Oil and gas pipeline hydrate generation risk early warning and model training method and device

Abstract

The invention discloses an oil and gas pipeline hydrate generation risk early warning and model training method and device. The method comprises the following steps: obtaining training set data obtained by a hydrate generation experiment, wherein the training set data comprises gas components, the generation temperature of the hydrate and the generation pressure of the hydrate; a hydrate generation risk early warning model is trained based on the training set data, and a hydrate generation risk early warning model with prediction errors meeting requirements is obtained; using the trained hydrate generation risk early warning model to determine whether the temperature and pressure conditions in the pipeline reach the temperature and pressure conditions of hydrate phase balance so as to predict whether the risk of generating hydrate exists in the pipeline, and when the risk exists, carrying out early warning prompt in time so as to carry out risk elimination in time, such as injecting an inhibitor and preventing the occurrence of the hydrate phase balance. Generation of hydrates is avoided or reduced, oil and gas pipeline blockage caused by generation of the hydrates is avoided, and safe and normal production and transportation of natural gas are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas pipeline flow safety assurance, and in particular to an oil and gas pipeline hydrate formation risk early warning and model training method and device. Background Art

[0002] During the transportation of natural gas in oil and gas pipelines, when it flows through the throttle valve, the throttling effect will cause large changes in temperature and pressure, and hydrates are easily generated in the throttle valve. Natural gas hydrates are crystals formed by water and hydrocarbon gases bonded by hydrogen bonds. They look like ice or dense snow. They are unstable hydrates encapsulated by cage-shaped crystals. They are unconventional energy sources with high density and high calorific value. Due to the hydrate form, the flow area of ​​natural gas through the throttle valve will be reduced, and in severe cases, it may even cause serious blockage, thus affecting the normal production and transportation of natural gas. Summary of the invention

[0003] The applicant has discovered that there is no corresponding risk warning solution in the prior art for the problem that natural gas is prone to hydrate formation during oil and gas pipeline transportation. It is impossible to predict the possibility of hydrate formation in advance and eliminate risks in a timely manner. Severe blockage of throttle valves due to hydrate formation often occurs, and the normal and safe production and transportation of natural gas cannot be guaranteed.

[0004] In view of the above problems, the present invention is proposed to provide an oil and gas pipeline hydrate formation risk warning and model training method and device that overcomes the above problems or at least partially solves the above problems.

[0005] The embodiment of the present invention provides a method for training an oil and gas pipeline hydrate formation risk warning model, comprising:

[0006] Acquire training set data obtained from a hydrate generation experiment, wherein the training set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure;

[0007] The training set data are grouped to obtain multiple groups of sub-training set data, and the gas quality components and hydrate formation temperature in each group of sub-training set data are respectively input into a pre-constructed hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the sub-training set data; according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained; if the training set prediction error meets the preset error requirement, a trained hydrate formation risk warning model is obtained; if the training set prediction error does not meet the preset error requirement, the hydrate formation risk warning model is retrained using the training set data.

[0008] In some preferred embodiments, obtaining training set data obtained from a hydrate generation experiment includes:

[0009] Using a hydrate reaction device, the gas components and water are introduced into the reactor of the hydrate reaction device in a certain ratio;

[0010] By controlling the temperature and pressure in the reactor, an experiment for generating hydrates is performed to obtain the generation temperature and the generation pressure of the hydrates;

[0011] Based on the hydrate formation temperature and hydrate formation pressure corresponding to different gas quality components, a training set data for training the hydrate formation risk warning model is constructed.

[0012] In a preferred embodiment, the training set prediction error of the hydrate formation risk warning model is obtained according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the formation pressure of hydrate in the multiple groups of sub-training set data;

[0013] According to the predicted pressure of hydrate formation corresponding to each group of sub-training set data and the formation pressure of hydrate in the sub-training set data, the sub-training set prediction error of the hydrate formation risk warning model is obtained; the prediction errors of multiple sub-training sets are weighted averaged to obtain the training set prediction error of the hydrate formation risk warning model; or

[0014] Based on the predicted pressure of hydrate formation corresponding to each group of sub-training set data, the predicted pressure of hydrate formation corresponding to the sub-training set data is obtained. According to the predicted pressure of hydrate formation corresponding to the training set data and the formation pressure of hydrate in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained.

[0015] In a preferred embodiment, a method for training an oil and gas pipeline hydrate formation risk warning model further includes:

[0016] Acquire validation set data obtained from a hydrate generation experiment, wherein the validation set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure;

[0017] Inputting the gas quality components and hydrate formation temperature in the validation set data into the trained hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the validation set data;

[0018] According to the predicted pressure of hydrate formation corresponding to the validation set data and the hydrate formation pressure in the validation set, the validation set prediction error of the hydrate formation risk warning model is determined; according to the validation set prediction error and the training set prediction error, the quality of the trained hydrate formation risk warning model is evaluated.

[0019] Preferably, in some embodiments, the hydrate formation risk warning model is a convolutional neural network model, which includes an input layer, a convolution layer, a batch normalization layer, an activation layer, a dropout layer, a fully connected layer, and a regression layer;

[0020] The input layer serves as an input interface for gas quality components and hydrate formation temperature;

[0021] The convolution layer is used to perform convolution operation and mapping on the gas quality components and hydrate formation temperature input by the input layer to obtain characteristic data of the gas quality components and temperature;

[0022] The batch normalization layer is used to normalize the characteristic data of the temperament components and temperature to obtain the normalized characteristic data of the temperament components and temperature;

[0023] The activation layer is used to perform nonlinear mapping on the normalized characteristic data of the temperament components and temperature to obtain a continuous probability density space of the characteristic data of the temperament components and temperature;

[0024] The dropout layer is used to inactivate the parameters of the convolutional neural network model, and use the continuous probability density space as input data to obtain test parameters that meet the hydrate formation risk warning model;

[0025] The fully connected layer is used to take the continuous probability density space as input data and the test parameter as a parameter, and output the predicted pressure of hydrate formation.

