Supercritical carbon dioxide hydrate phase diagram generation method, device and equipment

Abstract

The invention provides a supercritical carbon dioxide hydrate phase diagram generation method, device and equipment wherein the supercritical carbon dioxide hydrate phase diagram generation method comprises the following steps: establishing an SC-CO2 hydrate phase state data test system based on a multi-source monitoring technology, a physical data set and an image data set are obtained through the SC-CO2 hydrate phase state data testing system; dividing the physical data set and the image data set into a training data set and a test data set; constructing an SC-CO2 hydrate phase state data prediction generation model, and performing training optimization on the SC-CO2 hydrate phase state data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; and obtaining to-be-analyzed data under different environmental conditions, and predicting the to-be-analyzed data through the optimal generalization model to obtain a prediction result. According to the invention, the visual experiment of the phase evolution of the SC-CO2 hydrate is realized, and the generation precision and efficiency of the phase diagram of the SC-CO2 hydrate can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical fields of carbon dioxide geological sequestration and machine learning, and particularly relates to a method, device and equipment for generating a phase diagram of supercritical carbon dioxide hydrate. Background Art

[0002] Since supercritical CO 2 (SC-CO 2 ) hydrate can carry a large amount of CO 2 in the form of a clathrate through a small amount of water and perform CO 2 geological sequestration, deep reservoir CO 2 geological sequestration projects have increasingly attracted wide attention. CO 2 will transform into a supercritical state at an environment of 31.1 °C and 7.38 MPa, and its physical and mechanical properties will change greatly. At the same time, due to the deep reservoir being in an environment of high geothermal temperature - high geostress - high permeability, the state of SC-CO 2 hydrate will change greatly with the changes in temperature and pressure. Therefore, calculating the phase diagram of SC-CO 2 hydrate seriously affects the CO 2 content in the deep reservoir and the safety and stability of the CO 2 geological sequestration project.

[0003] Currently, the phase diagram calculation models of SC-CO 2 hydrate mainly include the calculation models based on hydrate phase equilibrium and the calculation models based on hydrate chemical potential. The former mainly includes the Chen-Guo model and the thermodynamic and thermo-dynamic equilibrium model, and the latter mainly includes the Van der Waals-Platteeuw model. However, both of these two types of models have certain simplified model assumptions, which are different from the true phase state of SC-CO 2 hydrate. Moreover, due to the difficulty in determining the specific morphology and number of layers of the SC-CO 2 hydrate clathrate structure, it can only predict the solubility of CO 2 at different temperatures and pressures, resulting in a low calculation accuracy of the phase diagram of SC-CO 2 hydrate. The deficiencies of these two types of models are specifically manifested in the following aspects: (1) The calculation models based on hydrate phase equilibrium have a large number of assumptions based on stable equilibrium conditions and different state equations, and are mostly applicable to predicting hydrate phase equilibrium conditions, which have a large difference from the true phase state evolution of SC-CO 2 hydrate; (2) The calculation models based on hydrate chemical potential can only calculate the phase diagram and solubility of CO 2 hydrate in the generation process and the stable process, and cannot calculate the SC-CO 2The transitional critical state of the hydrate phase evolution process, thus making it impossible to construct an accurate SC-CO 2 hydrate phase diagram; (3) In existing phase diagram calculation studies, there is little consideration of the calculation of the hydrate phase diagram in porous media. However, real CO 2 hydrates exist in the porous structures and fractures of deep reservoir rocks. There is little consideration in existing phase diagram experiment and theoretical model studies of the influence of the heterogeneous structure of reservoir rocks.

[0004] Thus, existing SC-CO 2 hydrate phase diagram calculation studies all have deficiencies such as a large number of assumptions, low accuracy of the phase diagram calculation model, and little consideration of the influence of the heterogeneous structure of reservoir rocks. Therefore, it is impossible to obtain an accurate SC-CO 2 hydrate phase diagram, resulting in unclear understanding of the evolution state and capacity during CO 2 geological sequestration in deep reservoirs, seriously affecting the evaluation of the CO 2 sequestration volume and the safety and stability of CO 2 geological sequestration projects.

[0005] Regarding the problem of poor accuracy of existing hydrate phase diagram generation methods, no effective solution has been proposed yet. Summary of the Invention

[0006] The present invention provides a method, device, and equipment for generating a supercritical carbon dioxide hydrate phase diagram to solve the defect of poor accuracy of existing hydrate phase diagram generation methods.

