Method and system for predicting fracturing wellhead pressure in real time

Abstract

The invention discloses a method and system for predicting fracturing wellhead pressure in real time. The method comprises the following steps of collecting historical construction data of a fractured wellhead, and performing data processing and normalization processing on the historical construction data of the fractured wellhead to obtain standard input data of the fractured wellhead; acquiring historical wellhead pressure characteristic curve data based on the standard input data of the fractured wellhead; constructing a fracturing wellhead pressure prediction model, and training, verifying and testing the fracturing wellhead pressure prediction model to obtain a tested fracturing wellhead pressure prediction model; and acquiring a real-time predicted value of the fracturing wellhead pressure and a corresponding working condition label by using the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model, and visualizing the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition label. According to the method, preparation can be made according to the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition label to deal with emergencies of different working conditions in advance.

Description

Technical Field

[0001] The present invention relates to the technical field of fracturing wellhead pressure prediction, and particularly relates to a method and system for real-time prediction of fracturing wellhead pressure. Background Art

[0002] In the existing methods for obtaining fracturing wellhead pressure, information of current equipment is usually collected, parameters that have occurred are checked, and then through parameter analysis, a working condition judgment result is obtained and corresponding measures are taken. This will lead to a delay in measures and untimely remedies, resulting in serious consequences. In addition, the existing methods for obtaining fracturing wellhead pressure do not utilize big data to predict the possible trend of construction curves, and thus cannot obtain experience from construction data, deduce, and feedback it into the construction process. Summary of the Invention

[0003] In view of the above deficiencies in the prior art, the present invention provides a method and system for real-time prediction of fracturing wellhead pressure.

[0004] To achieve the above invention purpose, the technical solution adopted by the present invention is as follows:

[0005] A method for real-time prediction of fracturing wellhead pressure includes the following steps:

[0006] S1. Collect historical construction data of the fracturing wellhead, and perform data processing and normalization processing on the historical construction data of the fracturing wellhead to obtain standard input data of the fracturing wellhead;

[0007] S2. Obtain a wellhead pressure curve and a wellhead displacement curve based on the standard input data of the fracturing wellhead, determine a working condition label according to the wellhead pressure curve and the wellhead displacement curve, and fuse the standard input data and the working condition label of the fracturing wellhead to the wellhead pressure curve to obtain historical wellhead pressure characteristic curve data;

[0008] S3. Construct a fracturing wellhead pressure prediction model, and use the historical wellhead pressure characteristic curve data to train, verify, and test the fracturing wellhead pressure prediction model to obtain a tested fracturing wellhead pressure prediction model;

[0009] S4. Collect real-time construction data of the fracturing wellhead, obtain real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead, use the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain real-time prediction values of the fracturing wellhead pressure and corresponding working condition labels, and visualize the real-time prediction values of the fracturing wellhead pressure and the corresponding working condition labels.

[0010] Further, in step S1, the historical construction data of the fracturing wellhead includes the stage type, wellhead pressure, wellhead displacement, total wellhead liquid volume, sand concentration, sand ratio, total sand transportation volume, and proppant type of the fracturing wellhead.

[0011] Further, data processing is performed on the historical construction data of the fracturing wellhead. The specific process is as follows: The historical construction data of the fracturing wellhead is divided into numerical type data and non-numerical type data. The stage type of the fracturing wellhead in the non-numerical type data is sorted by serial number to obtain the stage type value of the fracturing wellhead, and the proppant type of the fracturing wellhead in the non-numerical type data is numbered to obtain the proppant type value of the fracturing wellhead.

[0012] Further, in step S2, the working condition labels include pumping working condition, temporary plugging fracturing working condition, preflush acid fracturing working condition, sand plugging working condition, and equipment abnormal working condition.

[0013] Further, in step S2, the working condition labels are determined according to the wellhead pressure curve and the wellhead displacement curve. The specific process is as follows: According to the pressure difference on the wellhead pressure curve and the displacement difference on the displacement curve, it is judged whether a pressure gradient is formed at the perforation to determine the fracturing ball pumping working condition; According to the characteristic of first increasing and then decreasing of the wellhead pressure curve and the characteristic of remaining stable and unchanged of the displacement curve, the preflush acid fracturing working condition is determined; According to the characteristic of sudden rise of the wellhead pressure curve and the characteristic of downward trend of the displacement curve, the sand plugging working condition is determined; According to the characteristic of sudden drop of the wellhead pressure curve and the characteristic of first sudden drop and then flattening of the displacement curve, the equipment abnormal working condition is determined.

