gas wells based on the aforementioned mechanism model, and labeling the characteristic variables of the hybrid virtual metering model according to the influencing factors; S5, predicting wellhead flow data based on the hybrid virtual metering model. This invention, by using existing data for mining and analysis, can achieve accurate measurement of wellhead flow, replacing the current complex back-matching mechanism, and can enhance the core competitiveness in the field of efficient oil and gas field development.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field development, and more specifically, to a method for predicting oil and gas field flow data, an apparatus for predicting oil and gas field flow data, and a device and computer-readable storage medium for implementing the method for predicting oil and gas field flow data. Background Technology

[0002] Single-well metering is a crucial aspect of oil and gas field development and management, directly impacting vital development and production activities such as reservoir analysis and production planning. However, due to the unique characteristics of unconventional oil and gas fields, traditional single-well metering instruments (such as differential pressure flow meters, turbine flow meters, and ultrasonic flow meters) suffer from issues like instrument deviation, susceptibility to damage, and maintenance difficulties. This results in the inability to obtain accurate metering results, negatively affecting daily pressure optimization, production scheduling, and the operation of TEG dehydration units (gathering stations). For example, high-sulfur gas fields, with hydrogen sulfide (H2S) concentrations exceeding 0.1%, are a special type of unconventional gas reservoir characterized by high temperature, high pressure, high corrosiveness, and high toxicity, posing significant challenges and risks to the development, construction, and operation of these gas fields.

[0003] Virtual metering technology primarily expands the collected data through sensors and utilizes advanced data analysis and processing algorithms to achieve accurate estimation of physical quantities at the software level, rather than through traditional direct hardware measurement. Early virtual metering was mainly based on mechanistic models, using dynamic simulation systems of these models to estimate the oil, gas, and water flow rates of a well or wellhead system. Hybrid virtual metering, utilizing mechanistic models, data models, or a combination of both, is now widely used in oil and gas field development as a supplement or alternative to physical metering devices. Furthermore, with the rapid development of big data, the Internet of Things (IoT), and sensor technology, big data technology has made it possible to process massive amounts of sensor-collected data. Through big data analysis, valuable information can be extracted from huge datasets, and correlations and patterns between data can be discovered, thereby optimizing and improving virtual metering models. In addition, the application of machine learning and artificial intelligence can further enhance the intelligence level of virtual metering systems, enabling self-learning and adaptation, and improving the accuracy and reliability of measurements. With the rapid development of deep learning technology, many deep learning models have been applied to the study of time series data. Recurrent Neural Networks (RNNs) introduce time series data into neural networks, improving the model's adaptability to time series data. However, based on the RNN structure, some problems still exist, the most significant of which is gradient vanishing or gradient exploding, meaning that the model's memory value becomes smaller and smaller as the time step increases. Therefore, there is an urgent need for a new hybrid virtual econometric method that can meet the needs of modern development, overcome the above problems, and be applicable to the development of large and unconventional oil and gas fields. Summary of the Invention

[0004] The purpose of this invention is to address at least one of the aforementioned shortcomings of the existing technology. For example, one objective of this invention is to propose a virtual metering technology method that combines a transient mechanism model and a data model for high-sulfur multiphase flow based on machine learning algorithms. This method can quickly analyze relevant real-time data at the wellhead, explore the correlation between various characteristic parameters and the wellhead flow trend, establish a wellhead flow prediction model, and realize real-time virtual metering for a single well.

[0005] To achieve the above objectives, the present invention provides a method for predicting oil and gas field flow data.

[0006] The method for predicting oil and gas field flow data includes the following steps:

[0007] S1. Obtain wellhead data of oil and gas wells and perform preprocessing.

[0008] S2. The influencing factors of wellhead flow rate data are identified by analyzing the correlation between the wellhead data and the wellhead flow rate data.

[0009] S3. Conduct sensitivity analysis on the production capacity mechanism model of oil and gas wells to identify the relevant influencing factors of wellhead flow rate, and revise the mechanism model of oil and gas fields based on the relevant influencing factors.

