Method and device for identifying different fracturing events in fracturing process, electronic equipment and storage medium

Abstract

The invention discloses a method and a device for identifying different fracturing events in a fracturing process, electronic equipment and a medium. The method comprises the following steps: collecting fracturing data in a fracturing process by using a data acquisition module; performing feature enhancement on the fracturing data based on a data feature enhancement module to obtain target data; pre-processing the target data according to a data pre-processing module to obtain pre-processed data; sampling the preprocessed data by using a data sampling module to obtain sampled data; and classifying and identifying the sampled data by using a neural network model, determining a classifier, and determining a fracturing event classification result of the fracturing data based on the classifier. According to the technical scheme, through the data classification neural network model and the semantic segmentation network model, the classifiers of the fracturing data and the different fracturing event relation models are obtained through training and testing respectively, the different fracturing events in the fracturing process are intelligently recognized, the time and cost of manual analysis are reduced, and the working efficiency is improved. And the identification accuracy and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas exploration and development, and in particular to a method, device, electronic equipment and medium for identifying different fracturing events during a fracturing process. Background Art

[0002] In oil and gas exploration and production, fracturing technology is a key means of increasing oil and gas production by enhancing the permeability of oil and gas reservoirs. Different fracturing events can have varying impacts on the reservoir. For example, formation breakdown is a crucial component of formation pressure monitoring; instantaneous pump stop pressure affects fracture morphology; and sand plugging can significantly harm fracturing production. Therefore, accurately identifying different fracturing events is crucial for oil and gas exploration and production.

[0003] At present, traditional recognition methods are mainly based on manual experience and rules.

[0004] This type of method is time-consuming, inefficient, and easily affected by human factors, and lacks intelligence and automation. Summary of the Invention

[0005] The present invention provides a method, device, electronic equipment and medium for identifying different fracturing events during a fracturing process, which can intelligently identify different fracturing events during a fracturing process, reduce the time and cost of manual analysis, and improve the accuracy and efficiency of identification.

[0006] According to one aspect of the present invention, a method for identifying different fracturing events during a fracturing process is provided, comprising:

[0007] Using the data acquisition module to collect fracturing data during the fracturing process;

[0008] Performing feature enhancement on the fracturing data based on a data feature enhancement module to obtain target data;

[0009] Preprocessing the target data according to the data preprocessing module to obtain preprocessed data;

[0010] Using a data sampling module to sample the preprocessed data to obtain sampled data;

[0011] A neural network model is used to classify and identify the sampled data, a classifier is determined, and a fracturing event classification result of the fracturing data is determined based on the classifier.

[0012] According to another aspect of the present invention, a device for identifying different fracturing events during a fracturing process is provided, comprising:

[0013] A fracturing data collection unit, configured to collect fracturing data during the fracturing process using a data acquisition module;

[0014] A data feature enhancement unit, configured to enhance the features of the fracturing data based on the data feature enhancement module to obtain target data;

[0015] A data preprocessing unit, configured to preprocess the target data according to the data preprocessing module to obtain preprocessed data;

[0016] A data sampling unit, configured to sample the preprocessed data using a data sampling module to obtain sampled data;

[0017] A data classification unit is configured to classify and identify the sampled data using a neural network model, determine a classifier, and determine a fracturing event classification result of the fracturing data based on the classifier.

[0018] According to another aspect of the present invention, an electronic device is provided, comprising:

[0019] at least one processor; and

[0020] a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program executable by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the method for identifying different fracturing events in a fracturing process according to any embodiment of the present invention.

[0022] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying different fracturing events in a fracturing process according to any embodiment of the present invention when executed.

[0023] The technical solution of the embodiment of the present invention uses a data classification neural network model and a semantic segmentation network model to train and test a classifier for the relationship model between fracturing data and different fracturing events, thereby realizing intelligent identification of different fracturing events during the fracturing process, reducing the time and cost of manual analysis, and improving the accuracy and efficiency of identification.