[0007] In the first aspect, the present invention provides a method for generating a supercritical carbon dioxide hydrate phase diagram, including: Establishing an SC-CO 2 hydrate phase state data test system based on multi-source monitoring technology, and obtaining a physical data set and an image data set through the SC-CO 2 hydrate phase state data test system; Dividing the physical data set and the image data set into a training data set and a test data set; Constructing an SC-CO 2 hydrate phase state data prediction generation model, and training and optimizing the SC-CO 2 hydrate phase state data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; Obtaining data to be analyzed under different environmental conditions, predicting the data to be analyzed through the optimal generalization model to obtain a prediction result, and generating an SC-CO 2 hydrate phase diagram according to the prediction result.

[0008] According to a method for generating a supercritical carbon dioxide hydrate phase diagram provided by the present invention, the SC-CO 2 hydrate phase state data test system includes SC-CO 2 control generation system, topology control imaging system and nuclear magnetic resonance control imaging system; the SC-CO 2 hydrate phase state data test system is connected to the data control and processing module.

[0009] According to a method for generating a supercritical carbon dioxide hydrate phase diagram provided by the present invention, the SC-CO 2 control generation system includes a gas cylinder, the gas outlet end of the gas cylinder is connected to a high-pressure plunger pump, the output end of the high-pressure plunger pump is connected to a numerically controlled constant temperature water tank, an erosion-resistant container is arranged outside the numerically controlled constant temperature water tank, the output end of the numerically controlled constant temperature water tank is connected to a deionized water bottle through an erosion-resistant conduit, a quick-close valve is arranged on the erosion-resistant conduit, and the output end of the deionized water bottle is connected to the topology control imaging module.

[0010] According to a method for generating a supercritical carbon dioxide hydrate phase diagram provided by the present invention, the topology control imaging system includes a base table, an installation track is fixedly arranged on the base table through fastening bolts, a plurality of mutually corresponding X-ray probes and segmented detection plates are arranged on the installation track, and the X-ray probes and the segmented detection plates are connected to the data control and processing module through data wires.

[0011] According to a method for generating a supercritical carbon dioxide hydrate phase diagram provided by the present invention, the installation track is composed of two semi-circular multi-layer grooves, and the X-ray probes and the segmented detection plates can freely move in each layer of groove under the control of the data control and processing module.

[0012] According to a method for generating a supercritical carbon dioxide hydrate phase diagram provided by the present invention, the nuclear magnetic resonance control imaging system includes a magnetic pole system cabin body, and a magnet is arranged on the magnetic pole system cabin body; the nuclear magnetic resonance control imaging system further includes a wavelet signal amplifier, an image conversion generator and a nuclear magnetic resonance radio frequency source generator, and moreover, the nuclear magnetic resonance radio frequency source generator, the magnet, the wavelet signal amplifier, the image conversion generator and the data control and processing module are sequentially connected to form a closed loop; The nuclear magnetic resonance control imaging system further includes a loading bin and a porous medium, and the porous medium is arranged inside the loading bin; the loading bin includes an outer bin plate and an inner bin plate, both the outer bin plate and the inner bin plate are transparent; a plurality of temperature sensors and heating components are arranged between the outer bin plate and the inner bin plate; the nuclear magnetic resonance control imaging system further includes a pressure controller, and the pressure controller is connected to the temperature sensors, the heating components and the data control and processing module.

[0013] A method for generating a phase diagram of supercritical carbon dioxide hydrate according to the present invention optimizes the prediction and generation model of the phase state data of SC-CO hydrate through the training data set and the test data set to obtain an optimal generalization model, including: 2 Fine-tuning the prediction and generation model of the phase state data of SC-CO hydrate through the training data set; Inputting the test data set into the prediction and generation model of the phase state data of SC-CO hydrate to obtain predicted values; 2 Determining the relative error of the prediction and generation model of the phase state data of SC-CO hydrate according to the predicted values and the true values, and determining the optimal generalization model according to the relative error. Inputting the test data set into the prediction and generation model of the phase state data of SC-CO hydrate to obtain predicted values; 2 Determining the relative error of the prediction and generation model of the phase state data of SC-CO hydrate according to the predicted values and the true values, and determining the optimal generalization model according to the relative error. Determining the relative error of the prediction and generation model of the phase state data of SC-CO hydrate according to the predicted values and the true values, and determining the optimal generalization model according to the relative error. 2 Determining the relative error of the prediction and generation model of the phase state data of SC-CO hydrate according to the predicted values and the true values, and determining the optimal generalization model according to the relative error.

[0014] A method for generating a phase diagram of supercritical carbon dioxide hydrate according to the present invention obtains data to be analyzed under different environmental conditions, predicts the data to be analyzed through the optimal generalization model to obtain a prediction result, and generates a phase diagram of SC-CO hydrate according to the prediction result, including: 2 Inputting the data to be analyzed into the optimal generalization model, and predicting the data to be analyzed through the optimal generalization model to obtain critical temperature and critical pressure data of the phase state evolution of SC-CO hydrate; Generating the phase diagram of SC-CO hydrate according to the critical temperature and the critical pressure data. 2 Generating the phase diagram of SC-CO hydrate according to the critical temperature and the critical pressure data. Generating the phase diagram of SC-CO hydrate according to the critical temperature and the critical pressure data. 2 Generating the phase diagram of SC-CO hydrate according to the critical temperature and the critical pressure data.