[0026] The embodiment of the present invention provides an oil and gas pipeline hydrate formation risk early warning model method, comprising:

[0027] Obtain the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain pipeline temperature and pressure data;

[0028] The gas quality components of the transported material in the oil and gas pipeline and the pipeline temperature are input into the trained oil and gas pipeline hydrate formation risk warning model, and the predicted pressure is output to obtain predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure; the oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method as described above;

[0029] Compare the detected pipeline temperature and pressure data with the predicted temperature and pressure data;

[0030] The temperature and pressure range for hydrate formation is determined based on the predicted stabilization pressure data; if the pipeline temperature and pressure data is not within the temperature and pressure range for hydrate formation, no warning is issued; if the pipeline temperature and pressure data is within the temperature and pressure range for hydrate formation, a warning is issued.

[0031] The preferred embodiment also includes: if the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, determining the injection time and injection position of the injection inhibitor, and injecting the inhibitor at the injection position in the pipeline at the injection time.

[0032] An embodiment of the present invention provides an oil and gas pipeline hydrate formation risk warning model training device, comprising:

[0033] A sample data acquisition module is used to acquire training set data obtained from a hydrate generation experiment, wherein the training set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure;

[0034] The model training module is used to group the training set data to obtain multiple groups of sub-training set data, and input the gas quality components and hydrate formation temperature in each group of sub-training set data into a pre-constructed hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the sub-training set data; according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained; if the training set prediction error meets the preset error requirement, a trained hydrate formation risk warning model is obtained; if the training set prediction error does not meet the preset error requirement, the hydrate formation risk warning model is retrained using the training set data.

[0035] The embodiment of the present invention further provides an oil and gas pipeline hydrate formation risk warning device, comprising:

[0036] The temperature and pressure data acquisition module is used to acquire the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain the pipeline temperature and pressure data;

[0037] A prediction module is used to input the gas quality components of the transported material in the oil and gas pipeline and the pipeline temperature into a trained oil and gas pipeline hydrate formation risk warning model, output a predicted pressure, and obtain predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure; the oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method as described above;

[0038] The early warning module is used to determine the temperature and pressure range of hydrate formation based on the predicted temperature and pressure data; if the pipeline temperature and pressure data are not within the temperature and pressure range of hydrate formation, no early warning prompt is given; if the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, an early warning prompt is given.

[0039] An embodiment of the present invention provides a computer storage medium, in which computer executable instructions are stored. When the computer executable instructions are executed by a processor, the above-mentioned oil and gas pipeline hydrate formation risk warning model training method and / or the above-mentioned oil and gas pipeline hydrate formation risk warning method are implemented.

[0040] An embodiment of the present invention provides an oil and gas pipeline hydrate formation risk warning 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 above-mentioned oil and gas pipeline hydrate formation risk warning model training method and / or the above-mentioned oil and gas pipeline hydrate formation risk warning method are implemented.

[0041] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:

[0042] The oil and gas pipeline hydrate formation risk warning model training method provided by the embodiment of the present invention trains a pre-constructed hydrate formation risk warning model based on the training set data obtained from the hydrate formation experiment, and obtains a warning model that can accurately predict the temperature and pressure conditions for hydrate formation. The warning model can be used to accurately predict the possibility of hydrate formation. When hydrate formation is likely to occur, early warning prompts are given in time to eliminate risks in time, avoid oil and gas pipeline blockage caused by hydrate formation, and ensure the safe and normal production and transportation of natural gas.

[0043] The oil and gas pipeline hydrate formation risk warning method provided in the embodiment of the present invention uses a trained hydrate formation risk warning model to determine whether the temperature and pressure conditions in the pipeline have reached the temperature and pressure conditions for hydrate crystallization based on the detected pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline, thereby predicting whether there is a risk of hydrate formation in the pipeline, and promptly giving an early warning when the risk exists, so as to timely eliminate the risk, such as injecting inhibitors to avoid or reduce hydrate formation, avoid oil and gas pipeline blockage caused by hydrate formation, and ensure the safe and normal production and transportation of natural gas.

[0044] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0045] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0047] Figure 1 This is a flow chart of a method for training a hydrate formation risk warning model for an oil and gas pipeline in an embodiment of the present invention;

[0048] Figure 2 This is a flow chart of another oil and gas pipeline hydrate formation risk warning model verification method in an embodiment of the present invention;

[0049] Figure 3 This is an overall flow chart of the early warning of hydrate formation risk in oil and gas pipelines in an embodiment of the present invention;

[0050] Figure 4 A hydrate reaction device in an embodiment of the present invention;

[0051] Figure 5 This is an example diagram of the prediction effect of the training set in an embodiment of the present invention;

[0052] Figure 6 This is an example diagram of the prediction effect of the validation set in an embodiment of the present invention;

[0053] Figure 7 This is a flow chart of a method for early warning of hydrate formation risk in an oil and gas pipeline according to an embodiment of the present invention;

[0054] Figure 8 Schematic diagram of a hydrate formation risk early warning system in an embodiment of the present invention;

[0055] Fig. 9 This is a comparison effect diagram of the predicted data and experimental data of the hydrate formation risk warning model in the embodiment of the present invention;

[0056] Fig.10 Schematic diagram of the structure of the oil and gas pipeline hydrate formation risk warning model training device in an embodiment of the present invention;

[0057] Fig.11 It is a schematic diagram of the structure of the oil and gas pipeline hydrate formation risk warning device in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0059] In order to solve the problems existing in the prior art, the embodiments of the present invention provide a method for training a hydrate formation risk warning model for oil and gas pipelines and a method for warning the hydrate formation risk for oil and gas pipelines, which can be used for the technical guarantee of flow safety at the wellhead throttle valve position. The method substitutes the existing experimental data into the convolutional neural network model training model parameters to obtain a hydrate formation risk warning model, establishes a corresponding data acquisition system at a specified position in the pipeline, such as before and after the wellhead throttle valve, collects temperature and pressure data, obtains the flow state at the specified position in the pipeline by inputting the collected pipeline state parameters into the warning model, and determines the hydrate non-formation area and the formation area. The warning method can effectively avoid the risk of hydrate formation, improve economic efficiency and ensure flow safety.