[0024] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 This is a flow chart of a method for identifying different fracturing events during a fracturing process according to the first embodiment of the present invention;

[0027] Figure 2 This is a timestamp data classification diagram of three types of fracturing events, namely, fracturing start / end, formation rupture, and instantaneous pump stop, provided in accordance with the first embodiment of the present invention;

[0028] Figure 3 is a pump pressure-pump displacement curve diagram provided according to the first embodiment of the present invention;

[0029] Figure 4 This is a pixel value classification diagram of four types of fracturing events, namely, ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging, provided in accordance with the first embodiment of the present invention;

[0030] Figure 5 2. It is a schematic diagram of a feature enhancement technology for fracturing image data provided according to the first embodiment of the present invention;

[0031] Figure 6 1 is a structural diagram of a bidirectional long-short-term neural network model classifier based on an attention mechanism according to the first embodiment of the present invention;

[0032] Figure 7 This is a structural diagram of a U-shaped neural network model classifier provided according to the first embodiment of the present invention;

[0033] Figure 8 This is a diagram showing the recognition effect of the Att-BiLSTM model classifier for fracturing start / end, rupture, and instantaneous pump stop fracturing events according to the first embodiment of the present invention;

[0034] Figure 9 This is a diagram showing the recognition effect of the U-net model classifier provided in Example 1 of the present invention on ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging fracturing events;

[0035] Figure 10 2 is a schematic structural diagram of a device for identifying different fracturing events during a fracturing process according to a second embodiment of the present invention;

[0036] Figure 11 3 is a schematic diagram of the structure of an electronic device for implementing the method for identifying different fracturing events in a fracturing process according to the third embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0038] It should be noted that the terms "original", "target", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices.

[0039] Example 1

[0040] Figure 1 A flowchart of a method for identifying different fracturing events during a fracturing process is provided for the first embodiment of the present invention. This embodiment is applicable to the identification of different fracturing events during a fracturing process. The method can be executed by a device for identifying different fracturing events during a fracturing process. The device can be implemented in the form of hardware and / or software. The device can be configured in a well site system. Figure 1 As shown, the method includes:

[0041] S110: Collecting fracturing data during the fracturing process using a data acquisition module.

[0042] In an embodiment of the present application, the data acquisition module is used to collect fracturing data during the fracturing process. The fracturing data represents the real-time data of the fracturing well during the fracturing process, which is transmitted and stored in the on-site well site database and can be obtained through the well site database or real-time exploration.

[0043] As an optional but non-limiting implementation, the data acquisition module is used to collect fracturing data during the fracturing process, including:

[0044] Use the data acquisition module to collect raw data during the fracturing process;

[0045] A fracturing text dataset is established based on the time, pump pressure, and pump displacement in the original data, and a fracturing image dataset is established based on the pump pressure and pump displacement graph in the original data.

[0046] In this embodiment of the present application, raw data from the fracturing process includes parameters such as time, pump pressure, pump displacement, sand concentration, and pumping speed. This raw data is used to create a fracturing text dataset and a fracturing image dataset. Specifically, the raw data of time, pump pressure, and pump displacement are selected to create the fracturing text dataset, and the raw data of pump pressure and pump displacement curves are selected to create the fracturing image dataset. The fracturing text dataset and the fracturing image dataset are then stored in the fracturing database.

[0047] By collecting fracturing data and establishing fracturing text datasets and fracturing image datasets respectively, a data source is provided for subsequent model classifier input.

[0048] As an optional but non-limiting implementation, after establishing a fracturing text dataset based on the time, pump pressure, and pump displacement in the raw data, and establishing a fracturing image dataset based on the pump pressure and pump displacement graph in the raw data, the method further includes:

[0049] The predetermined fracturing events are used to perform timestamp data classification on the fracturing text dataset, and the predetermined fracturing events are used to perform fracturing event marking on the fracturing image dataset.

[0050] In the present application, the fracturing process refers to the use of hydraulic forces to create fractures in oil and gas layers during oil or gas production. Seven types of fracturing events may occur during the fracturing process: fracturing start / end, bottom layer rupture, instantaneous pump stop, ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging. Ball delivery fracturing events refer to the process of injecting ball-like materials through the wellhead to seal the perforations, thereby creating multiple fractures.

[0051] In practical applications, the fracturing text dataset constructed with time, pump pressure, and displacement data is classified into timestamp data according to the actual occurrence time of the fracturing events "fracturing start / end, formation rupture, instantaneous pump stop" and their classification table. See Table 1 for the classification table of fracturing event categories and timestamp data. Figure 2 This is a timestamp data classification diagram for three types of fracturing events: "fracturing start / end, formation rupture, and instantaneous pump stop", showing the pump pressure and pump displacement data at different fracturing event timestamps.