[0014] Further, in step S3, the fracturing wellhead pressure prediction model includes an input layer, a fully connected layer, a position encoding layer, an encoder, a decoder, and an output layer connected in sequence; The encoder includes a multi-head attention mechanism layer, a linear activation layer, and a feed-forward neural network module connected in sequence; The multi-head attention mechanism layer is used to map the wellhead pressure characteristic curve data output by the position encoding layer into query vectors, key vectors, and value vectors respectively through linear transformation, calculate attention scores according to the query vectors and key vectors, and perform normalization processing on the attention scores using Softmax normalization to generate attention weights.

[0015] Further, the data processing process of the position encoding layer is expressed as:

[0016]

[0017] Input pos =X pos +PE pos

[0018] Where: PE (pos,2i) is the encoding result of the wellhead pressure characteristic curve data in the even dimension, PE (pos,2i+1)It is the encoding result of the wellhead pressure characteristic curve data in the odd dimension. sin is the sine function, pos is the position of the current time step, i is the feature dimension index, d is the dimension of the embedding space, coS is the cosine function, Input pos It is the wellhead pressure characteristic curve data with position information, X pos It is the wellhead pressure characteristic curve data, PE pos It is the encoding result of the wellhead pressure characteristic curve data.

[0019] Furthermore, the data processing process of the multi-head attention mechanism layer is expressed as:

[0020]

[0021] Among them: Attention(Q, K, V) is the attention score obtained by the multi-head attention mechanism layer, Q is the query vector, K is the key vector, V is the value vector, softmax is the softmax activation function, d is the dimension of the embedding space, d k is the dimension of the key vector.

[0022] Furthermore, the number of attention heads in the multi-head attention mechanism layer is set to 4.

[0023] A system for real-time prediction of fracturing wellhead pressure applying the above method includes a data acquisition unit of the fracturing wellhead, a data processing unit of the fracturing wellhead, a pressure real-time prediction unit of the fracturing wellhead, and a visualization unit;

[0024] The data acquisition unit of the fracturing wellhead is used to collect the historical construction data and real-time construction data of the fracturing wellhead;

[0025] The data processing unit of the fracturing wellhead is used to perform data processing and normalization processing on the historical construction data of the fracturing wellhead, obtain the standard input data of the fracturing wellhead, obtain the wellhead pressure curve and wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition label according to the wellhead pressure curve and wellhead displacement curve, and fuse the standard input data and working condition label of the fracturing wellhead to the wellhead pressure curve to obtain the historical wellhead pressure characteristic curve data; obtain the real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead;

[0026] The pressure real-time prediction unit of the fracturing wellhead is used to construct a fracturing wellhead pressure prediction model, and use the historical wellhead pressure characteristic curve data to train, verify and test the fracturing wellhead pressure prediction model to obtain the tested fracturing wellhead pressure prediction model; use the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain the real-time prediction value of the fracturing wellhead pressure and the corresponding working condition label;

[0027] The visualization unit is used to visualize the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition labels.

[0028] The present invention has the following beneficial effects:

[0029] (1) By constructing a fracturing wellhead pressure prediction model and using the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model, the present invention can obtain the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition labels, and can prepare in advance to cope with sudden situations under different working conditions according to the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition labels;

[0030] (2) By obtaining the wellhead pressure curve and the wellhead displacement curve according to the standard input data of the fracturing wellhead, determining the working condition labels according to the wellhead pressure curve and the wellhead displacement curve, and fusing the standard input data of the fracturing wellhead and the working condition labels to the wellhead pressure curve, the present invention can obtain the historical wellhead pressure characteristic curve data, determine the data under different working conditions in the data processing stage, and input the wellhead pressure characteristic curve data with working condition labels into the fracturing wellhead pressure prediction model for separate training, enabling the fracturing wellhead pressure prediction model to identify the types of curves under different working conditions, which can improve the prediction accuracy of the fracturing wellhead pressure prediction model under different working conditions;