[0015] In a second aspect, the present invention further provides a device for generating a phase diagram of supercritical carbon dioxide hydrate, including: An acquisition module for establishing an SC-CO hydrate phase state data test system based on multi-source monitoring technology, and acquiring a physical data set and an image data set through the SC-CO hydrate phase state data test system; 2 An acquisition module for establishing an SC-CO hydrate phase state data test system based on multi-source monitoring technology, and acquiring a physical data set and an image data set through the SC-CO hydrate phase state data test system; 2 An acquisition module for establishing an SC-CO hydrate phase state data test system based on multi-source monitoring technology, and acquiring a physical data set and an image data set through the SC-CO hydrate phase state data test system; A partitioning module for partitioning the physical data set and the image data set into a training data set and a test data set; An optimization module for constructing an SC-CO hydrate phase state data prediction and generation model, and training and optimizing the SC-CO hydrate phase state data prediction and generation model through the training data set and the test data set to obtain an optimal generalization model; 2 An optimization module for constructing an SC-CO hydrate phase state data prediction and generation model, and training and optimizing the SC-CO hydrate phase state data prediction and generation model through the training data set and the test data set to obtain an optimal generalization model; 2 An optimization module for constructing an SC-CO hydrate phase state data prediction and generation model, and training and optimizing the SC-CO hydrate phase state data prediction and generation model through the training data set and the test data set to obtain an optimal generalization model; A generation module for obtaining data to be analyzed under different environmental conditions, predicting the data to be analyzed through the optimal generalization model to obtain a prediction result, and generating SC-CO based on the prediction result 2 hydrate phase diagram

[0016] In a third aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the supercritical carbon dioxide hydrate phase diagram generation method as described in the first aspect above

[0017] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the supercritical carbon dioxide hydrate phase diagram generation method as described in the first aspect above

[0018] In a fifth aspect, the present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the supercritical carbon dioxide hydrate phase diagram generation method as described in the first aspect above

[0019] Compared with the prior art, the present invention has the following beneficial effects The supercritical carbon dioxide hydrate phase diagram generation method provided by the present invention constructs an adaptive single-layer - double-layer SC-CO 2 hydrate phase state short-term memory network (ASD-SC-CO 2 PS-LSTM), that is, an SC-CO 2 hydrate phase state data prediction and generation model. Using the training data and test data as the input of the SC-CO 2 hydrate phase state data prediction and generation model, training and optimizing the SC-CO 2 hydrate phase state data prediction and generation model to obtain an optimal generalization model. The obtained optimal generalization model can directly extract the critical data of the SC-CO 2 hydrate phase state evolution. By predicting the data to be analyzed through the optimal generalization model, the accuracy and efficiency of the prediction result can be improved. On this basis, an SC-CO 2 hydrate phase diagram is generated according to the prediction result

[0020] In addition, the present invention also breaks through the deficiency that the SC-CO 2 hydrate phase state evolution cannot be directly observed, realizes the visualization experiment of the SC-CO 2 hydrate phase state evolution, thereby reducing the economic cost of the SC-CO 2 hydrate phase state evolution observation experiment, and an SC-CO 2 hydrate phase diagram can be obtained by using a small amount of experimental data Brief Description of the Drawings

[0021] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0022] Figure 1 is a flowchart of the method for generating the phase diagram of supercritical carbon dioxide hydrate provided by the present invention; Figure 2 is a schematic diagram of the SC-CO 2 hydrate phase state data test system in an embodiment of the present invention; Figure 3 is a schematic diagram of the topology control imaging system in an embodiment of the present invention; Figure 4 is the SC-CO in an embodiment of the present invention 2 computational flowchart of the hydrate phase state data prediction generation model; Figure 5 is a schematic diagram comparing the experimental predicted values and the true values in an embodiment of the present invention; Figure 6 is the SC-CO constructed in an embodiment of the present invention 2 schematic diagram of the hydrate phase diagram; Figure 7 is a structural block diagram of the device for generating the phase diagram of supercritical carbon dioxide hydrate provided by the present invention; Figure 8 is a schematic diagram of the structure of the electronic device provided by the present invention.