[0010] S4. Establish a hybrid virtual metering model for oil and gas wells based on the aforementioned mechanism model, and label the characteristic variables of the hybrid virtual metering model according to the aforementioned influencing factors.

[0011] S5. Predict wellhead flow data based on the hybrid virtual metering model.

[0012] In an exemplary embodiment of the method for predicting oil and gas field flow data of the present invention, the oil and gas field may be a high-sulfur gas field, and the oil and gas wells include one or more of vertical wells and horizontal wells.

[0013] In an exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the production capacity mechanism model may include one or more of the following: conventional oil and gas well production capacity calculation formula, production capacity calculation formula with single-factor influence, and production capacity calculation formula with dual-factor influence.

[0014] In an exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the correlation can be confirmed by one or more of the following methods: multivariate correlation analysis, multivariate regression analysis, principal component analysis, and grey relational analysis.

[0015] In an exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the neural network may be a recurrent neural network.

[0016] In an exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the gradient descent algorithm can be used to optimize the hybrid virtual metering model.

[0017] In an exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the measurement error of the hybrid virtual metering model can be less than 10%.

[0018] In another aspect, the present invention provides a device for predicting oil and gas field flow data. The device includes a wellhead data acquisition and processing module, an influencing factor confirmation module, a mechanism model correction module, a hybrid virtual metering model establishment module, and a wellhead flow data prediction module.

[0019] The wellhead data acquisition and processing module is configured to acquire wellhead data of oil and gas wells and perform preprocessing.

[0020] The influencing factor confirmation module is connected to the wellhead data acquisition and processing module and is configured to confirm the influencing factors of the wellhead flow rate data by analyzing the correlation between the wellhead data and the wellhead flow rate data.

[0021] The mechanism model correction and key influencing factor confirmation module is connected to the wellhead data acquisition and processing module and is configured to perform sensitivity analysis on the production capacity mechanism model of oil and gas wells to confirm the relevant influencing factors of wellhead flow rate, and correct the mechanism model of oil and gas fields based on the relevant influencing factors.

[0022] The hybrid virtual metering model establishment module is connected to the influencing factor confirmation module and the mechanism model correction module, respectively, and is configured to establish a hybrid virtual metering model of oil and gas wells based on the mechanism model, and to label the feature variables of the hybrid virtual metering model according to the influencing factors.

[0023] The wellhead flow rate prediction module is connected to the hybrid virtual metering model establishment module and is configured to predict wellhead flow rate data based on the hybrid virtual metering model.

[0024] In another aspect, the present invention provides a computer device, the computer device comprising:

[0025] Processor; memory storing a computer program that, when executed by the processor, implements the method for predicting oil and gas field flow data as described above.

[0026] In another aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for predicting oil and gas field flow data as described above.

[0027] Compared with the prior art, the beneficial effects of the present invention include at least one of the following:

[0028] (1) This invention enables precise measurement of wellhead flow rate by using existing data for mining and analysis. Through big data analysis and machine learning algorithms, this method achieves accurate flow rate estimation at the software level, reducing equipment wear and tear, lowering maintenance costs, and improving production stability and safety. This technology not only enables precise calculation of daily production for each well, replacing the current complex rework mechanism, but also enhances the core competitiveness in the field of efficient oil and gas field development.

[0029] (2) This invention can improve the accuracy and efficiency of single-well metering, reduce metering costs and risks, and provide valuable information and suggestions for the development and management of high-sulfur gas fields.

[0030] (3) The present invention uses a long short-term neural network to adjust the gradient through a forget gate and an update gate. By changing the backpropagation multiplication operation to an addition operation, the problem of gradient vanishing or exploding is solved, so that the recurrent neural network can more effectively process and analyze time-series information with long step lengths. Attached Figure Description

[0031] The above and other objects and / or features of the present invention will become clearer from the following description taken in conjunction with the accompanying drawings, in which:

[0032] Figure 1 A schematic diagram of the prediction process of an embodiment of the oil and gas field flow data prediction method of the present invention is shown.

[0033] Figure 2 A schematic diagram of the correlation analysis process of an embodiment of the oil and gas field flow data prediction method of the present invention is shown.