[0060] The embodiment of the present invention provides a method for training a hydrate formation risk warning model for an oil and gas pipeline, the process of which is as follows: Figure 1 As shown, the following steps are included:

[0061] Step S101: obtaining training set data obtained from a hydrate formation experiment, wherein the training set data includes gas quality components, hydrate formation temperature, and hydrate formation pressure.

[0062] Step S102: Group the training set data to obtain multiple groups of sub-training set data.

[0063] Step S103: inputting the gas quality components and hydrate formation temperature in each group of sub-training set data into the pre-built hydrate formation risk warning model respectively, and obtaining the predicted pressure of hydrate formation corresponding to the sub-training set data.

[0064] Step S104: obtaining a training set prediction error of the hydrate formation risk warning model according to the predicted pressures of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data.

[0065] Step S105: Determine whether the prediction error meets the preset error requirement.

[0066] If the training set prediction error meets the preset error requirement, execute step S106; if the training set prediction error does not meet the preset error requirement, execute step S107, return to step S102 or step S103, and use the training set data to retrain the hydrate formation risk warning model;

[0067] Step S106: obtaining a trained hydrate formation risk warning model.

[0068] In some optional embodiments, see Figure 2The above-mentioned oil and gas pipeline hydrate formation risk warning model training method also includes, after obtaining the trained oil and gas pipeline hydrate formation risk warning model, verifying the trained oil and gas pipeline hydrate formation risk warning model, the verification process includes:

[0069] Step S201: Acquire validation data obtained from a hydrate formation experiment, where the validation data includes gas quality components, hydrate formation temperature, and hydrate formation pressure.

[0070] Step S202: inputting the gas quality components and hydrate formation temperature in the validation set data into the trained hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the validation set data.

[0071] Step S203: Determine the validation set prediction error of the hydrate formation risk warning model according to the predicted pressure of hydrate formation corresponding to the validation set data and the formation pressure of hydrate in the validation set.

[0072] Step S204: evaluating the quality of the trained hydrate formation risk warning model according to the prediction error of the validation set and the prediction error of the training set.

[0073] Optionally, the training set data and the validation set data in the above method can be obtained by separately obtaining the training set data through experiments, and then separately obtaining the validation set data through experiments, or by obtaining sample data through experiments and then dividing the obtained sample data into training set and validation set. For the implementation method of dividing the obtained sample data after obtaining the sample data, see Figure 3 The overall process of early warning of hydrate formation risk in oil and gas pipelines shown in FIG. 1 includes:

[0074] Step S301: selecting a database, wherein the data in the selected database may be sample data obtained by experiments, including, referring to steps S101 and S201, the sample data including gas quality components, hydrate formation temperature, and hydrate formation pressure.

[0075] Step S302: using the first proportion of sample data in the database as a training set, where the training set is used to debug model parameters;

[0076] Step S303: using the second proportion of sample data in the database as a validation set, where the validation set is used to verify the actual learning ability of the model;

[0077] According to the needs of model training, the data in the selected database can be divided into two parts, one as a training set and the other as a validation set. For example, 80% of the sample data is used as a training set and 20% of the sample data is used as a validation set.

[0078] Step S304: constructing a hydrate formation risk warning model.

[0079] Step S305: Use the training set and the validation set to complete the training of the constructed hydrate formation risk warning model.

[0080] After the hydrate formation risk warning model is built, it can be trained using the training set, see steps S102-S106, to obtain a trained hydrate formation risk warning model; the trained model is verified using the validation set, see steps S201-204, to obtain a verified hydrate formation risk warning model, and finally the model training is completed.

[0081] Step S306: Acquire on-site data and predict new on-site data;

[0082] The new on-site data may be the temperature and pressure data of a specified location in the oil and gas pipeline to be predicted.

[0083] Step S307: Use the model to predict the field data, and if it is determined that there is a risk, issue a hydrate formation risk warning. The specific implementation process of using the model for warning will be described in detail later.

[0084] In the above step S101, the training set data is obtained by using a hydrate reaction device, and gas components and water are introduced into the reactor of the hydrate reaction device in a certain ratio; by controlling the temperature and pressure in the reactor, an experiment of generating hydrates is carried out to obtain the hydrate generation temperature and the hydrate generation pressure; based on the hydrate generation temperature and the hydrate generation pressure corresponding to different gas components, the training set data for training the hydrate generation risk warning model is constructed.

