Drilling data transmission method based on self-encoder

Abstract

The invention relates to a drilling data transmission method based on an auto-encoder, and the method comprises the steps: obtaining a data compression rate set by a user, and the data compression rate is related to the data distribution type, data transmission quantity and data quality demands of a drilling well; according to the data compression rate, neural network training of the auto-encoder is carried out to obtain a trained auto-encoder, including an encoder and a decoder, and the encoder and the decoder are deployed; acquiring and preprocessing drilling data through a sensor arranged at the position of a drilling bit; compressing the preprocessed data by using the encoder to obtain low-dimensional feature data corresponding to the data compression rate; transmitting the low-dimensional feature data to the decoder; performing data decompression and reduction on the low-dimensional feature data by using the decoder to obtain decompressed high-dimensional data; and performing data analysis based on the high-dimensional data.

Description

Technical Field

[0001] The invention relates to the technical field of oil reservoir exploration and development, and in particular to a drilling data transmission method based on an autoencoder. Background Art

[0002] In the oil and gas production industry, the monitored drilling data is crucial. These data include not only key parameters such as the position, speed, pressure and temperature of the drill bit, but also important information such as formation characteristics and wellbore stability. Sensors are arranged at the location of the drill bit to collect data, and then the data is transmitted to the ground for data analysis.

[0003] As drilling depth continues to increase, real-time and effective data transmission faces severe challenges. Since the underground communication environment is limited by distance, signal attenuation and hardware performance, the data generated by the drill bit often cannot be transmitted to the ground control end at a sufficient speed, which directly affects the efficiency and safety of drilling operations.

[0004] Research has found that traditional drilling data transmission methods usually use lossless compression algorithms to reduce the amount of data transmitted in order to increase data transmission speed. The core technology of traditional compression algorithms is to build a dictionary that replaces the long character sequence in the original data with a short character sequence, and then uses the principle of information entropy coding to encode according to the probability distribution of the character sequence appearing in the data, thereby achieving efficient data compression.

[0005] However, traditional data compression and transmission methods have achieved certain technical effects in scenarios with large amounts of data duplication. However, when faced with randomly distributed data or data with less duplication, compression algorithms often have poor effects, and sometimes even cause the compressed data volume to increase instead of decrease, which is the so-called "data expansion" phenomenon. This situation seriously affects the efficiency and cost of data transmission. Summary of the invention

[0006] The present invention provides a drilling data transmission method based on an autoencoder, which can enable data analysts to flexibly set different data compression rates for different reservoirs and different drilling conditions, and achieve data decompression and recovery with a higher restoration rate, thereby effectively responding to the challenges of downhole communications and improving the overall efficiency and safety of drilling operations.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In a first aspect, the present application provides a drilling data transmission method based on an autoencoder, comprising:

[0009] Obtaining a data compression rate set by a user, wherein the data compression rate is related to the data distribution type, data transmission volume, and data quality requirements of the drilling;

[0010] According to the data compression rate, neural network training of the autoencoder is performed to obtain a trained autoencoder, including an encoder and a decoder, and the encoder and the decoder are deployed;

[0011] The drilling data is collected and preprocessed by sensors arranged at the drilling bit position;

[0012] Using the encoder, compressing the preprocessed data to obtain low-dimensional feature data corresponding to the data compression rate;

[0013] transmitting the low-dimensional feature data to the decoder;

[0014] Decompressing and restoring the low-dimensional feature data using the decoder to obtain decompressed high-dimensional data;

[0015] Based on the high-dimensional data, data analysis is performed.

[0016] In one implementation, the data compression rate increases as the data transmission volume increases, and decreases as the data quality requirement increases.

[0017] In one implementation, the network structure of the autoencoder includes the encoder and the decoder;

[0018] The network architecture of the encoder includes: a convolution layer for extracting feature maps; a pooling layer for further downsampling the feature maps; and a fully connected layer for obtaining the final low-dimensional features based on the downsampled feature maps;

[0019] The network architecture of the decoder includes: a deconvolution layer for feature reconstruction and a fully connected layer for reconstructing the original data. In one implementation, the training process of the autoencoder includes:

[0020] The training data is serialized and divided into training set and test set;

[0021] Construct a loss function based on mean square error;

[0022] Based on the serialized training set and loss function, batch normalization is used for training until the trained autoencoder processes the test set to meet the set requirements.

[0023] In one implementation, the acquisition of drilling data includes acquiring drill bit speed, depth, vibration, and temperature data through various types of sensors;

[0024] The preprocessing includes: outlier detection and correction, missing value interpolation and data segmentation processing.