[0034] Figure 3 A schematic diagram illustrating the mechanism model construction process of an embodiment of the oil and gas field flow data prediction method of the present invention is shown.

[0035] Figure 4 A schematic diagram of the recurrent neural network construction of an embodiment of the oil and gas field flow data prediction method of the present invention is shown.

[0036] Figure 5 A schematic diagram showing a comparison between the corrected formula and the original formula of an embodiment of the oil and gas field flow data prediction method of the present invention is illustrated.

[0037] Figure 6 A schematic diagram showing a comparison of model data prediction values, formula calculation values, and EC back-matching values ​​in an embodiment of the oil and gas field flow data prediction method of the present invention is illustrated. Detailed Implementation

[0038] In the following sections, the method, apparatus, equipment, and medium for predicting oil and gas field flow data of the present invention will be described in detail with reference to exemplary embodiments.

[0039] It should be noted that the terms “S1”, “S2”, “S3”, etc. used in this invention are only for the convenience of description and distinction, and should not be construed as indicating or implying relative importance or used to describe a specific order or sequence.

[0040] Virtual metering, utilizing mechanistic models, data models, or a hybrid of both, serves as a supplement or alternative to physical metering devices and is widely applied in oil and gas field development. Internationally, many advanced oil and gas fields, such as deepwater oil fields in the North Sea and the Gulf of Mexico, have begun to utilize virtual metering technology. Domestically, virtual metering technology is also gradually being promoted, especially in the development of large oil and gas fields and unconventional gas fields. Therefore, the inventors propose a new method for predicting oil and gas field flow data. This method uses machine learning and deep learning modeling to virtually meter oil and gas flow, further combining the advantages of mechanistic models to optimize data models, thus combining the strengths and weaknesses of both mechanistic and data models to form a hybrid virtual metering technology.

[0041] To achieve the above objectives, the present invention provides a method for predicting oil and gas field flow data.

[0042] In a first exemplary embodiment of the oil and gas field flow data prediction method of the present invention, the oil and gas field flow data prediction method includes the following steps:

[0043] S1. Obtain wellhead data of oil and gas wells and perform preprocessing.

[0044] Optionally, the oil and gas field may be a high-sulfur gas field, and the oil and gas well may include one or more of vertical and horizontal wells.

[0045] S2. Identify the influencing factors of wellhead flow rate data by analyzing the correlation between wellhead data and wellhead flow rate data.

[0046] Optionally, one or more of the following methods can be used to confirm the correlation: multivariate correlation analysis, multivariate regression analysis, principal component analysis, and grey relational analysis.

[0047] S3. Conduct sensitivity analysis on the production mechanism model of oil and gas wells to identify relevant influencing factors on wellhead flow rate, and revise the mechanism model of the oil and gas field based on these factors. Optionally, the production mechanism model may include one or more of the following: conventional oil and gas well production calculation formula, production calculation formula based on a single factor, and production calculation formula based on the influence of two factors.

[0048] S4. Establish a hybrid virtual metering model for oil and gas wells based on the mechanism model, and label the characteristic variables of the hybrid virtual metering model according to the influencing factors.

[0049] Optionally, the neural network can be a recurrent neural network.

[0050] Optionally, the gradient descent algorithm can be used to optimize the hybrid virtual econometric model.

[0051] S5. Predict wellhead flow data based on a hybrid virtual metering model.

[0052] Optionally, the measurement error of the hybrid virtual measurement model is less than 10%.

[0053] In another aspect, the present invention provides a second exemplary embodiment of an oil and gas field flow data prediction device. The oil and gas field flow data prediction device includes a wellhead data acquisition and processing module, an influencing factor confirmation module, a mechanism model correction module, a hybrid virtual metering model establishment module, and a wellhead flow data prediction module.