[0085] An optional specific structure of the hydrate reaction device is shown in FIG. Figure 4 , which may be a high-pressure visual hydrate reaction device, the hydrate reaction device includes a gas supply system, a pressure control system, a hydrate reaction system, a data acquisition system, a temperature control system and a display;

[0086] The gas supply system also includes gas cylinders, which are used to store the gas required for the hydrate synthesis reaction test, for example, the gas stored is the natural gas transported in the oil and gas pipeline; the hydrate reaction system also includes a hydrate reactor, and the model, size, material, etc. of the hydrate reactor can be selected and designed according to needs, for example: the reactor body of the hydrate reactor is made of 316L stainless steel, with a net volume of 4L, equipped with a sapphire window, and the reactor volume is controlled by a supporting iron block (316L, 0.8L); the temperature control system also includes a water bath control, which also includes a reactor water bath jacket, a DC-3010 low-temperature water bath and a constant temperature water bath, wherein the temperature control range is -20℃-80℃, and the control accuracy is 0.1K or other accuracies that meet the requirements; the data acquisition system also includes a temperature sensor, a pressure sensor and a data acquisition card. The data acquisition system collects the temperature and pressure of the hydrate formation experiment. For example, temperature detection uses a T-type thermocouple, and its accuracy can be 0.1K or other accuracies that meet the requirements. Pressure monitoring uses a pressure sensor, and the pressure sensor is used to collect the pressure of hydrate formation. The accuracy of the pressure sensor can be selected as needed, for example, a pressure sensor with an accuracy of 0.01Mpa is selected. The information acquisition card can be selected as needed, and its performance can meet the requirements of data acquisition processing. The information acquisition card can use a T-type thermocouple to check the temperature of hydrate formation; the display is used to display the detected and collected data.

[0087] The high-pressure visible hydrate reactor system is used to conduct experiments. Gas and water can be introduced into the reactor in a certain proportion. The high-pressure visible hydrate reactor system is used to conduct hydrate formation experiments. The temperature (T) and pressure (P) of hydrate formation are collected by the data acquisition system of the reactor system. The operation process of the hydrate formation experiment includes:

[0088] Before the experiment, the hydrate reaction experimental device was cleaned internally with deionized water and carbon dioxide, and then an appropriate amount of deionized water was added after cleaning;

[0089] The temperature of the reactant in the hydrate reactor is controlled by a water bath. When the temperature of the deionized water in the hydrate reactor reaches a predetermined temperature, the exhaust valve is opened and the gas in the gas cylinder is introduced. After the air in the hydrate reactor is exhausted, the exhaust valve is closed. The ventilation time is about 3 minutes.

[0090] When the pressure obtained by the data acquisition system reaches a predetermined pressure, the air inlet is closed and the magnetic stirring is turned on. The speed can be set, for example, to 200 rpm;

[0091] The temperature and pressure are detected by the temperature sensor and pressure sensor of the data acquisition system, and the data acquisition system records the temperature and pressure values, and displays the changes in temperature and pressure on the display; the experimental process in the reactor is recorded by the camera at the sapphire window; when the pressure obtained by the above data acquisition system remains unchanged for a certain period of time, such as 1 hour, and the temperature has no obvious rising trend, it is considered that the hydrate reaction is basically completed, and the group of experiments is ended. This group of experiments is repeated multiple times, for example, three times.

[0092] After the experiment is completed, the temperature and pressure data of hydrate formation in the experimental data are extracted, and the gas composition is normalized to obtain the experimental data. The experimental data obtained include gas composition, hydrate formation temperature, and hydrate formation pressure. The obtained experimental data can be used as training set data.

[0093] In the above step S102, when the training set data is grouped, the number of groups can be determined according to the needs of model training, for example, 100 groups or 200 groups can be divided to obtain multiple groups of sub-training set data. The number of sample data in each group can also be set as needed, and each sample data includes gas quality components, hydrate formation temperature, and hydrate formation pressure.

[0094] In the above step S103, the pre-constructed hydrate formation risk warning model is trained using the grouped sub-training set data, and each sub-training set data may be trained once. The hydrate formation risk warning model may use a convolutional neural network model.

[0095] The convolutional neural network model includes input layer, convolution layer, batch normalization layer, activation layer, dropout layer, fully connected layer, and regression layer;

[0096] The input layer serves as the input interface for gas composition and hydrate formation temperature;

[0097] The convolution layer is used to perform convolution operation and mapping on the gas composition and hydrate formation temperature input by the input layer to obtain the characteristic data of gas composition and temperature;

[0098] The convolution layer extracts features from the input parameters. The specific gas composition and temperature initial data are mapped through one-dimensional convolution operation to obtain the data features of gas composition and temperature. The convolution layer can be expressed as: Where * represents a one-dimensional convolution operation, Represents the convolution kernel The jth feature map generated, j∈[1,n c ],n c Indicates the number of convolution kernels, is a matrix composed of weights, where m is the size of the convolution kernel and b is the bias. r(·) is the activation function, which can provide the nonlinear modeling ability of the network and realize the nonlinear mapping learning ability of the deep neural network layer. Common activation functions include: ReLU, Sigmoid, tanh and other functions. In this example, two convolution kernels are used, with sizes of 3*1 and 2*1 respectively, and the stride is 1.

[0099] The batch normalization layer is used to normalize the characteristic data of the gas composition and temperature to obtain the normalized characteristic data of the gas composition and temperature;

[0100] The specific batch normalization layer normalizes the data features obtained after the convolution operation, and the output of each layer of the network is activated. Through the batch normalization layer, the original data distribution is destroyed, which alleviates overfitting to a certain extent and reduces the artificial selection of parameters and the requirements for the learning rate. You can use a large initial learning rate or a smaller learning rate, and the algorithm can also quickly train and converge; reduce gradient vanishing, speed up convergence, and improve training accuracy.

[0101] The activation layer is used to perform nonlinear mapping on the normalized characteristic data of the temperament components and temperature to obtain the continuous probability density space of the characteristic data of the temperament components and temperature;

[0102] The activation layer performs a nonlinear mapping on the data feature results of the convolution calculation of the previous layer to form a continuous probability density space. In this example, the activation layer uses an activation function (ReLU) to avoid gradient vanishing during training.