[0023] Reference Numerals: 1: gas cylinder; 2: fluid flow direction; 3: high-pressure plunger pump; 4: erosion-resistant container; 5: numerically controlled constant-temperature water tank; 6: erosion-resistant conduit; 7: quick-close valve; 8: deionized water bottle; 9: magnetic pole system cabin; 10: data wire; 11: magnet; 12: wavelet signal amplifier; 13: column; 14: image conversion generator; 15: nuclear magnetic resonance radio frequency generator; 16: data control and processing module; 17: pressure controller; 18: bearing platform; 19: fastening bolt; 20: installation track; 21: outer cabin plate; 22: loading cabin; 23: injection port; 24: temperature sensor; 25: heating component; 26: segmented detection plate; 27: X-ray probe; 28: inner cabin plate; 29: porous medium; 30: fixing bolt. Detailed Embodiments

[0024] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0025] The present invention provides a method for generating a phase diagram of supercritical carbon dioxide hydrate. Figure 1 It is a flowchart of the method for generating a phase diagram of supercritical carbon dioxide hydrate provided by the present invention. As Figure 1 shown, the method includes the following steps: Step S101, establish an SC-CO 2 hydrate phase state data test system based on multi-source monitoring technology, and obtain a physical data set and an image data set through the SC-CO 2 hydrate phase state data test system; Step S102, divide the physical data set and the image data set into a training data set and a test data set; Step S103, construct an SC-CO 2 hydrate phase state data prediction generation model, and train and optimize the SC-CO 2 hydrate phase state data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; Step S104, obtain data to be analyzed under different environmental conditions, predict the data to be analyzed through the optimal generalization model to obtain a prediction result, and generate an SC-CO 2 hydrate phase diagram according to the prediction result.

[0026] The following describes the phase state evolution of SC-CO 2 hydrate with a ratio of 8:2 to water. In this method, first, an SC-CO 2 hydrate phase state data test system based on multi-source monitoring technology is established, and physical data sets such as temperature and pressure and an image data set are obtained through the SC-CO 2 hydrate phase state data test system. Then, the obtained physical data set and image data set are divided into a training data set and a test data set. Then, an adaptive single-layer and double-layer SC-CO 2 hydrate phase state short-term memory network (ASD-SC-CO 2 PS-LSTM) is constructed, that is, an SC-CO 2 hydrate phase state data prediction generation model. The training data and the test data are used as SC-CO 2 hydrate phase state data prediction generation model. The training data and the test data are used as SC-CO 2Input for the hydrate phase data prediction generation model for SC-CO 2 Train and optimize the hydrate phase data prediction generation model to obtain the best generalization model. The best generalization model obtained through training can directly extract the critical data of the SC-CO 2 hydrate phase evolution. By using the best generalization model to predict the data to be analyzed, the accuracy and efficiency of the prediction results can be improved. On this basis, generate the SC-CO 2 hydrate phase diagram according to the prediction results.

[0027] Figure 2 is the schematic diagram of the SC-CO 2 hydrate phase data test system in the embodiment of the present invention, as Figure 2 shown. In some of the embodiments, the SC-CO 2 hydrate phase data test system includes an SC-CO 2 control generation system, a topology control imaging system, and a nuclear magnetic resonance control imaging system; the SC-CO 2 hydrate phase data test system is connected to the data control and processing module 16.

[0028] Specifically, the SC-CO 2 control generation system includes a gas cylinder 1. The gas outlet end of the gas cylinder 1 is connected to a high-pressure plunger pump 3. The output end of the high-pressure plunger pump 3 is connected to a numerically controlled constant temperature water tank 5. An erosion-resistant container 4 is arranged outside the numerically controlled constant temperature water tank 5. The output end of the numerically controlled constant temperature water tank 5 is connected to a deionized water bottle 8 through an erosion-resistant conduit 6. A quick-close valve 7 is arranged on the erosion-resistant conduit 6. The output end of the deionized water bottle 8 is connected to the topology control imaging module.

[0029] Furthermore, the erosion-resistant container 4 and the erosion-resistant conduit 6 are made of supermartensitic stainless steel HP2-13Cr to prevent corrosion and leakage of the container during the SC-CO 2 injection process, resulting in the inability to reach the pressure to generate SC-CO 2 hydrate.

[0030] Figure 3 is the schematic diagram of the topology control imaging system in the embodiment of the present invention, as Figure 3 shown. The topology control imaging system includes a base 18. An installation track 20 is fixedly arranged on the base 18 through fastening bolts 19. A plurality of mutually corresponding X-ray probes 27 and segmented detection plates 26 are arranged on the installation track 20. The X-ray probes 27 and the segmented detection plates 26 are connected to the data control and processing module 16 through data wires 10.

[0031] Among them, the installation track 20 is composed of two semi-circular multi-layer grooves. The X-ray probe 27 and the segmented detection plate 26 can move freely in each layer of groove under the control of the data control and processing module 16 so as to perform SC-CO 2 phase evolution imaging. Further, the segmented detection plate 26 adopts an arc design, and each segmented detection plate 26 is provided with an independent switch to ensure SC-CO 2 phase evolution imaging at different resolutions.