[0025] In one implementation, in the data segmentation process, streaming data obtained by various types of sensors are segmented into data of fixed time window lengths for compression processing by an encoder.

[0026] In one implementation, the low-dimensional feature data is transmitted to the decoder based on optical fiber communication technology or wireless communication technology.

[0027] In one implementation, the data transmission process also includes using data encryption technology or data verification scheme.

[0028] In one implementation, the method further includes integrating and formatting the obtained high-dimensional data.

[0029] The present invention adopts the above technical solution, which has the following advantages:

[0030] 1. The autoencoder used in the present invention is a deep learning model that can learn effective representation of data. The encoder compresses the original data into low-dimensional features, and then the decoder restores these features to a shape close to the original. The format of the initial data.

[0031] 2. The present invention allows users to dynamically adjust the compression rate according to actual data transmission requirements and environmental constraints. Under the premise of ensuring data processing speed, users can choose a higher compression rate to reduce the amount of data transmission, or reduce the compression rate when higher data quality is required.

[0032] 3. Through the carefully designed autoencoder structure and training process, the present invention ensures that the accuracy of data recovery remains at a high level even under highly compressed conditions, thereby ensuring the availability and reliability of drilling data.

[0033] 4. The present invention can adapt to drilling data of different types and distributions, and can achieve effective compression regardless of whether the data has a large amount of repeated content or randomly distributed data, thereby avoiding the phenomenon of "data expansion".

[0034] In summary, the present invention can be widely used in the technical field of oil reservoir exploration and development. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of a method flow provided by an embodiment of the present invention;

[0036] Figure 2 is a schematic diagram of the overall structure of an autoencoder provided by an embodiment of the present invention;

[0037] Figure 3 is the network structure of the encoder;

[0038] Figure 4 It is the network structure of the decoder. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0040] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0041] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0042] In view of the defects and problems of existing technologies, such as Figure 1 , the present application provides a drilling data transmission method based on an autoencoder, comprising:

[0043] Obtaining a data compression rate set by a user, wherein the data compression rate is related to the data distribution type, data transmission volume, and data quality requirements of the drilling;

[0044] According to the data compression rate, neural network training of the autoencoder is performed to obtain a trained autoencoder, including an encoder and a decoder, and the encoder and the decoder are deployed;

[0045] The drilling data is collected and preprocessed by sensors arranged at the drilling bit position;

[0046] Using the encoder, compressing the preprocessed data to obtain low-dimensional feature data corresponding to the data compression rate;

[0047] transmitting the low-dimensional feature data to the decoder;

[0048] Decompressing and restoring the low-dimensional feature data using the decoder to obtain decompressed high-dimensional data;

[0049] Based on the high-dimensional data, data analysis is performed.

[0050] The above method is further described below in a more detailed embodiment based on the drawings provided in this application.

[0051] See also Figure 1 The drilling data (compression) transmission method based on the autoencoder can be implemented by a data processing system, which is composed of multiple modules, including a data collection module, a data segmentation module, an encoder module (compression module), a decoder module (decompression module) and a data regularization module. The processing steps of the system are as follows:

[0052] Step 1) Model preset: preset the compression rate according to user needs, train the compression model, and divide the trained model into a compression part and a decompression part.

[0053] Specifically, the structure of the autoencoder includes an encoder and a decoder.

[0054] Figure 3 The network architecture of the encoder is shown: feature extraction is performed through four convolutional layers:

[0055] The first layer uses a 1D convolution layer with a convolution kernel size of 11. It is processed by the activation function Tanh.

[0056] The second layer is also a 1D convolution layer with a convolution kernel size of 7. This layer is also processed by the activation function Tanh.

[0057] The third layer continues to use the 1D convolution layer with a convolution kernel size of 5. This layer is also processed by the activation function Tanh.

[0058] The fourth layer uses a 1D convolution layer again with a kernel size of 3. This layer is also processed by the activation function Tanh.

[0059] After that, a maximum pooling layer is used with a pooling window size of 3 and a stride of 2 to downsample the feature map. Finally, two fully connected layers are passed to obtain the final low-dimensional features of the original data.

[0060] Figure 4 The decoder network architecture is shown.

[0061] The decoder's network architecture: feature reconstruction is performed through four deconvolution layers, which is symmetrical to the encoder structure. Specifically:

[0062] The first layer uses a 1D deconvolution layer with a convolution kernel size of 3. It is processed by the activation function Tanh.

[0063] The second layer is also a 1D deconvolution layer with a kernel size of 5. This layer is also processed by the activation function Tanh.