[0103] The dropout layer is used to inactivate the parameters of the convolutional neural network model and use the continuous probability density space as input data to obtain test parameters that meet the hydrate formation risk warning model;

[0104] The dropout layer can randomly ignore some weights of the convolutional neural network model parameters, and use the continuous probability density space as data to train the model that has ignored some weights of the convolutional neural network model parameters to obtain the predicted test parameters. Specifically, the dropout layer can avoid overfitting, and dropout is used in layers with more parameters such as the fully connected layer; in the training of a neural network containing a dropout layer, each batch of training data is randomly selected, and multiple sub-neural networks are actually trained. The positions of the randomly ignored weights are different in different sub-networks. Finally, during the test process, these small sub-networks are combined to make predictions similar to a voting mechanism.

[0105] The fully connected layer is used to take the continuous probability density space as input data, the test parameters as parameters, and output the predicted pressure of hydrate formation. The main function of the fully connected layer is to integrate the features calculated by the previous layers (convolutional layer, batch normalization layer) into one value. Its advantage is that it reduces the influence of feature position on the classification results and improves the robustness of the entire network.

[0106] The constructed model can also use the RMSProp optimizer to optimize the parameters of the convolutional neural network model, reducing the number of iterations required to reach the optimal value and improving the ability of the optimization algorithm.

[0107] In the above step S103, the temperament components and temperature in the training set are used as input variables, and the pressure is used as the output variable, and are substituted into the model to iteratively train the model parameters.

[0108] In the above step S104, the prediction error of the training set of the hydrate formation risk warning model is obtained according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the formation pressure of hydrate in the training set data. The prediction error of the training set can be determined in one of the following ways, or in other ways. The prediction error parameter can be selected as needed, for example, at least one of the root mean square error (RMSE), mean absolute error (MAE), and mean bias error (MBE) can be used.

[0109] Method 1:

[0110] According to the predicted pressure of hydrate formation corresponding to each group of sub-training set data and the hydrate formation pressure in the sub-training set data, the sub-training set prediction error of the hydrate formation risk warning model is obtained; the weighted average of multiple sub-training set prediction errors is performed to obtain the training set prediction error of the hydrate formation risk warning model.

[0111] Method 2:

[0112] Based on the predicted pressure of hydrate formation corresponding to each group of sub-training set data, the predicted pressure of hydrate formation corresponding to the training set data is obtained. According to the predicted pressure of hydrate formation corresponding to the training set data and the formation pressure of hydrate in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained.

[0113] The process of obtaining the validation set data in the above step S201 may refer to the description of step S101.

[0114] The above steps S202-203 use the validation set to input the model type, and the process of obtaining the prediction error of the validation set can be seen in the description of steps S103 and S104. The difference is that the validation set is not grouped, and the validation set data is used to input the model to obtain the prediction error.

[0115] In the above step S204, the prediction error of the validation set is compared with the prediction error of the training set, and the quality of the trained hydrate formation risk warning model is evaluated according to the degree of difference between the prediction error of the validation set and the prediction error of the training set. For example, if the difference between the prediction error of the validation set and the prediction error of the training set is small and the deviation between the two is less than a certain value, it is considered that the trained model meets the requirements. If the deviation between the two is greater than a certain value, it is considered that the trained model does not meet the requirements.

[0116] The above method of the embodiment of the present invention uses a training set to train the model, adjusts the model parameters, uses a validation set to validate the model, evaluates the model generalization performance, optimizes the model parameters, improves the model accuracy, and enables the trained network model to meet the wellhead throttle valve hydrate early warning effect. The prediction effect of the training set is specifically referred to in Figure 5 As shown in the figure, the hydrate formation risk warning model is trained using the training set data to obtain the training effect of the hydrate formation risk warning model. Among them, the solid line represents the baseline when there is no deviation between the test pressure and the predicted pressure, and the line composed of multiple points is the model predicted pressure. From the comparison of the two curves, a training hydrate formation risk warning model that meets expectations can be obtained.

[0117] For details on the prediction effect of the validation set in the embodiment of the present invention, see Figure 6 As shown in the figure, after the hydrate formation risk warning model is verified using the validation set data, the prediction effect of the hydrate formation risk warning model. Among them, the solid line represents the baseline when there is no deviation between the test pressure and the predicted pressure, and the line composed of multiple points is the model predicted pressure. From the comparison of the two curves, a verified hydrate formation risk warning model that meets expectations can be obtained.

[0118] In the embodiment of the present invention, the prediction error of the sub-training set and the prediction error of the validation set include the root mean square error (RMSE), the mean absolute error (MAE), and the mean bias error (MBE). The data in the chart are the prediction error of the training set and the prediction error of the validation set in the embodiment.

[0119]

[0120]

[0121] Based on the same inventive concept, an embodiment of the present invention further provides an oil and gas pipeline hydrate formation risk warning method, the flow chart of which is shown in FIG7 , comprising the following steps:

[0122] Step S701: Acquire the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain pipeline temperature and pressure data.

[0123] See Figure 8As shown, the data acquisition system includes a temperature sensor and a pressure sensor. The data acquisition system uses the temperature sensor and the pressure sensor to collect the temperature and pressure data before and after the specified position of the oil and gas pipeline, and obtains the temperature and pressure data of the specified position of the oil and gas pipeline. The specified position can be before and after the throttle valve in the pipeline, for example: the wellhead regulates the pressure of the produced gas and the angle throttle valve that adjusts the flow rate. The data acquisition system includes a temperature sensor and a pressure sensor. A temperature sensor and a pressure sensor can be set before and after the wellhead throttle valve to collect the temperature and pressure before and after the valve.

