Well cementation slurry density acquisition method and device based on vibration noise removal
By removing the influence of vibration noise through a neural network regression model, the problem of poor accuracy in mud density detection data was solved, achieving high-precision mud density detection and cementing quality control.
Patent Information
- Application Number
- CN202411133828.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2026-03-03
AI Technical Summary
In existing technologies, the density data collected by mud density sensors has large errors due to irregular vibration noise, resulting in poor accuracy of cementing detection data and difficulty in effectively removing the influence of irregular vibration noise.
A machine learning-based neural network regression model is used to train and optimize the vibration noise removal model. The model generates mud density values by using the directional components of vibration noise and preprocessed mud density monitoring data. The model drift is corrected by monitoring differences to improve detection accuracy.
It improves the accuracy of mud density detection and cementing quality, reduces manual labor input, lowers labor intensity, and increases the efficiency of the cement mixing process.
Smart Images
Figure CN121593764A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of oil and gas drilling and cementing technology, specifically relating to a cementing mud density acquisition method based on vibration and noise removal, a cementing mud density acquisition device based on vibration and noise removal, a cementing mud density control system based on vibration and noise removal, a computer device, and a machine-readable storage medium. Background Technology
[0002] During oil drilling, technicians need to constantly monitor the accurate density data of the drilling mud and adjust it according to the requirements of the cementing process to maintain balance inside and outside the wellbore at different stages, thus ensuring the scientific execution of cementing operations. Research has revealed that a major factor contributing to errors and distortions in the density data collected by mud density sensors is irregular vibration noise present during the acquisition process. Traditional filtering and noise reduction methods often perform poorly in handling complex noise, especially irregular vibration noise.
[0003] Due to the complex and variable cementing operation environment, and the highly random nature of irregular vibration and noise, how to remove the influence of vibration and noise on the detection signal of the mud density sensor and thereby obtain accurate and reliable mud density data has always been a difficult technical problem to overcome in the field of cementing mud density detection, and a bottleneck that has hindered the development of cementing detection technology. Summary of the Invention
[0004] The purpose of this invention is to provide a cementing mud density acquisition method based on vibration and noise removal, a cementing mud density acquisition device based on vibration and noise removal, a cementing mud density control system based on vibration and noise removal, a computer device, and a machine-readable storage medium, in order to overcome the technical problem of poor accuracy of cementing mud density detection data caused by the lack of feasible and effective irregular vibration and noise removal schemes in the prior art.
[0005] To achieve the above objectives, a first aspect of the present invention provides a method for acquiring cementing mud density based on vibration noise removal, comprising:
[0006] Collect vibration and noise data and mud density monitoring data of the current section of the target wellbore environment;
[0007] Preprocess the mud density monitoring data of the current layer;
[0008] The pre-processed mud density monitoring data and vibration noise factor are input into a pre-built vibration noise removal model to generate the mud density value of the current layer.
[0009] The vibration noise factor refers to the directional components of the vibration noise, and the vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
[0010] Optionally, the vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
[0011] Optional preprocessing includes filtering, denoising, and normalization.
[0012] Optionally, the regression model is a neural network regression model.
[0013] Optionally, when initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to the input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
[0014] Optionally, the hidden layer connected to the input layer in the neural network regression model includes four nodes, wherein:
[0015] The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise.
[0016] The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise.
[0017] The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise.
[0018] The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
[0019] Optionally, the method further includes:
[0020] The difference between the mud density value generated by the vibration and noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment is monitored. If the difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration and noise removal model are updated. Otherwise, the mud density value generated by the vibration and noise removal model is used as the mud density value of the current layer.
[0021] A second aspect of the present invention provides a cementing mud density acquisition device based on vibration noise removal, the device comprising:
[0022] The acquisition module is used to collect vibration and noise data in the current section of the target wellbore, as well as mud density monitoring data for the current section.
[0023] The preprocessing module is used to preprocess the mud density monitoring data of the current layer.
[0024] The prediction module is used to input the pre-processed mud density monitoring data and vibration noise factor into a pre-built vibration noise removal model to generate the mud density value of the current layer.
[0025] The vibration noise factor refers to the directional components of the vibration noise, and the vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
[0026] Optionally, the vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
[0027] Optional preprocessing includes filtering, denoising, and normalization.