[0064] The third layer continues to use the 1D deconvolution layer with a convolution kernel size of 7. This layer is also processed by the activation function Tanh.

[0065] The fourth layer uses a 1D deconvolution layer again with a kernel size of 11. This layer is also processed by the activation function Tanh.

[0066] Finally, the original data is reconstructed through two fully connected layers.

[0067] Furthermore, the training process of the autoencoder includes:

[0068] First, the original dataset is split into sequences of 256 data points in length. This is done to ensure that the continuity and contextual information of the input data are meaningful for the learning of the autoencoder. Next, these sequences are randomly shuffled to eliminate any possible sequential dependencies, and the dataset is divided into 70% as a training set and 30% as a test set. This division ratio can provide the model with enough data to learn, while also leaving enough data to evaluate the generalization ability of the model. Subsequently, the training set is used to train the autoencoder.

[0069] During the training process, the mean square error (MSE) is selected as the loss function because MSE can quantify the difference between the predicted value and the true value, providing a clear optimization target for the model. At the same time, the Adam optimizer is used to update the model parameters. The Adam optimizer combines the advantages of the momentum method, can accelerate the gradient descent process, and is suitable for processing large-scale data sets and parameter optimization problems by adaptively adjusting the learning rate. It is particularly suitable for processing non-stationary objectives and non-convex optimization problems. In addition, the hyperparameters of the Adam optimizer (such as learning rate and momentum) are relatively easy to adjust, making the training process more flexible.

[0070] During the network training process, batch normalization is introduced. Batch normalization can reduce internal covariate shift by normalizing the input of each small batch of data, stabilize the network input distribution during training, and thus accelerate the training speed of the neural network. At the same time, this operation also helps to improve the generalization ability of the model because it reduces the model's dependence on the distribution of specific input data.

[0071] After the training process is completed, the trained autoencoder is used to predict the test set. In the test phase, the loss function value on the test data set is calculated to evaluate the performance of the model on unseen data. By comparing the loss value on the test set with the loss value on the training set, it can be determined whether the model is overfitting, and the model structure and training strategy can be adjusted accordingly to achieve better generalization performance.

[0072] Step 2) Data collection and preprocessing: Through downhole sensors (such as pressure sensors, temperature sensors, acceleration These sensors record the operating parameters of the drill bit under different conditions, including speed, depth, vibration and temperature.

[0073] Step 3) Data segmentation: Split the streaming data into time windows of fixed length. For example, split the sensor data every 5 minutes to ensure that each data block contains enough information for subsequent analysis.

[0074] Step 4) Data compression: The processed data is passed into the compression model. The task of the compression model is to compress the high-dimensional input data into a low-dimensional representation, capturing the main features of the data while removing as much redundancy as possible.

[0075] Step 4) Data transmission: The integrated low-dimensional representation is transmitted to the receiving end through a network connection (such as wireless communication, optical fiber connection, etc.). In this process, the security and integrity of the transmission need to be considered, and encryption technology and data verification schemes may be used to ensure the accuracy of the data.

[0076] Step 5) Data reception and reconstruction: The received data is segmented as necessary so that it can be input into the decompression model as needed. The low-dimensional encoded data is input into the decompression model. The structure of the decompression model is usually symmetrical with the compression model and is responsible for restoring the low-dimensional representation to the original high-dimensional data.

[0077] Step 6) Data integration and application: Finally, the reconstructed raw data is integrated and formatted to meet specific subsequent work requirements, such as storage and analysis. Depending on the needs, the data is used for applications such as real-time monitoring, predictive analysis, fault detection, or optimization decision-making.

[0078] In summary, this application proposes an innovative drilling data compression method with adjustable compression rate based on autoencoders. The core of this method is to use autoencoders for efficient data compression. Autoencoders can extract low-dimensional feature distributions from complex data distributions, and this process fully utilizes the self-supervised learning ability of neural networks. Compared with traditional methods, the present invention has the following significant advantages:

[0079] Convenience: The compression method based on the autoencoder is simple to operate and easy to implement;

[0080] Efficiency: This method can process data quickly and improve compression efficiency;

[0081] Real-time: The present invention is applicable to real-time data streams and can meet dynamically changing data compression requirements.

[0082] The present invention solves the limitations of existing solutions on data requirements, can flexibly adjust the compression rate according to user needs, and achieve efficient data compression. At the same time, it will not be affected by the data distribution, and can compress the data at a stable compression rate. Through the transmission of compressed data and the data recovery at the receiving end, we can conduct in-depth analysis of the downhole situation in real time, provide strong support for decision makers, and make more favorable decisions. This compression method not only improves the efficiency of drilling data processing, but also makes the management of the oil and gas industry more efficient and scientific.