[0031] (3) The present invention proposes a system for real-time prediction of fracturing wellhead pressure, including a data acquisition unit of the fracturing wellhead, a data processing unit of the fracturing wellhead, a pressure real-time prediction unit of the fracturing wellhead, and a visualization unit; the data acquisition unit of the fracturing wellhead can be used to collect the historical construction data and the real-time construction data of the fracturing wellhead; the data processing unit of the fracturing wellhead can be used to perform data processing and normalization processing on the historical construction data of the fracturing wellhead, obtain the standard input data of the fracturing wellhead, obtain the wellhead pressure curve and the wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition labels according to the wellhead pressure curve and the wellhead displacement curve, and fuse the standard input data of the fracturing wellhead and the working condition labels to the wellhead pressure curve to obtain the historical wellhead pressure characteristic curve data; obtain the real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead; the pressure real-time prediction unit of the fracturing wellhead can be used to construct a fracturing wellhead pressure prediction model, train, verify, and test the fracturing wellhead pressure prediction model by using the historical wellhead pressure characteristic curve data, and obtain the tested fracturing wellhead pressure prediction model; obtain the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition labels by using the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model; the visualization unit can be used to visualize the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition labels. Description of the Drawings

[0032] Figure 1It is a schematic flow chart of a method for real-time prediction of the pressure at the fracturing wellhead;

[0033] Figure 2 It is a schematic diagram of the working condition label type;

[0034] Figure 3 It is a schematic structural diagram of a system for real-time prediction of the pressure at the fracturing wellhead. Specific embodiments

[0035] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of the present technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of the present technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0036] As Figure 1 shown, a method for real-time prediction of the pressure at the fracturing wellhead includes steps S1 - S4, specifically as follows:

[0037] S1. Collect the historical construction data of the fracturing wellhead, and perform data processing and normalization processing on the historical construction data of the fracturing wellhead to obtain the standard input data of the fracturing wellhead.

[0038] In an optional embodiment of the present invention, the present invention constructs a data acquisition unit for the fracturing wellhead, and collects the historical construction data of the fracturing wellhead through the data acquisition unit for the fracturing wellhead. The historical construction data of the fracturing wellhead includes the stage type, wellhead pressure, wellhead displacement, total wellhead liquid volume, sand concentration, sand ratio, total sand transportation volume, and proppant type of the fracturing wellhead.

[0039] The present invention constructs a data processing unit for the fracturing wellhead, and performs data processing and normalization processing on the historical construction data of the fracturing wellhead through the data processing unit for the fracturing wellhead to obtain the standard input data of the fracturing wellhead. The specific process of the present invention for performing data processing on the historical construction data of the fracturing wellhead is as follows: Divide the historical construction data of the fracturing wellhead into numerical type data and non - numerical type data, perform serial number sorting on the stage type of the fracturing wellhead in the non - numerical type data to obtain the stage type value of the fracturing wellhead, and number the proppant type of the fracturing wellhead in the non - numerical type data to obtain the proppant type value of the fracturing wellhead.

[0040] S2. Obtain the wellhead pressure curve and the wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition label according to the wellhead pressure curve and the wellhead displacement curve, and fuse the standard input data of the fracturing wellhead and the working condition label into the wellhead pressure curve to obtain the historical wellhead pressure characteristic curve data.

[0041] In an alternative embodiment of the present invention, the data processing unit of the fracturing wellhead constructed by the present invention can obtain the wellhead pressure curve and the wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition label according to the wellhead pressure curve and the wellhead displacement curve, and fuse the standard input data of the fracturing wellhead and the working condition label into the wellhead pressure curve to obtain the historical wellhead pressure characteristic curve data.

[0042] As Figure 2 shown, the working condition labels include pumping working condition, temporary plugging fracturing working condition, preflush acid fracturing working condition, sand plugging working condition and equipment abnormal working condition.

[0043] The present invention determines the working condition label according to the wellhead pressure curve and the wellhead displacement curve. The specific process is as follows: judging whether a pressure gradient is formed at the perforation according to the pressure difference on the wellhead pressure curve and the displacement difference on the displacement curve to determine the fracturing ball pumping working condition; determining the preflush acid fracturing working condition according to the characteristic of the wellhead pressure curve increasing first and then decreasing and the characteristic of the displacement curve remaining stable; determining the sand plugging working condition according to the characteristic of the wellhead pressure curve rising suddenly and the characteristic of the displacement curve decreasing; determining the equipment abnormal working condition according to the characteristic of the wellhead pressure curve dropping suddenly and the characteristic of the displacement curve dropping suddenly first and then flattening.

[0044] S3. Construct a fracturing wellhead pressure prediction model, and use the historical wellhead pressure characteristic curve data to train, verify and test the fracturing wellhead pressure prediction model to obtain the tested fracturing wellhead pressure prediction model.