[0054] The system comprises the following modules: a wellhead data acquisition and processing module, configured to acquire and preprocess wellhead data from oil and gas wells; an influencing factor identification module, connected to the wellhead data acquisition and processing module, configured to identify influencing factors of wellhead flow rate by analyzing the correlation between wellhead data and wellhead flow rate data; a mechanism model correction and key influencing factor identification module, also connected to the wellhead data acquisition and processing module, configured to perform sensitivity analysis on the oil and gas well productivity mechanism model to identify relevant influencing factors of wellhead flow rate and correct the oil and gas field mechanism model based on these influencing factors; a hybrid virtual metering model establishment module, connected to both the influencing factor identification and mechanism model correction modules, configured to establish a neural network-based hybrid virtual metering model for oil and gas wells based on the mechanism model and label the characteristic variables of the hybrid virtual metering model according to the influencing factors; and a wellhead flow rate data prediction module, connected to the hybrid virtual metering model establishment module, configured to predict wellhead flow rate data based on the hybrid virtual metering model.

[0055] In another aspect, the present invention provides a third exemplary embodiment of a computer device. The computer device includes a processor and a memory. The memory stores a computer program. The computer program is executed by the processor, causing the processor to perform a computer program for predicting oil and gas field flow data according to the present invention.

[0056] In another aspect, the present invention provides a fourth exemplary embodiment of a computer-readable storage medium storing a computer program. The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform a method for predicting oil and gas field flow data according to the present invention. The computer-readable recording medium is any data storage device capable of storing data readable by a computer system. Examples of computer-readable recording media include: read-only memory, random access memory, read-only optical disc, magnetic tape, floppy disk, optical data storage device, and carrier waves (such as data transmission via the Internet through wired or wireless transmission paths).

[0057] To better understand the exemplary embodiments of the present invention described above, further descriptions are provided below in conjunction with specific embodiments and accompanying drawings, but the examples given are not intended to limit the present invention.

[0058] Example 1

[0059] In this embodiment, the technical process of the oil and gas field flow data prediction method is as follows: Figure 1 As shown, this can be achieved through the following steps:

[0060] S1. Process the wellhead data, clean and format the data, and handle any missing or outlier values.

[0061] The data may include laboratory data, field data, and computational data based on mechanistic models.

[0062] S2. Conduct a correlation analysis on the wellhead flow rate and its related influencing factors, identify the strongly correlated influencing factors, and mark the characteristic variables of the hybrid virtual metering model in step S3 based on the strongly correlated influencing factors.

[0063] Wellhead data includes time-series data of key parameters such as wellhead flow rate, downhole pressure, downhole temperature, wellhead pressure, and wellhead temperature. Different methods can be used to analyze and mine this data, exploring the correlation patterns between wellhead flow rate and other parameters. Comparative studies of different methods can yield unified results, confirming the relative strength of each influencing factor's impact on wellhead flow rate.

[0064] like Figure 2As shown, correlation analysis can be performed using methods such as multivariate correlation analysis, multivariate regression analysis, principal component analysis, and grey relational analysis. Research on the correlation of factors influencing wellhead flow rate is crucial for establishing subsequent flow rate prediction models. Therefore, in-depth research on the correlation between factors influencing wellhead flow rate using data-driven methods can improve the understanding of oil and gas well production processes, provide scientific basis and data support for the oil and gas extraction industry, offer more effective decision-making references for production and operation personnel, and promote the intelligent and digital development of oil and gas well production processes. This research is of great significance for optimizing oil and gas well production processes, increasing production, and reducing costs.

[0065] like Figure 3 As shown, this paper reviews the current status of gas well productivity mechanism models, including those for vertical and horizontal wells. It analyzes conventional oil and gas well productivity calculation formulas, formulas influenced by single factors, and formulas influenced by dual factors, and performs sensitivity analysis on these formulas. Relevant factors for wellhead flow rate are extracted, and a modified formula (i.e., the mechanism model for high-sulfur gas fields) is developed. A mechanism model is a mathematical model established based on the principles of fluid motion and the structural characteristics of a flowmeter, used to describe the working mechanism and measurement characteristics of the flowmeter. For example, a mechanism model for flow measurement can help us understand the measurement principle of a flowmeter, analyze its measurement errors, optimize its design parameters, and improve its measurement accuracy and reliability.

[0066] S3. Construct data and mechanism models for high-sulfur gas fields, and use big data analysis to form a hybrid virtual metering model as a wellhead flow model. The model can also provide early warnings for real-time wellhead flow parameters, enabling the estimation of uncertainties in actual flow to correct metering.