[0083] On the other hand, an embodiment of the present application provides a processing device corresponding to a drilling data compression method with adjustable compression rate based on an autoencoder. The processing device can be suitable for a client processing device, such as a mobile phone, a laptop computer, a tablet computer, a desktop computer, etc., to execute the method of the aforementioned embodiment.

[0084] The processing device includes a processor, a memory, a communication interface and a bus. The processor, the memory and the communication interface are connected through the bus to complete the communication between them. The memory stores the calculations that can be run on the processing device. Machine program, When the processing device runs the computer program, the data transmission method provided in the embodiment of the present application is executed.

[0085] In some implementations, the memory may be a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory.

[0086] In some other implementations, the processor may be a central processing unit (CPU), a digital signal processor (DSP), or other general-purpose processors of various types, which are not limited herein.

[0087] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0088] Those skilled in the art will understand that the structure of the above-mentioned computing device is only a partial structure related to the scheme of the present application, and does not constitute a limitation on the computing device to which the scheme of the present application is applied. The specific computing device may include more or fewer components, or combine certain components, or have a different arrangement of components.

[0089] An embodiment of the present application also provides a computer program product for tight gas well productivity prediction. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing the data transmission method described in the present embodiment.

[0090] Computer readable storage media can be tangible devices that hold and store instructions used by instruction execution devices. Computer readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any combination thereof.

[0091] The above embodiment provides a computer-readable storage medium, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0092] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0093] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce the following: refer to A product of a command device, the command device is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 indivual Functions specified in a box or boxes.

[0094] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 Steps for specifying functions in a box or boxes

[0095] In the several embodiments provided by the present invention, it should be understood that the disclosed method can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0096] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A drilling data transmission method based on an autoencoder, characterized in that: include: Obtaining a data compression rate set by a user, wherein the data compression rate is related to the data distribution type, data transmission volume, and data quality requirements of the drilling; According to the data compression rate, neural network training of the autoencoder is performed to obtain a trained autoencoder, including an encoder and a decoder, and the encoder and the decoder are deployed; The drilling data is collected and preprocessed by sensors arranged at the drilling bit position; Using the encoder, compressing the preprocessed data to obtain low-dimensional feature data corresponding to the data compression rate; transmitting the low-dimensional feature data to the decoder; Decompressing and restoring the low-dimensional feature data using the decoder to obtain decompressed high-dimensional data; Based on the high-dimensional data, data analysis is performed.

2. The drilling data transmission method based on the autoencoder according to claim 1, characterized in that: The data compression rate increases as the data transmission volume increases, and decreases as the data quality requirement increases.

3. The drilling data transmission method based on the autoencoder according to claim 1, characterized in that: The network structure of the autoencoder includes the encoder and the decoder; The network architecture of the encoder includes: a convolution layer for extracting feature maps; a pooling layer for further downsampling the feature maps; and a fully connected layer for obtaining the final low-dimensional features based on the downsampled feature maps; The network architecture of the decoder includes: a deconvolution layer for feature reconstruction, and a fully connected layer for reconstructing original data.

4. The drilling data transmission method based on the autoencoder according to claim 3 is characterized in that: The training process of the autoencoder includes: The training data is serialized and divided into training set and test set; Construct a loss function based on mean square error; Based on the serialized training set and loss function, batch normalization is used for training until the trained autoencoder processes the test set to meet the set requirements.

5. The drilling data transmission method based on the autoencoder according to claim 1, characterized in that: The acquisition of drilling data includes acquisition of drill bit speed, depth, vibration and temperature data through various types of sensors; The preprocessing includes: outlier detection and correction, missing value interpolation and data segmentation processing.

6. The drilling data transmission method based on the autoencoder according to claim 5, characterized in that: In the data segmentation process, the streaming data obtained by various types of sensors are segmented into data of fixed time window lengths for compression processing by the encoder.

7. The drilling data transmission method based on the autoencoder according to claim 1, characterized in that: The low-dimensional feature data is transmitted to the decoder based on optical fiber communication technology or wireless communication technology.

8. The drilling data transmission method based on the autoencoder according to claim 7, characterized in that: The data transmission process also includes the use of data encryption technology or data verification schemes.

9. The drilling data transmission method based on the autoencoder according to claim 1, characterized in that: The method also includes integrating and formatting the obtained high-dimensional data.

Images