[0052] Table 1 Summary of fracturing event categories and timestamp data

[0053]

[0054] In addition, the pump pressure and pump displacement data are used to establish a pump pressure-pump displacement curve as a fracturing image dataset. Specifically, according to the response characteristics of "ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, sand plugging" on the fracturing curve, different events are marked on the fracturing image dataset to obtain a fracturing event identification map. Figure 3 The fracturing image dataset is a pump pressure-pump displacement curve, see Figure 4 This is a pixel value classification map of the four types of fracturing events: "ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging". Different pixel values ​​represent different fracturing event categories.

[0055] S120: Enhance the characteristics of the fracturing data based on a data feature enhancement module to obtain target data.

[0056] In the embodiment of the present application, it should be noted that the fracturing data includes a fracturing text dataset and a fracturing image dataset, and the data enhancement module performs data feature enhancement on the fracturing text dataset and the fracturing image dataset respectively, so as to enhance the feature information hidden between the data and improve the accuracy and generalization ability of the classification model.

[0057] As an optional but non-limiting implementation, the fracturing data is enhanced based on the data feature enhancement module to obtain target data, including:

[0058] The features of the fracturing text dataset and the fracturing image dataset are enhanced based on a data feature enhancement module to obtain a target fracturing text dataset and a target fracturing image dataset.

[0059] Specifically, we enhanced the data features of the fracturing text dataset, strengthening the connection between pump pressure and pump displacement data at different time steps. This process captured pump pressure and pump displacement data at different time steps and segmented the time range into consecutive small time intervals, each of which is called a time step. See Table 2 for the input data parameters for the Att-BiLSTM model classifier.

[0060] Table 2 Input data parameters of Att-BiLSTM model classifier

[0061]

[0062] The data features of the fracturing text dataset are enhanced using the following formula:

[0063]

[0064]

[0065]

[0066] Among them, f′(x i) represents the derivative of the i-th time data, h(x i ) represents the mean of the i-th time data, k(x i ) represents the mean difference of the data at time i, j represents the sliding window size, and N represents the number of data points in a section of fracturing data.

[0067] Specifically, the data features of the fracturing image dataset are enhanced, including extracting the pump pressure-pump displacement curve from the fracturing database, and using image data feature enhancement to perform image cropping, distortion, flipping, and color gamut changes on the pump pressure-pump displacement curve to highlight the characteristics of different fracturing events and enhance image features. Figure 5 This is a schematic diagram of the feature enhancement technology for fracturing image data, including the effects of graphic scaling, distortion, and rotation, color gamut change, and local feature enhancement.

[0068] S130 , preprocessing the target data according to a data preprocessing module to obtain preprocessed data.

[0069] In the embodiment of the present application, data preprocessing can obtain a better relationship recognition model between input values ​​and output values, thereby improving the model recognition accuracy. The target data includes a target fracturing text data set and a target fracturing image data set after data feature enhancement, and the target data is preprocessed using a data preprocessing module, including data cleaning, data denoising and outlier processing, so that the range of the preprocessed data is between [-1,1], eliminating the dimension effect, and making the input data meet the input requirements of the deep learning algorithm. Among them, data cleaning can be achieved by cleaning data by missing values, smoothing noise data, identifying or deleting outliers and resolving inconsistencies, so as to achieve format standardization, abnormal data removal, error correction, and duplicate data removal. Data denoising methods include deletion and interpolation of missing values. Interpolation of missing values ​​means filling missing values ​​with a certain method, such as mean interpolation, median interpolation, mode interpolation, etc. Outlier processing includes deletion, replacement and non-processing methods.

[0070] As an optional but non-limiting implementation, preprocessing the target data according to the data preprocessing module to obtain preprocessed data includes:

[0071] The target fracturing text dataset is preprocessed using a predetermined mean filter and StandardScaler function to obtain a preprocessed fracturing text dataset, and the target fracturing image dataset is preprocessed using a predetermined pixel mean square error normalization algorithm to obtain a preprocessed fracturing image dataset.