[0032] The nuclear magnetic resonance control imaging system includes a magnetic pole system cabin 9, and a magnet 11 is arranged on the magnetic pole system cabin 9; the nuclear magnetic resonance control imaging system further includes a wavelet signal amplifier 12, an image conversion generator 14 and a nuclear magnetic resonance radio frequency source generator 15, and the nuclear magnetic resonance radio frequency source generator 15, the magnet 11, the wavelet signal amplifier 12, the image conversion generator 14 and the data control and processing module 16 are connected in sequence to form a closed loop.

[0033] The nuclear magnetic resonance control imaging system further includes a loading bin 22 and a porous medium 29, and the porous medium 29 is arranged inside the loading bin 22; the loading bin 22 includes an outer bin plate 21 and an inner bin plate 28, and both the outer bin plate 21 and the inner bin plate 28 are transparent; a plurality of temperature sensors 24 and heating components 25 are arranged between the outer bin plate 21 and the inner bin plate 28; the nuclear magnetic resonance control imaging system further includes a pressure controller 17, and the pressure controller 17 is connected to the temperature sensors 24, the heating components 25 and the data control and processing module 16.

[0034] Further, the outer bin plate 21 is processed and manufactured from aluminum oxynitride material, with an average light transmittance of about 80%, a flexural strength of up to 300 MPa - 400 MPa, and a high-temperature resistance temperature of up to 2200 °C to ensure smooth imaging during the experiment and not damage the equipment of the nuclear magnetic resonance control imaging system. The inner bin plate 28 is processed and manufactured from graphene transparent glass material, with an average light transmittance of about 80%, a strength tens of times higher than that of steel, and a thermal conductivity as high as 5300 W / (mK) to ensure smooth imaging and rapid temperature control during the experiment. The loading bin 22 adopts an inner-empty circular ring structure design, and the temperature sensors 24 and the multi-layer arc-shaped heating components 25 are located between the outer bin plate 21 and the inner bin plate 28 to accurately control and quickly control the temperature and pressure changes.

[0035] Through the above system, the deficiency that the phase evolution of SC-CO 2 hydrate cannot be directly observed is overcome, and the visualization experiment of the phase evolution of SC-CO 2 hydrate is realized, thereby reducing the economic cost of the observation experiment of the phase evolution of SC-CO 2 hydrate, and the phase diagram of SC-CO 2 hydrate can be obtained with a small amount of experimental data.

[0036] In some of these embodiments, in step S103, the SC-CO hydrate phase data prediction generation model is trained and optimized using a training data set and a test data set to obtain an optimal generalization model, including: fine-tuning the SC-CO hydrate phase data prediction generation model using the training data set; inputting the test data set into the SC-CO hydrate phase data prediction generation model to obtain predicted values; determining the relative error of the SC-CO hydrate phase data prediction generation model based on the predicted values and the true values, and determining the optimal generalization model based on the relative error. 2 In this embodiment, for the division of the training data set and the test data set, it can be divided according to a preset ratio. Exemplarily, 30% of the obtained physical data set and image data set is used as the training data set to train the SC-CO hydrate phase data prediction generation model. The remaining 70% of the physical data set and image data set is used as the test data set to optimize the SC-CO hydrate phase data prediction generation model. When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as shown in 2 When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as shown in 2 When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as shown in 2 When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as shown in

[0037] In this embodiment, for the division of the training data set and the test data set, it can be divided according to a preset ratio. Exemplarily, 30% of the obtained physical data set and image data set is used as the training data set to train the SC-CO hydrate phase data prediction generation model. The remaining 70% of the physical data set and image data set is used as the test data set to optimize the SC-CO hydrate phase data prediction generation model. When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as 2 In this embodiment, for the division of the training data set and the test data set, it can be divided according to a preset ratio. Exemplarily, 30% of the obtained physical data set and image data set is used as the training data set to train the SC-CO hydrate phase data prediction generation model. The remaining 70% of the physical data set and image data set is used as the test data set to optimize the SC-CO hydrate phase data prediction generation model. When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as 2 In this embodiment, for the division of the training data set and the test data set, it can be divided according to a preset ratio. Exemplarily, 30% of the obtained physical data set and image data set is used as the training data set to train the SC-CO hydrate phase data prediction generation model. The remaining 70% of the physical data set and image data set is used as the test data set to optimize the SC-CO hydrate phase data prediction generation model. When the relative error between its predicted value and the true value obtained through experiments reaches 0.95, the currently optimized phase diagram generation model is established as the optimal generalization model, as Figure 4 shown, Figure 4 This is the calculation flowchart of the SC-CO hydrate phase data prediction generation model in the embodiment of the present invention. The analysis expression of the relative error is as follows: 2 This is the calculation flowchart of the SC-CO hydrate phase data prediction generation model in the embodiment of the present invention. The analysis expression of the relative error is as follows:

[0038] Among them, V prediction represents the predicted value, V ture represents the true value. If the relative error is less than 0.95, the physical data set and the image data set are re-obtained and the SC-CO hydrate phase data prediction generation model is trained and optimized until the optimal generalization model is obtained. 2 represents the true value. If the relative error is less than 0.95, the physical data set and the image data set are re-obtained and the SC-CO hydrate phase data prediction generation model is trained and optimized until the optimal generalization model is obtained.