[0124] Step S702: Input the gas quality components and pipeline temperature of the transported material in the oil and gas pipeline into the trained oil and gas pipeline hydrate formation risk warning model, output the predicted pressure, and obtain the predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure. The oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method as described above.

[0125] The oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method described above. The temperature data of the oil and gas pipeline and the gas quality components in the pipeline are collected through the data acquisition system and input into the prediction model, and the predicted pressure is output and the hydrate phase equilibrium curve is predicted. Fig. 9 As shown, the experimental natural gas hydrate phase equilibrium curve and the predicted hydrate phase equilibrium curve are shown. The two curves are relatively close, indicating that the accuracy of the model prediction results can meet the requirements.

[0126] Step S703: Determine the temperature and pressure range for hydrate formation based on the predicted temperature and pressure data.

[0127] See also Fig. 9 As shown in the figure, after the temperature and pressure before and after the valve collected by the sensor are input into the early warning model, if the temperature and pressure point is in the upper left corner of the predicted phase equilibrium curve, there is a risk of hydrate formation and a risk alarm is issued; if it is in the lower right corner of the phase equilibrium curve, there is no risk of hydrate formation and no alarm is issued.

[0128] Step S704: If the pipeline temperature and pressure data are not within the temperature and pressure range for hydrate formation, no warning prompt is issued.

[0129] Step S705: If the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, an early warning prompt is issued.

[0130] Compare the detected pipeline temperature and pressure data with the predicted temperature and pressure data.

[0131] See Fig. 9As shown, the collected temperature and pressure data at the specified position of the oil and gas pipeline are compared with the predicted hydrate phase equilibrium curve; if the temperature and pressure point is in the upper left corner of the predicted phase equilibrium curve, there is a risk of hydrate formation and a risk alarm is issued; if it is in the lower right corner of the phase equilibrium curve, there is no risk of hydrate formation and no alarm is issued.

[0132] In some optional embodiments, after the above step S705 determines that the pipeline temperature and pressure data is within the temperature and pressure range of hydrate formation, it also includes:

[0133] Step S706: Determine the injection time and injection position of the injection inhibitor, and inject the inhibitor at the injection position in the pipeline at the injection time. Figure 8 As shown, the prediction of hydrate formation risk can be realized on the computer equipment and displayed on the display port. When it is determined that there is a risk, the warning result is fed back to the throttle valve to indicate the risk of hydrate blockage, and an inhibitor is injected at the specified position of the pipeline. The inhibitor can be an alcohol substance.

[0134] For details on the comparison between the predicted data and experimental data of the hydrate formation risk warning model in the embodiment of the present invention, see Fig. 9 , wherein the solid line is the prediction effect of the hydrate phase equilibrium curve generated by the prediction data obtained by using the hydrate formation risk warning, and the dotted solid line is the hydrate phase equilibrium curve generated by the experimental data obtained by the hydrate formation experiment. By comparing the two curves, a hydrate formation risk warning model that meets expectations can be obtained.

[0135] The above method of the embodiment of the present invention is based on the wellhead throttle hydrate early warning of the convolutional neural network, which solves the problem that the hydrate generation before and after the wellhead throttle cannot be predicted in the prior art. The method trains a hydrate generation risk early warning model through a convolutional neural network according to experimental data, generates a predicted hydrate phase equilibrium curve through the model, establishes a corresponding data acquisition system before and after the wellhead throttle, obtains temperature and pressure data, and compares it with the predicted hydrate phase equilibrium curve. If the temperature and pressure point of the wellhead throttle is in the upper left of the hydrate phase equilibrium curve predicted by the network model, there is a risk of hydrate generation; if the temperature and pressure point is in the lower right, there is no risk of hydrate generation. Early warning of the risk of hydrate generation. Based on a large amount of experimental data, the deep learning algorithm is applied to the prediction of hydrate blockage in the pipeline, and the hydrate generation in the pipeline is accurately predicted to realize the early warning function, and the accurate time point of injecting the inhibitor is determined to prevent the generation of hydrates in the pipeline from blocking the pipeline, thereby improving economic benefits.

[0136] Based on the same inventive concept, the embodiment of the present invention also provides a device for training a hydrate formation risk warning model for oil and gas pipelines. The device can be arranged in a computer device with a computing and processing function. The structure of the device is shown in FIG. Fig.10As shown, it includes a sample data acquisition module 11 and a model training module 12;

[0137] The sample data acquisition module 11 is used to acquire the training set data obtained from the hydrate generation experiment, wherein the training set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure;

[0138] The model training module 12 is used to group the training set data to obtain multiple groups of sub-training set data, and input the gas quality components and hydrate formation temperature in each group of sub-training set data into the pre-constructed hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the sub-training set data; according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained; if the training set prediction error meets the preset error requirement, a trained hydrate formation risk warning model is obtained; if the training set prediction error does not meet the preset error requirement, the hydrate formation risk warning model is retrained using the training set data.

[0139] In some optional embodiments, the above device further includes: a model verification module 13;

[0140] The sample data acquisition module 11 is also used to obtain the validation set data obtained from the hydrate generation experiment, and the validation set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure;

[0141] The model verification module 13 is used to input the gas quality components and hydrate formation temperature in the verification set data into the trained hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the verification set data; determine the verification set prediction error of the hydrate formation risk warning model according to the predicted pressure of hydrate formation corresponding to the verification set data and the hydrate formation pressure in the verification set; and evaluate the quality of the trained hydrate formation risk warning model according to the verification set prediction error and the training set prediction error.