[0072] Specifically, the preprocessing process of the target fracturing text dataset includes using a mean filter with a size of 35 to smooth various parameters of the target fracturing text dataset to reduce data noise. Then, the standardization technology StandardScaler function is used to eliminate the adverse effects caused by singular sample data in the target fracturing text dataset, prevent certain features in the data from having a significant impact on the model, and improve the stability and accuracy of the model. The mean filter function formula is as follows:

[0073]

[0074] Where n represents the filter size, x j represents the jth number of parameter x, x′ n+j-1 is the new value in parameter x, and N represents the number of data points in a segment of fracturing data.

[0075] The StandardScaler function is a data normalization method with the following formula:

[0076]

[0077] Where x is the original data, z is the normalized data, μ is the mean, and s is the standard deviation. After model training is complete, all data are reverse-transformed back to their original range.

[0078] Specifically, the preprocessing process of the target fracturing image dataset includes scaling the image pixel value range of the target fracturing image dataset to the range of [0, 1] through the pixel mean square error normalization algorithm to better adapt to model training and inference. The mean square error normalization function formula is as follows:

[0079]

[0080] Among them, x is the original pixel value, x norm is the normalized pixel value, μ is the mean of the image pixel values, and s is the standard deviation of the image pixel values. Mean-variance normalization can make the model more stable and is suitable for models and algorithms that are sensitive to the range of image pixel values.

[0081] S140: Utilize a data sampling module to sample the preprocessed data to obtain sampled data.

[0082] In the embodiment of the present application, the pre-processed data is sampled to save time and resources when processing large-scale fracturing data. In actual application, after the data pre-processing is completed, the sliding window sampling technology in the data adoption module is used to sample the sample points.

[0083] As an optional but non-limiting implementation, using a data sampling module to sample the preprocessed data to obtain sampled data includes:

[0084] The pre-processed fracturing text dataset and the pre-processed fracturing image dataset are sampled using a sliding window sampling technology in a data sampling module to obtain a sampled fracturing text dataset and a sampled fracturing image dataset.

[0085] Specifically, the sliding window sampling technology in the data sampling module is used to sample the preprocessed fracturing text dataset. The data is sampled using a sliding window of size 15 based on the fracturing construction time. The sliding window sampling uses the direction of increasing time features as the moving direction and 1 second as the moving step. The data in the window is sampled to form a sample matrix. At the same time, the category corresponding to the last row of the sample matrix is ​​used as the category of the current sample matrix, thereby forming a sample point.

[0086] Specifically, the sliding window sampling technology in the data sampling module is used to sample the pre-processed fracturing image data set, which is to input the pre-processed fracturing image data into the model. i And the real identification map y i One-to-one sampling is performed to form a sample point, where each pixel value of the input image in the sample point corresponds to each pixel category of the real recognition image.

[0087] S150: Use a neural network model to classify and identify the sampled data, determine a classifier, and determine a fracturing event classification result of the fracturing data based on the classifier.

[0088] In the embodiment of the present application, the sampled data includes a sampled fracturing text dataset and a sampled fracturing image dataset. The sampled fracturing text dataset and the sampled fracturing image dataset are divided into training set data and test set data respectively using a neural network model and input into the established classifier module for sample identification and classification. The neural network model is a complex network system formed by a large number of simple processing units (called neurons) that are widely interconnected. It is a mathematical model that simulates biological neural networks for information processing. Commonly used neural network models include perceptrons, convolutional neural networks, feedforward neural networks, etc.

[0089] As an optional but non-limiting implementation, a neural network model is used to classify and identify the sampled data to determine a classifier, including:

[0090] The sampled fracturing text dataset is trained based on a bidirectional long short-term neural network with an attention mechanism to determine a first classifier, and the sampled fracturing image dataset is trained based on a U-shaped neural network to determine a second classifier.

[0091] It is understood that the first classifier is based on a bidirectional long-short-term neural network model with an attention mechanism, which is used to train on a sampled fracturing text dataset and output fracturing event recognition results for "fracturing start / end, rupture, and instantaneous pump stop." The second classifier is based on a U-shaped neural network model, which is used to train on a sampled fracturing image dataset and output fracturing event recognition results for "ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging."