[0067] like Figure 4 As shown, a recurrent neural network is constructed. This framework consists of an input layer, hidden layers, an output layer, and a model training layer. The functions of each module are as follows: 1) Input layer: Performs data preprocessing operations such as segmentation and standardization on the original time variable sequence set to meet the model input requirements. 2) Hidden layer: Uses the neural unit structure shown in the figure to update and optimize the parameters. 3) Output layer: Outputs the model's prediction results through destandardization to verify the prediction error. 4) Model training: Uses the gradient descent algorithm (here, we assume the Adam optimization algorithm is used) to update the weights of the prediction model. The main methods and steps are as follows:

[0068] The entire dataset is processed to remove irrelevant data. Data preprocessing is performed on these datasets to obtain normalized or standardized standard data.

[0069] The hidden layers of a recurrent neural network wellhead flow prediction model are constructed. Training is performed using the input training set data. The Adam function in the model optimizer is used to automatically process the model parameters. The Adam method allows the model to adapt to the learning rate, and parameter updates are not affected by gradient scaling transformations. Therefore, relying on the Adam optimizer, the hyperparameters in the library are automatically adjusted through each effective training iteration. The main hyperparameters of the model are shown in Table 1 below, and the optimal hyperparameters are shown in Table 2 below.

[0070] Table 1. Main hyperparameters of the model

[0071]

[0072] Keras Tuner is a toolkit specifically designed for optimizing hyperparameters in TensorFlow and Keras models. It can be used to automatically search for optimal combinations of hyperparameters to improve model performance. The search space for hyperparameters can be defined, including learning rate, number of units in a layer, Dropout, etc., and optimal parameters can be determined using a random search approach.

[0073] Table 2 Optimal Hyperparameters of the Model

[0074] hyperparameters value LSTM layers 2 Number of neurons in the first layer 256 First-level dropout 0.0 Number of neurons in the second layer 224 Second-level dropout 0.3 Learning rate 0.008 Batch size 32 Number of iterations 50 Optimizer Adam

[0075] Model prediction: The trained recurrent neural network wellhead flow model will be used to make predictions on the test set to obtain the predicted wellhead flow rate. The predicted wellhead flow rate will be compared with the actual wellhead flow rate, and RMSE and R2 will be used to verify the accuracy of the data. Finally, the prediction effect of the model will be displayed intuitively through visualization.

[0076] Example 2

[0077] In this embodiment, the oil and gas field flow rate prediction method as described in Example 1 is applied to predict the wellhead flow rate of a gas well, and the flow rate predicted by the existing formula is compared and analyzed.

[0078] 1) Using mechanistic models:

[0079] Original formula: Y = aX^2 + bX + C.

[0080] Where Y represents tubing pressure in MPa; X represents flow rate in cubic meters per hour; and C represents shut-in negative pressure, such as the most recent shut-in negative pressure of well LJ24 being 20.23 MPa.

[0081] Because the total flow rate calculated in the original formula differs significantly from the daily return value, the improvement approach in this embodiment is to replace C with a dynamic downhole pressure and correct C by multiplying it by a coefficient. For example, using well test data from May 12, 2023 to August 30, 2023 for formula fitting, the result is as follows: Y = -0.1149X^2 + 0.1459X + 0.7525Z. Furthermore, predicting daily production data from September 1, 2023 to December 31, 2023 yields the following result:

[0082] Fitting parameters: a = -0.1149, b = 0.1459, c = 0.7525.

[0083] R-squared value: R 2 =0.9878.

[0084] Comparison of the corrected formula and the original formula's back-matching value, for example Figure 5 As shown, by replacing the fixed shut-in negative pressure with CZ and correcting the original formula, the calculated flow rate becomes more accurate and closer to the reflow value.

[0085] 2) Using the wellhead flow rate model:

[0086] Using a recurrent neural network model, the production data of this gas well from January 10th to 16th, 2023, was used as input to predict the wellhead flow rate. This prediction was then compared with the flow rate calculated using a formula and the company's EC return value (which can be simply understood as the accurate wellhead flow rate). The results are as follows: Figure 6 As shown in Table 3 below.