[0045] In an alternative embodiment of the present invention, the present invention constructs a real-time pressure prediction unit for the fracturing wellhead. The real-time pressure prediction unit of the fracturing wellhead constructs a fracturing wellhead pressure prediction model, and uses the historical wellhead pressure characteristic curve data to train, verify and test the fracturing wellhead pressure prediction model to obtain the tested fracturing wellhead pressure prediction model.

[0046] The fracturing wellhead pressure prediction model includes an input layer, a fully connected layer, a position encoding layer, an encoder, a decoder and an output layer connected in sequence; the encoder includes a multi-head attention mechanism layer, a linear activation layer and a feed-forward neural network module connected in sequence; the multi-head attention mechanism layer is used to map the wellhead pressure characteristic curve data output by the position encoding layer into a query vector, a key vector and a value vector respectively through linear transformation, calculate the attention score according to the query vector and the key vector, and perform normalization processing on the attention score by using Softmax normalization to generate the attention weight. The decoder includes a masked self-attention layer, a linear activation layer, a multi-head attention mechanism layer and a feed-forward neural network module connected in sequence. The masked self-attention layer is used to predict the predicted value of the fracturing wellhead pressure of the next time series according to the predicted sequence value of the fracturing wellhead pressure generated currently.

[0047] In the present invention, the wellhead pressure characteristic curve data is transformed into a space with a dimension of 64 through a fully connected layer by means of a linear neuron, so as to retain the key features in the wellhead pressure characteristic curve data that affect the wellhead pressure prediction.

[0048] In the present invention, the position information of the wellhead pressure characteristic curve data is introduced through a position encoding layer. Specifically, torch.sin is used in the even dimensions and torch.cos is used in the odd dimensions to perform position encoding for each time step (0, 1, 2,...), calculate the position information of each time step, and add the obtained position encoding data to the time step features in the input data, and then add it to the wellhead pressure characteristic curve data.

[0049] The data processing process of the position encoding layer is expressed as:

[0050]

[0051] Input pos =X pos +PE pos

[0052] where: PE (pos,2i) is the encoding result of the wellhead pressure characteristic curve data in the even dimensions, PE (pos,2i+1) is the encoding result of the wellhead pressure characteristic curve data in the odd dimensions, sin is the sine function, pos is the position of the current time step, i is the feature dimension index, d is the dimension of the embedding space, coS is the cosine function, Input pos is the wellhead pressure characteristic curve data with position information, X pos is the wellhead pressure characteristic curve data, and PE pos is the encoding result of the wellhead pressure characteristic curve data.

[0053] Each input feature in the fracturing wellhead pressure prediction model is embedded into a vector with a dimension of 128. Therefore, the dimension of the embedding space is 128. The position index represents each element in the position encoding vector, which is 0 to 127 in this model.

[0054] The data processing process of the multi-head attention mechanism layer is expressed as:

[0055]

[0056] where: Attention(Q, K, V) is the attention score obtained by the multi-head attention mechanism layer, Q is the query vector, K is the key vector, V is the value vector, softmax is the softmax activation function, d is the dimension of the embedding space, and d k is the dimension of the key vector.

[0057] The number of attention heads in the multi - head attention mechanism layer is set to 4. By setting the number of attention heads in the multi - head attention mechanism layer to 4, the present invention can capture different information in the input sequence from multiple different perspectives and can reduce the computational complexity while ensuring the accuracy of the model.

[0058] The present invention can capture the interdependent relationship between each time point in the input wellhead pressure characteristic curve data through the multi - head attention mechanism layer, and predict the upcoming fluctuation trend (xt + 1, xt + 2, xt + 3..., xt + 30) by learning the interdependent relationship between historical data points (x1, x2, x3,..., xt).

[0059] The feed - forward neural network module consists of two linear transformation layers and a ReLU activation function. The present invention strengthens the non - linear relationship between the processing progress and the operating parameters by using the feed - forward neural network module, thereby improving the prediction accuracy. Specifically, after the output wellhead pressure characteristic curve data is transformed through the linear activation layer, it is processed by the feed - forward network for the attention output and mapped to the target space to obtain the output data. Then the output data is restored through the decoder to obtain the readable feature values, and the wellhead fracturing prediction value and the working condition label are taken as the final output results.