[0092] According to the sampled fracturing text dataset, a sample point is formed (including the model input x i and its true category y i ), and then form the Att-BiLSTM (Attention Mechanism Bi-directional LongShort-Term Memory, a bidirectional long-term short-term neural network based on the attention mechanism) model input-output dataset based on all sample points.

[0093] See also Figure 6 This figure shows the classifier structure of the Attention-Based Bidirectional Long Short-Term Memory (Att-BiLSTM) model. AttBiLSTM is a deep learning-based neural network model primarily used for classifying and predicting sequence data. It combines BiLSTM (Bidirectional Long Short-Term Memory) and the Attention Mechanism to effectively handle the long-term dependencies and complex features of time series data, improving classification accuracy. The Att-BiLSTM model classifier consists of the following components: ① Input layer: The original sequence data is input into the network; ② BiLSTM layer: This layer contains two LSTM layers, a forward LSTM and a backward LSTM, which can learn and extract the input data in both forward and backward directions, grasping the long-term dependencies in the time series data; ③ Attention layer: This layer learns the weight of each time step and performs a weighted average on the output of the LSTM layer, highlighting important time steps and suppressing irrelevant time steps, thereby improving classification accuracy; ④ Fully connected layer: The output of the attention layer is input into the fully connected layer for further feature abstraction and processing; ⑤ Output layer: This layer is a Sigmoid layer used to classify the target value. The Att-BiLSTM model classifier can automatically learn features and patterns in sequence data, avoiding tedious manual feature engineering and is applicable to various types of sequence data classification tasks.

[0094] In addition, a sample point (including the model input image xi and the real recognition image y) is formed according to the sampled fracturing image dataset. i), where each pixel value of the model input image in the sample point corresponds to each pixel category of the real recognition image, and then a U-net (U-shaped Network) model input-output dataset is formed based on all sample points.

[0095] See also Figure 7 This figure shows the architecture of a U-net model classifier. U-net is a deep learning-based neural network model primarily used for image segmentation tasks, effectively extracting image features and segmenting target regions. The U-net architecture can be divided into two parts: a downsampling path and an upsampling path. The U-net model classifier consists of the following components: ① The downsampling path primarily extracts image features and typically consists of multiple convolutional and pooling layers. Convolutional layers extract local image features, while pooling layers reduce the image size while preserving important feature information. In the downsampling path, each pooling layer is followed by a convolutional layer to preserve more detailed information. ② The upsampling path primarily restores image resolution and typically consists of multiple upsampling and convolutional layers. Upsampling layers magnify the image while preserving important feature information, while convolutional layers further extract features. In the upsampling path, each upsampling layer is followed by a convolutional layer to transfer feature information from the downsampling path. In the U-net structure, there is a skip connection between the downsampling path and the upsampling path, which is used to pass the feature information extracted in the downsampling path directly to the upsampling path to avoid information loss and blurring.

[0096] Then, the Att-BiLSTM model input-output dataset and the U-net model input-output dataset were randomly divided into training sets and test sets at a ratio of 9:1 for classifier training and testing.

[0097] The Att-BiLSTM model classifier identifies and classifies each timestamp in the sampled fracturing text dataset, while the U-net model classifier identifies and classifies each pixel value in the sampled fracturing image dataset. The Att-BiLSTM model classifier was then trained and evaluated using the evaluation metrics timestamp accuracy, the cross-entropy loss function, and the adaptive learning rate optimization algorithm (Adam). The U-net model classifier was trained and evaluated using the cross-entropy loss function and the adaptive learning rate optimization algorithm (Adam).

[0098] The evaluation indicator timestamp accuracy formula is:

[0099]

[0100] Among them, Accuracy represents the recognition accuracy of the test set, N t Represents the test set size, True Batch_size_i Represents the number of samples correctly identified in the i-th Batch_size, and m represents the number of Batch_size in the test set.

[0101] The evaluation indicator cross entropy loss function formula is:

[0102]

[0103] Among them, CE_loss represents the classification loss value of the sample, class is the true category number of sample x, j is the category number of the sample, and x j It is the model output value of sample x in category j.