[0087] Table 3. Error Comparison Table of Predicted Values, Actual Values, and Feedback Values ​​for a Certain Gas Well

[0088]

[0089] Based on the above comparison Figure 6 As shown in Table 3, the method can accurately predict wellhead flow rate. Compared with existing calculation formulas, it can significantly improve the accuracy. The average error with the back-matching value is only about 2%, which is less than 3%.

[0090] Although the present invention has been described above in conjunction with exemplary embodiments and accompanying drawings, those skilled in the art should understand that various modifications can be made to the above embodiments without departing from the spirit and scope of the claims.

Claims

1. A method for predicting oil and gas field flow data, characterized in that, The oil and gas field flow correction method includes the following steps: S1. Obtain wellhead data of oil and gas wells and perform preprocessing; S2. Identify the influencing factors of wellhead flow rate data by analyzing the correlation between the wellhead data and the wellhead flow rate data; S3. Conduct sensitivity analysis on the production capacity mechanism model of oil and gas wells to identify the relevant influencing factors of wellhead flow rate, and revise the mechanism model of oil and gas fields based on the relevant influencing factors; S4. Establish a hybrid virtual metering model for oil and gas wells based on the aforementioned mechanism model, and label the characteristic variables of the hybrid virtual metering model according to the aforementioned influencing factors; S5. Predict wellhead flow data based on the hybrid virtual metering model.

2. The method for predicting oil and gas field flow data according to claim 1, characterized in that, The oil and gas field is a high-sulfur gas field, and the oil and gas wells include one or more of vertical wells and horizontal wells.

3. The method for predicting oil and gas field flow data according to claim 1, characterized in that, The production capacity mechanism model includes one or more of the following: conventional oil and gas well production capacity calculation formula, production capacity calculation formula based on a single factor, and production capacity calculation formula based on the influence of two factors.

4. The method for predicting oil and gas field flow data according to claim 1, characterized in that, The correlation was confirmed using one or more of the following methods: multivariate correlation analysis, multivariate regression analysis, principal component analysis, and grey relational analysis.

5. The method for predicting oil and gas field flow data according to claim 1, characterized in that, The neural network is a recurrent neural network.

6. The method for predicting oil and gas field flow data according to claim 5, characterized in that, The gradient descent algorithm is used to optimize the hybrid virtual econometric model.

7. The method for predicting oil and gas field flow data according to claim 1, characterized in that, The measurement error of the hybrid virtual metrology model is less than 10%.

8. A device for predicting oil and gas field flow data, characterized in that, The oil and gas field flow rate prediction device includes a wellhead data acquisition and processing module, an influencing factor confirmation module, a mechanism model correction module, a hybrid virtual metering model establishment module, and a wellhead flow rate prediction module. The wellhead data acquisition and processing module is configured to acquire wellhead data of oil and gas wells and perform preprocessing. The influencing factor confirmation module is connected to the wellhead data acquisition and processing module and is configured to confirm the influencing factors of the wellhead flow rate data by analyzing the correlation between the wellhead data and the wellhead flow rate data. The mechanism model correction and key influencing factor confirmation module is connected to the wellhead data acquisition and processing module and is configured to perform sensitivity analysis on the production capacity mechanism model of oil and gas wells to confirm the relevant influencing factors of wellhead flow rate, and correct the mechanism model of oil and gas fields based on the relevant influencing factors. The hybrid virtual metering model establishment module is connected to the influencing factor confirmation module and the mechanism model correction module, respectively, and is configured to establish a hybrid virtual metering model of oil and gas wells based on the mechanism model, and to label the feature variables of the hybrid virtual metering model according to the influencing factors. The wellhead flow rate prediction module is connected to the hybrid virtual metering model establishment module and is configured to predict wellhead flow rate data based on the hybrid virtual metering model.

9. A computer device, characterized in that, The computer device includes: At least one processor; and A memory storing program instructions configured to be executed by the at least one processor, the program instructions including instructions for performing the method for predicting oil and gas field flow data according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method for predicting oil and gas field flow data as described in any one of claims 1 to 7.

Images