[0142] Based on the same inventive concept, the embodiment of the present invention also provides an oil and gas pipeline hydrate formation risk warning device, which can be set in a computer device with a computing and processing function. The structure of the device is shown in FIG. Fig.11 As shown, it includes a temperature and pressure data acquisition module 21, a prediction module 22, and an early warning module 23;

[0143] The temperature and pressure data acquisition module 21 is used to acquire the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain pipeline temperature and pressure data;

[0144] The prediction module 22 is used to input the gas quality components of the transported material in the oil and gas pipeline and the pipeline temperature into the trained oil and gas pipeline hydrate formation risk warning model, output the predicted pressure, and obtain the predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure; the oil and gas pipeline hydrate formation risk warning model is trained using the above-mentioned oil and gas pipeline hydrate formation risk warning model training method;

[0145] The early warning module 23 is used to determine the temperature and pressure range of hydrate formation according to the predicted temperature and pressure data; if the pipeline temperature and pressure data are not within the temperature and pressure range of hydrate formation, no early warning prompt is given; if the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, an early warning prompt is given.

[0146] In some optional embodiments, the early warning module 23 is further used to determine the injection time and injection position of the inhibitor, and the inhibitor is injected at the injection position in the pipeline at the injection time.

[0147] Regarding the oil and gas pipeline hydrate formation risk warning model training and oil and gas pipeline hydrate formation risk warning device in the above-mentioned embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0148] An embodiment of the present invention provides a computer storage medium, in which computer executable instructions are stored. When the computer executable instructions are executed by a processor, the oil and gas pipeline hydrate formation risk warning model training method and / or the oil and gas pipeline hydrate formation risk warning method as described above are implemented.

[0149] An embodiment of the present invention provides an oil and gas pipeline hydrate formation risk warning 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 oil and gas pipeline hydrate formation risk warning model training method and / or the oil and gas pipeline hydrate formation risk warning method as described above are implemented.

[0150] Unless otherwise specifically stated, terms such as processing, computing, calculating, determining, displaying, etc. may refer to the actions and / or processes of one or more processing or computing systems, or similar devices, which operate and convert data represented as physical (e.g., electronic) quantities within registers or memories of a processing system into other data similarly represented as physical quantities within memories, registers, or other such information storage, transmission, or display devices of the processing system. Information and signals may be represented using any of a variety of different techniques and methods. For example, data, instructions, commands, information, signals, bits, symbols, and chips mentioned throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, light fields or particles, or any combination thereof.

[0151] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of protection of the present disclosure. The attached method claims present the elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.

[0152] In the above detailed description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure should not be interpreted as reflecting an intention that the embodiments of the claimed subject matter require more features than are clearly stated in each claim. On the contrary, as reflected in the appended claims, the invention is in a state of having less than all the features of the disclosed individual embodiments. Therefore, the appended claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.

[0153] Those skilled in the art will also appreciate that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the embodiments herein can all be implemented as electronic hardware, computer software, or a combination thereof. In order to clearly illustrate the interchangeability between hardware and software, various illustrative components, blocks, modules, circuits, and steps are generally described above around their functions. Whether such functions are implemented as hardware or software depends on specific applications and the design constraints imposed on the entire system. A skilled person can implement the described functions in an alternative manner for each specific application, but such implementation decisions should not be interpreted as departing from the scope of protection of the present disclosure.

[0154] The steps of the method or algorithm described in conjunction with the embodiments herein may be directly embodied as hardware, a software module executed by a processor, or a combination thereof. The software module may be located in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a mobile disk, a CD-ROM, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor so that the processor can read information from the storage medium and can write information to the storage medium. Of course, the storage medium may also be an integral part of the processor. The processor and the storage medium may be located in an ASIC. The ASIC may be located in a user terminal. Of course, the processor and the storage medium may also be present in a user terminal as discrete components.

[0155] For software implementation, the techniques described in this application can be implemented with modules (e.g., procedures, functions, etc.) that perform the functions described in this application. These software codes can be stored in a memory unit and executed by a processor. The memory unit can be implemented within the processor or outside the processor. In the latter case, it is coupled to the processor in a communication manner via various means, which are well known in the art.

[0156] The above description includes examples of one or more embodiments. Of course, it is impossible to describe all possible combinations of components or methods for the purpose of describing the above embodiments, but it should be recognized by those skilled in the art that the various embodiments may be further combined and arranged. Therefore, the embodiments described herein are intended to cover all such changes, modifications and variations that fall within the scope of protection of the appended claims. In addition, with respect to the term "including" used in the specification or claims, the word is covered in a manner similar to the term "including", just as "including," is explained as a transitional word in the claims. In addition, any term "or" used in the specification of the claims is intended to mean "non-exclusive or".

Claims

1. A method for training an oil and gas pipeline hydrate formation risk warning model, characterized in that: include: Acquire training set data obtained from a hydrate generation experiment, wherein the training set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure; Group the training set data to obtain multiple groups of sub-training set data; The gas quality components and the hydrate formation temperature in each group of sub-training set data are respectively input into the pre-constructed hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the sub-training set data; according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained; if the training set prediction error meets the preset error requirement, a trained hydrate formation risk warning model is obtained; If the prediction error of the training set does not meet the preset error requirement, the hydrate formation risk warning model is retrained using the training set data.

2. The method according to claim 1, characterized in that The training set data obtained from the hydrate generation experiment includes: Using a hydrate reaction device, the gas components and water are introduced into the reactor of the hydrate reaction device in a certain ratio; By controlling the temperature and pressure in the reactor, an experiment for generating hydrates is conducted to obtain the hydrate generation temperature and the hydrate generation pressure; Based on the hydrate formation temperature and hydrate formation pressure corresponding to different gas quality components, a training set data for training the hydrate formation risk warning model is constructed.