[0104] In the text fracturing data recognition results, the trained optimal fracturing event classifier is used to obtain the classification results of each timestamp data in the fracturing data. Then, the classification conversion results and the fracturing data are visualized and output by the output module using the data conversion method and saved in the corresponding folder. (0→1) output is 80 (fracturing start), (1→2) output is 90 (fracture pressure), (2→3) output is 70 (fracturing end), (3→0) output is 60 (instantaneous pump stop pressure), and the output of other cases is 0, see Figure 8 This is the Att-BiLSTM model classifier's recognition effect diagram for the fracturing events of "fracturing start / end, formation rupture, and instantaneous pump stop", showing the time domain corresponding to each fracturing event.

[0105] In the recognition results of the fracturing image data, the recognition results are combined with the pump pressure-pump displacement curve, see Figure 9 This image shows the U-net model classifier's performance for identifying events such as "ball delivery, pre-acid pressure reduction, temporary plugging and diversion fracturing, and sand plugging." It displays the time domain corresponding to each fracturing event. The time-varying accuracy and error of the classifier training process are also visualized, providing an intuitive understanding of the classifier's performance and optimization process.

[0106] The present invention discloses a method, device, electronic device and medium for identifying different fracturing events in a fracturing process. The method comprises: using a data acquisition module to collect fracturing data in the fracturing process; based on a data feature enhancement module, performing feature enhancement on the fracturing data to obtain target data; preprocessing the target data according to a data preprocessing module to obtain preprocessed data; using a data sampling module to sample the preprocessed data to obtain sampled data; using a neural network model to classify and identify the sampled data, determine a classifier, and determine the fracturing event classification result of the fracturing data based on the classifier. The technical solution of the present invention, through a data classification neural network model and a semantic segmentation network model, respectively trains and tests a classifier of a relationship model between fracturing data and different fracturing events, respectively classifies and identifies different fracturing events for fracturing text data and fracturing image data, thereby realizing intelligent identification of different fracturing events in the fracturing process, reducing the time and cost of manual analysis, and improving the accuracy and efficiency of identification.

[0107] Example 2

[0108] Figure 10 This is a schematic diagram of the structure of a device for identifying different fracturing events during a fracturing process provided by the second embodiment of the present invention. Figure 10 As shown, the device includes:

[0109] The fracturing data collection unit 1010 is used to collect fracturing data during the fracturing process using the data acquisition module;

[0110] A data feature enhancement unit 1020 is configured to enhance the features of the fracturing data based on a data feature enhancement module to obtain target data;

[0111] A data preprocessing unit 1030 is configured to preprocess the target data according to a data preprocessing module to obtain preprocessed data;

[0112] The data sampling unit 1040 is configured to sample the pre-processed data using a data sampling module to obtain sampled data;

[0113] The data classification unit 1050 is configured to classify and identify the sampled data using a neural network model, determine a classifier, and determine a fracturing event classification result of the fracturing data based on the classifier.

[0114] Optionally, the fracturing data collection unit 1010 includes:

[0115] A raw data collection subunit, used to collect raw data during the fracturing process using a data acquisition module;

[0116] The data set establishment subunit is used to establish a fracturing text data set based on the time, pump pressure and pump displacement in the original data, and to establish a fracturing image data set based on the pump pressure and pump displacement graph in the original data.

[0117] Optionally, the dataset creation subunit includes:

[0118] The predetermined fracturing events are used to perform timestamp data classification on the fracturing text dataset, and the predetermined fracturing events are used to perform fracturing event marking on the fracturing image dataset.

[0119] Optionally, the data feature enhancement unit 1020 includes:

[0120] The data set enhancement subunit is used to enhance the features of the fracturing text data set and the fracturing image data set based on the data feature enhancement module to obtain a target fracturing text data set and a target fracturing image data set.

[0121] Optionally, the data preprocessing unit 1030 includes:

[0122] The dataset preprocessing subunit is configured to preprocess the target fracturing text dataset using a predetermined mean filter and StandardScaler function to obtain a preprocessed fracturing text dataset, and to preprocess the target fracturing image dataset using a predetermined pixel mean square error normalization algorithm to obtain a preprocessed fracturing image dataset.

[0123] Optionally, the data sampling unit 1040 includes:

[0124] The data set sampling subunit is used to sample the pre-processed fracturing text data set and the pre-processed fracturing image data set using the sliding window sampling technology in the data sampling module to obtain a sampled fracturing text data set and a sampled fracturing image data set.