[0039] Based on the above embodiment, in step S104, the data to be analyzed under different environmental conditions is obtained, and the data to be analyzed is predicted through the optimal generalization model to obtain a prediction result, and an SC-CO hydrate phase diagram is generated according to the prediction result, including: inputting the data to be analyzed into the optimal generalization model, and predicting the data to be analyzed through the optimal generalization model to obtain the critical temperature and critical pressure data of the SC-CO hydrate phase evolution; generating an SC-CO hydrate phase diagram according to the critical temperature and critical pressure data; 2 Based on the above embodiment, in step S104, the data to be analyzed under different environmental conditions is obtained, and the data to be analyzed is predicted through the optimal generalization model to obtain a prediction result, and an SC-CO hydrate phase diagram is generated according to the prediction result, including: inputting the data to be analyzed into the optimal generalization model, and predicting the data to be analyzed through the optimal generalization model to obtain the critical temperature and critical pressure data of the SC-CO hydrate phase evolution; generating an SC-CO hydrate phase diagram according to the critical temperature and critical pressure data; 2 Based on the above embodiment, in step S104, the data to be analyzed under different environmental conditions is obtained, and the data to be analyzed is predicted through the optimal generalization model to obtain a prediction result, and an SC-CO hydrate phase diagram is generated according to the prediction result, including: inputting the data to be analyzed into the optimal generalization model, and predicting the data to be analyzed through the optimal generalization model to obtain the critical temperature and critical pressure data of the SC-CO hydrate phase evolution; generating an SC-CO hydrate phase diagram according to the critical temperature and critical pressure data; 2Hydrate phase diagram.

[0040] Exemplarily, the optimal generalization model is used to predict SC-CO 2 Critical temperature and adjacent pressure data of hydrate phase state evolution to construct SC-CO 2 Hydrate phase diagram, as Figures 5 - 6 shown, Figure 5 is a comparison schematic diagram of experimental predicted values and true values in the embodiments of the present invention, Figure 6 is a schematic diagram of the SC-CO 2 hydrate phase diagram constructed in the embodiments of the present invention.

[0041] In the above embodiments, the expression of the SC-CO 2 hydrate phase state data prediction generation model is as follows:

[0042] Wherein, represents the selection gate, represents the forgetting gate, represents the input gate, represents the candidate state, represents the sigmoid function, represents the weight of the image input data, represents the weight of the physical input data, represents element-wise multiplication, represents element-wise addition, W ASD represents the selection gate weight matrix, W forget represents the forgetting gate weight matrix, W input represents the input gate weight matrix, W CS represents the weight matrix candidate state, b ASD represents the selection gate bias vector, b forget represents the forgetting gate bias vector, b input represents the input gate bias vector, b CS represents the bias vector candidate state, H t-1 represents the hidden state at time t-1, X t represents the current input at time t, and tanh represents the hyperbolic tangent function, with a value range of -1 to 1.

[0043] SC-CO 2The unit update, hidden state update, and output calculation process of the hydrate phase state data prediction generation model can be expressed as follows:

[0044] Among them, represents the unit state, represents the output gate, b output represents the output gate bias vector.

[0045] In addition, when constructing the SC-CO 2 hydrate phase state data prediction generation model for the SC-CO 2 hydrate phase diagram, software such as Excel, Origin, and MATLAB can be used for plotting to draw the SC-CO 2 hydrate phase diagram.

[0046] The present invention also provides a supercritical carbon dioxide hydrate phase diagram generation device. The supercritical carbon dioxide hydrate phase diagram generation device provided by the present invention will be described below. The supercritical carbon dioxide hydrate phase diagram generation device described below can be mutually corresponding and referenced with the supercritical carbon dioxide hydrate phase diagram generation method described above. Figure 7 is the structural block diagram of the supercritical carbon dioxide hydrate phase diagram generation device provided by the present invention. As shown in Figure 7 shown, the device includes: An acquisition module 701, configured to establish an SC-CO 2 hydrate phase state data test system based on multi-source monitoring technology, and acquire a physical data set and an image data set through the SC-CO 2 hydrate phase state data test system; A division module 702, configured to divide the physical data set and the image data set into a training data set and a test data set; An optimization module 703, configured to construct an SC-CO 2 hydrate phase state data prediction generation model, and train and optimize the SC-CO 2 hydrate phase state data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; A generation module 704, configured to acquire data to be analyzed under different environmental conditions, predict the data to be analyzed through the optimal generalization model to obtain a prediction result, and generate an SC-CO 2 hydrate phase diagram according to the prediction result.