3. The method according to claim 1, characterized in that According to the predicted pressure of hydrate formation corresponding to multiple groups of sub-training set data and the formation pressure of hydrate in multiple groups of sub-training set data, the training set prediction error of the hydrate formation risk warning model is obtained, including: According to the predicted pressure of hydrate formation corresponding to each group of sub-training set data and the formation pressure of hydrate in the sub-training set data, the sub-training set prediction error of the hydrate formation risk warning model is obtained; the prediction errors of multiple sub-training sets are weighted averaged to obtain the training set prediction error of the hydrate formation risk warning model; or Based on the predicted pressure of hydrate formation corresponding to each group of sub-training set data, the predicted pressure of hydrate formation corresponding to the training set data is obtained. According to the predicted pressure of hydrate formation corresponding to the training set data and the formation pressure of hydrate in the training set data, the training set prediction error of the hydrate formation risk warning model is obtained.

4. The method according to claim 1, characterized in that Also includes: Acquire validation set data obtained from a hydrate generation experiment, wherein the validation set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure; Inputting the gas quality components and hydrate formation temperature in the validation set data into the trained hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the validation set data; According to the predicted pressure of hydrate formation corresponding to the validation set data and the hydrate formation pressure in the validation set, the validation set prediction error of the hydrate formation risk warning model is determined; according to the validation set prediction error and the training set prediction error, the quality of the trained hydrate formation risk warning model is evaluated.

5. The method according to any one of claims 1 to 4, characterized in that: The hydrate formation risk warning model is a convolutional neural network model, which includes an input layer, a convolution layer, a batch normalization layer, an activation layer, a dropout layer, a fully connected layer, and a regression layer; The input layer serves as an input interface for gas quality components and hydrate formation temperature; The convolution layer is used to perform convolution operation and mapping on the gas quality components and hydrate formation temperature input by the input layer to obtain characteristic data of the gas quality components and temperature; The batch normalization layer is used to normalize the characteristic data of the temperament components and temperature to obtain the normalized characteristic data of the temperament components and temperature; The activation layer is used to perform nonlinear mapping on the normalized characteristic data of the temperament components and temperature to obtain a continuous probability density space of the characteristic data of the temperament components and temperature; The dropout layer is used to inactivate the parameters of the convolutional neural network model, and use the continuous probability density space as input data to obtain test parameters that meet the hydrate formation risk warning model; The fully connected layer is used to take the continuous probability density space as input data and the test parameter as a parameter, and output the predicted pressure of hydrate formation.

6. A method for early warning of hydrate formation risk in oil and gas pipelines, characterized in that: include: Obtain the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain pipeline temperature and pressure data; The gas quality components of the transported material in the oil and gas pipeline and the pipeline temperature are input into the trained oil and gas pipeline hydrate formation risk warning model, and the predicted pressure is output to obtain predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure; the oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method as described in any one of claims 1 to 5; Determine the temperature and pressure range for hydrate formation based on the predicted temperature and pressure data; if the pipeline temperature and pressure data is not within the temperature and pressure range for hydrate formation, no warning will be issued; If the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, an early warning will be issued.

7. The method according to claim 6, characterized in that If the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, the method further includes: determining an injection time and an injection position of the inhibitor, and injecting the inhibitor at the injection position in the pipeline at the injection time.

8. A training device for a hydrate formation risk warning model for oil and gas pipelines, characterized in that: include: A sample data acquisition module is used to acquire training set data obtained from a hydrate generation experiment, wherein the training set data includes gas quality components, hydrate generation temperature, and hydrate generation pressure; A model training module is used to group the training set data to obtain multiple groups of sub-training set data, and input the gas quality components and hydrate formation temperature in each group of sub-training set data into a pre-constructed hydrate formation risk warning model to obtain the predicted pressure of hydrate formation corresponding to the sub-training set data; according to the predicted pressure of hydrate formation corresponding to the multiple groups of sub-training set data and the hydrate formation pressure in the training set data, obtain the training set prediction error of the hydrate formation risk warning model; if the training set prediction error meets the preset error requirement, a trained hydrate formation risk warning model is obtained; If the prediction error of the training set does not meet the preset error requirement, the hydrate formation risk warning model is retrained using the training set data.

9. An oil and gas pipeline hydrate formation risk warning device, characterized in that: include: The temperature and pressure data acquisition module is used to acquire the pipeline temperature and pipeline pressure at a specified position in the oil and gas pipeline detected by the sensor to obtain the pipeline temperature and pressure data; A prediction module, for inputting the gas quality components of the transported material in the oil and gas pipeline and the pipeline temperature into a trained oil and gas pipeline hydrate formation risk warning model, outputting a predicted pressure, and obtaining predicted temperature and pressure data including the pipeline temperature and the corresponding predicted pressure; the oil and gas pipeline hydrate formation risk warning model is trained using the oil and gas pipeline hydrate formation risk warning model training method as described in any one of claims 1 to 5; An early warning module is used to determine the temperature and pressure range of hydrate formation based on the predicted temperature and pressure data; if the pipeline temperature and pressure data is not within the temperature and pressure range of hydrate formation, no early warning prompt is issued; If the pipeline temperature and pressure data are within the temperature and pressure range of hydrate formation, an early warning will be issued.

10. A computer storage medium, characterized in that: The computer storage medium stores computer executable instructions, which, when executed by the processor, implement the oil and gas pipeline hydrate formation risk warning model training method described in any one of claims 1-5 and / or the oil and gas pipeline hydrate formation risk warning method described in any one of claims 6-7.

11. An oil and gas pipeline hydrate formation risk warning device, characterized in that: include: 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 method for training a risk warning model for hydrate formation in oil and gas pipelines as described in any one of claims 1 to 5 and / or the method for warning risk of hydrate formation in oil and gas pipelines as described in any one of claims 6 to 7 are implemented.

Images