[0125] Optionally, the data classification unit 1050 includes:

[0126] The classifier determination subunit is used to train the sampled fracturing text dataset based on a bidirectional long short-term neural network with an attention mechanism to determine a first classifier, and to train the sampled fracturing image dataset based on a U-shaped neural network to determine a second classifier.

[0127] The device for identifying different fracturing events in a fracturing process provided by an embodiment of the present invention can execute the method for identifying different fracturing events in a fracturing process provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0128] Example 3

[0129] Figure 11 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0130] like Figure 11 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0131] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0132] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the method for identifying different fracturing events during a fracturing process.

[0133] In some embodiments, the method for identifying different fracturing events during a fracturing process may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for identifying different fracturing events during a fracturing process described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the method for identifying different fracturing events during a fracturing process in any other appropriate manner (e.g., by means of firmware).

[0134] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0135] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0136] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0137] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0138] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0139] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0140] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0141] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for identifying different fracturing events during a fracturing process, characterized in that: include: Using the data acquisition module to collect fracturing data during the fracturing process; Performing feature enhancement on the fracturing data based on a data feature enhancement module to obtain target data; Preprocessing the target data according to the data preprocessing module to obtain preprocessed data; Using a data sampling module to sample the preprocessed data to obtain sampled data; A neural network model is used to classify and identify the sampled data, a classifier is determined, and a fracturing event classification result of the fracturing data is determined based on the classifier.

2. The method according to claim 1, characterized in that The data acquisition module is used to collect fracturing data during the fracturing process, including: Use the data acquisition module to collect raw data during the fracturing process; A fracturing text dataset is established based on the time, pump pressure, and pump displacement in the original data, and a fracturing image dataset is established based on the pump pressure and pump displacement graph in the original data.

3. The method according to claim 2, characterized in that After establishing a fracturing text dataset based on the time, pump pressure, and pump displacement in the original data, and establishing a fracturing image dataset based on a pump pressure and pump displacement graph in the original data, the method further includes: The predetermined fracturing events are used to perform timestamp data classification on the fracturing text dataset, and the predetermined fracturing events are used to perform fracturing event marking on the fracturing image dataset.

4. The method according to claim 3, characterized in that The fracturing data is enhanced based on the data feature enhancement module to obtain target data, including: The features of the fracturing text dataset and the fracturing image dataset are enhanced based on a data feature enhancement module to obtain a target fracturing text dataset and a target fracturing image dataset.

5. The method according to claim 4, characterized in that Preprocessing the target data according to the data preprocessing module to obtain preprocessed data includes: The target fracturing text dataset is preprocessed using a predetermined mean filter and StandardScaler function to obtain a preprocessed fracturing text dataset, and the target fracturing image dataset is preprocessed using a predetermined pixel mean square error normalization algorithm to obtain a preprocessed fracturing image dataset.

6. The method according to claim 5, characterized in that The pre-processed data is sampled by a data sampling module to obtain sampled data, including: The pre-processed fracturing text dataset and the pre-processed fracturing image dataset are sampled using a sliding window sampling technology in a data sampling module to obtain a sampled fracturing text dataset and a sampled fracturing image dataset.

7. The method according to claim 6, characterized in that Using a neural network model to classify and identify the sampled data and determine a classifier includes: The sampled fracturing text dataset is trained based on a bidirectional long short-term neural network with an attention mechanism to determine a first classifier, and the sampled fracturing image dataset is trained based on a U-shaped neural network to determine a second classifier.

8. A device for identifying different fracturing events during a fracturing process, characterized in that: include: A fracturing data collection unit, configured to collect fracturing data during the fracturing process using a data acquisition module; A data feature enhancement unit, configured to enhance the features of the fracturing data based on the data feature enhancement module to obtain target data; A data preprocessing unit, configured to preprocess the target data according to the data preprocessing module to obtain preprocessed data; A data sampling unit, configured to sample the preprocessed data using a data sampling module to obtain sampled data; A data classification unit is configured to classify and identify the sampled data using a neural network model, determine a classifier, and determine a fracturing event classification result of the fracturing data based on the classifier.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the method for identifying different fracturing events in a fracturing process according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying different fracturing events in a fracturing process according to any one of claims 1 to 7 when executed.

Images