[0047] When the device is in use, first, the acquisition module 701 establishes an SC-CO 2 hydrate phase state data test system based on multi-source monitoring technology, and through the SC-CO 2The hydrate phase data test system acquires physical data sets such as temperature and pressure, as well as image data sets. Then, the partitioning module 702 partitions the acquired physical data sets and image data sets into training data sets and test data sets. The optimization module 703 then constructs an adaptive single-layer - double-layer SC-CO 2 Hydrate Phase Short-Term Memory Network (ASD-SC-CO 2 PS-LSTM), that is, SC-CO 2 Hydrate phase data prediction generation model. Using the training data and test data as the input of the SC-CO 2 Hydrate phase data prediction generation model, the SC-CO 2 Hydrate phase data prediction generation model is trained and optimized to obtain the best generalization model. The best generalization model obtained by the generation module 704 can directly extract the critical data of the SC-CO 2 Hydrate phase evolution. By using the best generalization model to predict the data to be analyzed, the accuracy and efficiency of the prediction results can be improved. On this basis, a SC-CO 2 Hydrate phase diagram is generated according to the prediction results.

[0048] Figure 8 An example of the entity structure diagram of an electronic device is shown, such as Figure 8 shown. The electronic device may include: a processor 801, a communication interface 802, a memory 803, and a communication bus 804. Among them, the processor 801, the communication interface 802, and the memory 803 complete communication with each other through the communication bus 804. The processor 801 can call the logical instructions in the memory 803 to execute the method for generating the supercritical carbon dioxide hydrate phase diagram, and this method includes: Establish a SC-CO 2 Hydrate phase data test system based on multi-source monitoring technology, and acquire physical data sets and image data sets through the SC-CO 2 Hydrate phase data test system; Partition the physical data sets and image data sets into training data sets and test data sets; Construct a SC-CO 2 Hydrate phase data prediction generation model, and train and optimize the SC-CO 2 Hydrate phase data prediction generation model through the training data sets and test data sets to obtain the best generalization model; Acquire the data to be analyzed under different environmental conditions, predict the data to be analyzed through the best generalization model to obtain the prediction results, and generate a SC-CO 2 Hydrate phase diagram according to the prediction results.

[0049] In addition, the logic instructions in the above-mentioned memory 803 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0050] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the supercritical carbon dioxide hydrate phase diagram generation method provided by the above methods, the method comprising: Establishing SC-CO based on multi-source monitoring technology 2 Hydrate phase data testing system, and through SC-CO 2 The hydrate phase data testing system acquires physical data sets and image data sets; Divide the physical dataset and the image dataset into a training dataset and a test dataset; Construction of SC-CO 2 The hydrate phase data prediction model was generated, and the SC-CO 2 The hydrate phase data prediction generation model is trained and optimized to obtain the best generalization model; Obtain the data to be analyzed under different environmental conditions, predict the data to be analyzed through the optimal generalization model, obtain the prediction results, and generate SC-CO according to the prediction results. 2 Hydrate phase diagram.

[0051] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the supercritical carbon dioxide hydrate phase diagram generation method provided by the above methods, the method comprising: Establishing SC-CO based on multi-source monitoring technology 2 Hydrate phase data testing system, and through SC-CO 2The hydrate phase data test system obtains a physical data set and an image data set; Divide the physical data set and the image data set into a training data set and a test data set; Build an SC-CO 2 Hydrate phase data prediction generation model, and use the training data set and the test data set to train and optimize the SC-CO 2 Hydrate phase data prediction generation model to obtain the best generalization model; Obtain the data to be analyzed under different environmental conditions, predict the data to be analyzed through the best generalization model to obtain the prediction result, and generate an SC-CO 2 Hydrate phase diagram.

[0052] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0053] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the above technical solutions, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to make a computer device (which can be a personal computer, server, or network device, etc.) execute the methods of each embodiment or some parts of the embodiments.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating a supercritical carbon dioxide hydrate phase diagram, characterized in that: include: Establishing a SC-CO2 hydrate phase data testing system based on multi-source monitoring technology, and acquiring a physical data set and an image data set through the SC-CO2 hydrate phase data testing system; Dividing the physical data set and the image data set into a training data set and a test data set; Constructing a SC-CO2 hydrate phase data prediction generation model, and training and optimizing the SC-CO2 hydrate phase data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; The data to be analyzed under different environmental conditions are obtained, the data to be analyzed are predicted by the optimal generalization model to obtain prediction results, and the SC-CO2 hydrate phase diagram is generated according to the prediction results.

2. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 1, characterized in that: The SC-CO2 hydrate phase data testing system comprises a SC-CO2 control generation system, a topology control imaging system and a nuclear magnetic resonance control imaging system; the SC-CO2 hydrate phase data testing system is connected to a data control processing module (16).

3. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 2, characterized in that: The SC-CO2 control generation system comprises a gas cylinder (1), the gas outlet end of the gas cylinder (1) is connected to a high-pressure plunger pump (3), the output end of the high-pressure plunger pump (3) is connected to a numerically controlled constant temperature water tank (5), a corrosion-resistant container (4) is arranged outside the numerically controlled constant temperature water tank (5), the output end of the numerically controlled constant temperature water tank (5) is connected to a deionized water bottle (8) via a corrosion-resistant conduit (6), a quick-closing valve (7) is arranged on the corrosion-resistant conduit (6), and the output end of the deionized water bottle (8) is connected to the topological control imaging module.

4. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 2, characterized in that: The topological control imaging system comprises a support platform (18), a mounting track (20) being fixedly arranged on the support platform (18) by means of fastening bolts (19), a plurality of mutually corresponding X-ray probes (27) and block detection plates (26) being arranged on the mounting track (20), and the X-ray probes (27) and the block detection plates (26) being connected to the data control processing module (16) by means of data wires (10).

5. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 4, characterized in that: The mounting track (20) is composed of two semicircular arc-shaped multi-layer grooves, and the X-ray probe (27) and the segmented detection plate (26) are controlled to move freely in each layer of the groove by the data control processing module (16).

6. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 2, characterized in that: The nuclear magnetic resonance controlled imaging system comprises a magnetic pole system cabin (9), on which a magnet (11) is arranged; the nuclear magnetic resonance controlled imaging system also comprises a wavelet signal amplifier (12), an image conversion generator (14) and a nuclear magnetic radio frequency source generator (15), and the nuclear magnetic radio frequency source generator (15), the magnet (11), the wavelet signal amplifier (12), the image conversion generator (14) and the data control processing module (16) are sequentially connected to form a closed loop; The nuclear magnetic resonance controlled imaging system further comprises a loading bin (22) and a porous medium (29), wherein the porous medium (29) is arranged inside the loading bin (22); the loading bin (22) comprises an outer bin plate (21) and an inner bin plate (28), wherein both the outer bin plate (21) and the inner bin plate (28) are transparent; a plurality of temperature sensors (24) and a heating assembly (25) are arranged between the outer bin plate (21) and the inner bin plate (28); the nuclear magnetic resonance controlled imaging system further comprises a pressure controller (17), wherein the pressure controller (17) is connected to the temperature sensor (24), the heating assembly (25) and the data control processing module (16).

7. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 1, characterized in that: The SC-CO2 hydrate phase data prediction generation model is trained and optimized by using the training data set and the test data set to obtain an optimal generalization model, including: Training and fine-tuning the SC-CO2 hydrate phase data prediction generation model using the training data set; Inputting the test data set into the SC-CO2 hydrate phase state data prediction generation model to obtain a predicted value; The relative error of the SC-CO2 hydrate phase data prediction generation model is determined according to the predicted value and the true value, and the optimal generalization model is determined according to the relative error.

8. The method for generating a supercritical carbon dioxide hydrate phase diagram according to claim 1, characterized in that: Acquiring data to be analyzed under different environmental conditions, predicting the data to be analyzed by using the optimal generalization model to obtain prediction results, and generating a SC-CO2 hydrate phase diagram according to the prediction results, including: Inputting the data to be analyzed into the optimal generalization model, predicting the data to be analyzed by the optimal generalization model, and obtaining the critical temperature and critical pressure data of the phase evolution of SC-CO2 hydrate; The SC-CO2 hydrate phase diagram is generated according to the critical temperature and the critical pressure data.

9. A supercritical carbon dioxide hydrate phase diagram generating device, characterized in that: include: An acquisition module is used to establish a SC-CO2 hydrate phase state data test system based on a multi-source monitoring technology, and to acquire a physical data set and an image data set through the SC-CO2 hydrate phase state data test system; A division module, used for dividing the physical data set and the image data set into a training data set and a test data set; An optimization module is used to construct a SC-CO2 hydrate phase data prediction generation model, and train and optimize the SC-CO2 hydrate phase data prediction generation model through the training data set and the test data set to obtain an optimal generalization model; The generation module is used to obtain the data to be analyzed under different environmental conditions, predict the data to be analyzed by using the optimal generalization model to obtain the prediction results, and generate the SC-CO2 hydrate phase diagram according to the prediction results.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the supercritical carbon dioxide hydrate phase diagram generation method according to any one of claims 1 to 8 is implemented.

Images