[0060] The present invention uses the historical wellhead pressure characteristic curve data to train, validate and test the fracturing wellhead pressure prediction model. In this process, the loss functions are selected as the nn.SmoothL1Loss function and the CurvatureLoss, and the optimizer uses AdamW.

[0061] S4. Collect the real - time construction data of the fracturing wellhead, obtain the real - time wellhead pressure characteristic curve data according to the real - time construction data of the fracturing wellhead, use the real - time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain the real - time prediction value of the fracturing wellhead pressure and the corresponding working condition label, and visualize the real - time prediction value of the fracturing wellhead pressure and the corresponding working condition label.

[0062] In an alternative embodiment of the present invention, the present invention collects the real - time construction data of the fracturing wellhead through the data acquisition unit of the fracturing wellhead, then obtains the real - time wellhead pressure characteristic curve data by using the data processing unit of the fracturing wellhead, finally obtains the real - time prediction value of the fracturing wellhead pressure and the corresponding working condition label by using the real - time pressure prediction unit of the fracturing wellhead, and visualizes the real - time prediction value of the fracturing wellhead pressure and the corresponding working condition label by using the visualization unit.

[0063] As Figure 3As shown, a system for real-time prediction of the pressure at a fracturing wellhead applying the above method includes a data acquisition unit, a data processing unit, a real-time pressure prediction unit, and a visualization unit for the fracturing wellhead.

[0064] In an alternative embodiment of the present invention, the data acquisition unit for the fracturing wellhead is configured to acquire historical construction data and real-time construction data of the fracturing wellhead.

[0065] In an alternative embodiment of the present invention, the data processing unit for the fracturing wellhead is configured to perform data processing and normalization on the historical construction data of the fracturing wellhead to obtain standard input data for the fracturing wellhead, obtain a wellhead pressure curve and a wellhead displacement curve based on the standard input data for the fracturing wellhead, determine a working condition label according to the wellhead pressure curve and the wellhead displacement curve, and fuse the standard input data and the working condition label for the fracturing wellhead to the wellhead pressure curve to obtain historical wellhead pressure characteristic curve data; obtain real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead.

[0066] In an alternative embodiment of the present invention, the real-time pressure prediction unit for the fracturing wellhead is configured to construct a fracturing wellhead pressure prediction model, and use the historical wellhead pressure characteristic curve data to train, validate, and test the fracturing wellhead pressure prediction model to obtain a tested fracturing wellhead pressure prediction model; use the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain real-time predicted values of the pressure at the fracturing wellhead and corresponding working condition labels.

[0067] In an alternative embodiment of the present invention, the visualization unit is configured to visualize the real-time predicted values of the pressure at the fracturing wellhead and corresponding working condition labels.

[0068] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0069] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in one or more of the blocks or blocks.

[0070] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in one or more of the blocks or blocks.

[0071] Specific embodiments are used in the present invention to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only for helping to understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

[0072] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principles of the present invention and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A method for real-time prediction of fracturing wellhead pressure, characterized in that: The following steps are involved: S1. Collect historical construction data of the fracturing wellhead, and perform data processing and normalization on the historical construction data of the fracturing wellhead to obtain standard input data of the fracturing wellhead; S2. Obtain a wellhead pressure curve and a wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition label according to the wellhead pressure curve and the wellhead displacement curve, and merge the standard input data and working condition label of the fracturing wellhead into the wellhead pressure curve to obtain historical wellhead pressure characteristic curve data; S3, constructing a fracturing wellhead pressure prediction model, and using historical wellhead pressure characteristic curve data to train, verify and test the fracturing wellhead pressure prediction model to obtain a tested fracturing wellhead pressure prediction model; S4. Collect real-time construction data of the fracturing wellhead, obtain real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead, use the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain the real-time prediction value of the fracturing wellhead pressure and the corresponding working condition label, and visualize the real-time prediction value of the fracturing wellhead pressure and the corresponding working condition label.

2. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: In step S1, the historical construction data of the fracturing wellhead includes the stage type of the fracturing wellhead, wellhead pressure, wellhead displacement, total wellhead liquid volume, sand concentration, sand ratio, total sand delivery and proppant type.

3. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: The historical construction data of the fracturing wellhead is processed, and the specific process is: the historical construction data of the fracturing wellhead is divided into digital type data and non-digital type data, the stage types of the fracturing wellhead in the non-digital type data are sorted by serial number to obtain the stage type values ​​of the fracturing wellhead, and the proppant types of the fracturing wellhead in the non-digital type data are numbered to obtain the proppant type values ​​of the fracturing wellhead.

4. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: In step S2, the operating condition labels include pumping conditions, temporary plugging and fracturing conditions, pre-acid fracturing conditions, sand plugging conditions and equipment abnormal conditions.

5. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: In step S2, the operating condition label is determined according to the wellhead pressure curve and the wellhead displacement curve. The specific process is as follows: according to the pressure difference on the wellhead pressure curve and the displacement difference on the displacement curve, it is judged whether a pressure ladder is formed at the hole to determine the fracturing ball pumping condition; according to the first increase and then decrease characteristics of the wellhead pressure curve and the stable and unchanged characteristics of the displacement curve, the pre-acid fracturing condition is determined; according to the rising and sudden rise characteristics of the wellhead pressure curve and the downward trend characteristics of the displacement curve, the sand plugging condition is determined; according to the sudden drop characteristics of the wellhead pressure curve and the first sudden drop and then gentle characteristics of the displacement curve, the abnormal equipment condition is determined.

6. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: In step S3, the fracturing wellhead pressure prediction model includes an input layer, a fully connected layer, a position encoding layer, an encoder, a decoder and an output layer connected in sequence; the encoder includes a multi-head attention mechanism layer, a linear activation layer and a feedforward neural network module connected in sequence; the multi-head attention mechanism layer is used to map the wellhead pressure characteristic curve data output by the position encoding layer into a query vector, a key vector and a value vector respectively through a linear transformation, calculate the attention score according to the query vector and the key vector, and use Softmax normalization to normalize the attention score to generate an attention weight.

7. The method for real-time prediction of fracturing wellhead pressure according to claim 6, characterized in that: The data processing process of the position encoding layer is expressed as: Input pos =X pos +PE pos Among them: PE (pos,2i) is the encoding result of the wellhead pressure characteristic curve data in even dimensions, PE (pos,2i+1) is the encoding result of the wellhead pressure characteristic curve data in odd dimensions, sin is the sine function, pos is the current time step position, i is the feature dimension index, d is the dimension of the embedding space, cos is the cosine function, Input pos is the wellhead pressure characteristic curve data with position information, X pos is the wellhead pressure characteristic curve data, PE pos It is the encoding result of the wellhead pressure characteristic curve data.

8. The method for real-time prediction of fracturing wellhead pressure according to claim 6, characterized in that: The data processing process of the multi-head attention mechanism layer is expressed as: Where: Attention(Q,K,V) is the attention score obtained by the multi-head attention mechanism layer, Q is the query vector, K is the key vector, V is the value vector, softmax is the softmax activation function, d is the dimension of the embedding space, and d k is the dimension of the key vector.

9. The method for real-time prediction of fracturing wellhead pressure according to claim 1, characterized in that: The number of attention heads in the multi-head attention mechanism layer is set to 4.

10. A system for real-time prediction of fracturing wellhead pressure using the method described in any one of claims 1 to 9, characterized in that: It includes a data acquisition unit at the fracturing wellhead, a data processing unit at the fracturing wellhead, a real-time pressure prediction unit at the fracturing wellhead, and a visualization unit; The data acquisition unit of the fracturing wellhead is used to collect the historical construction data of the fracturing wellhead and the real-time construction data of the fracturing wellhead; The data processing unit of the fracturing wellhead is used to process and normalize the historical construction data of the fracturing wellhead, obtain the standard input data of the fracturing wellhead, obtain the wellhead pressure curve and the wellhead displacement curve based on the standard input data of the fracturing wellhead, determine the working condition label according to the wellhead pressure curve and the wellhead displacement curve, and integrate the standard input data and working condition label of the fracturing wellhead into the wellhead pressure curve to obtain the historical wellhead pressure characteristic curve data; obtain the real-time wellhead pressure characteristic curve data according to the real-time construction data of the fracturing wellhead; The real-time prediction unit for the pressure at the fracturing wellhead is used to construct a fracturing wellhead pressure prediction model, and use the historical wellhead pressure characteristic curve data to train, verify and test the fracturing wellhead pressure prediction model to obtain the tested fracturing wellhead pressure prediction model; use the real-time wellhead pressure characteristic curve data and the tested fracturing wellhead pressure prediction model to obtain the real-time prediction value of the fracturing wellhead pressure and the corresponding working condition label; The visualization unit is used to visualize the real-time predicted value of the fracturing wellhead pressure and the corresponding working condition label.

Images