[0028] Optionally, the regression model is a neural network regression model.
[0029] Optionally, when initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to the input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
[0030] Optionally, the hidden layer connected to the input layer in the neural network regression model includes four nodes, wherein:
[0031] The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise.
[0032] The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise.
[0033] The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise.
[0034] The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
[0035] Optionally, the device further includes a monitoring and correction module, which is used to monitor the degree of difference between the mud density value generated by the vibration and noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment. If the degree of difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration and noise removal model are updated. Otherwise, the mud density value generated by the vibration and noise removal model is used as the mud density value of the current layer.
[0036] A third aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the cementing mud density acquisition method based on vibration noise removal as described in the first aspect of the present invention.
[0037] A fourth aspect of the present invention provides a cementing mud density control system based on vibration and noise removal. The system includes the computer device and control device described in the third aspect of the present invention. The control device is used to receive the mud density value of the current layer output by the computer device, and to determine whether the difference between the mud density value of the current layer output by the computer device and the mud density design value for cementing operations in the current layer exceeds a second threshold. If so, a control command is generated based on the difference value to change the mud density in the current layer.
[0038] A fifth aspect of the present invention provides a machine-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the cementing mud density acquisition method based on vibration noise removal as described in the first aspect of the present invention.
[0039] The above technical solution abandons vibration noise removal methods based on hardware circuits or software filtering. Instead, it utilizes the advantages of machine learning in linear and nonlinear regression to construct a vibration noise removal model after training and parameter optimization of the initial regression model. The functional expression of the vibration noise removal model is the fitting relationship between the input parameters and the output parameters. Thus, the mud density value of the current layer is predicted by using preprocessed mud density monitoring data with vibration noise interference. Here, the input parameters refer to the vibration noise factor input to the vibration noise removal model and the preprocessed mud density monitoring data, and the output parameter refers to the predicted mud density value.
[0040] Based on this, the impact of irregular vibration noise on mud density acquisition is eliminated by removing the vibration noise model, which improves the accuracy of mud density detection and cementing quality, and also has high efficiency. As a result, the efficiency of the cement mixing process is also improved. At the same time, the high-precision automatic detection of mud density value based on the vibration noise removal model also reduces the manual input in the cement mixing process, reduces labor intensity, and saves labor costs.
[0041] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description
[0042] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:
[0043] Figure 1 A schematic diagram illustrates a first flowchart of a cementing mud density acquisition method based on vibration noise removal according to an embodiment of the present invention;
[0044] Figure 2 A schematic diagram illustrates a second flowchart of a cementing mud density acquisition method based on vibration noise removal according to an embodiment of the present invention;
[0045] Figure 3 This diagram illustrates a network architecture of a neural network regression model in a specific application example.
[0046] Figure 4 A schematic diagram illustrating the components of a cementing mud density acquisition device based on vibration noise removal according to an embodiment of the present invention is shown.
[0047] Figure 5 A schematic block diagram of a computer device according to an embodiment of the present invention is shown. Detailed Implementation
[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustrating and explaining the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0049] It should be noted that if the embodiments of the present invention involve descriptions such as "first" and "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" and "second" may explicitly or implicitly include at least one of those features. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.
[0050] Example 1
[0051] See Figure 1 The cementing mud density acquisition method based on vibration noise removal provided in this embodiment of the invention includes the following implementation steps:
[0052] Step S100: Collect vibration and noise data and mud density monitoring data of the current section of the target wellbore environment.
[0053] As is known, in a typical embodiment, mud density monitoring data of the cementing section is collected by a mud sensor, and vibration and noise can be collected by a gyroscope or a pump-flush sensor, etc.
[0054] Step S200: Preprocess the mud density monitoring data of the current layer.
[0055] For example, in one specific embodiment, the preprocessing of the mud density monitoring data for the current layer includes filtering, denoising, and normalization. It should be understood that filtering and denoising are used to remove interference other than vibration noise, and normalization is a scaling process performed on the mud density monitoring data to scale it to a reasonable range, facilitating better training and parameter optimization, so as to use the mud density monitoring data as input data for the vibration noise removal model. Therefore, preprocessing the mud density monitoring data improves the training effect and generalization ability of the vibration noise removal model, thereby improving the accuracy of cementing mud density acquisition. Furthermore, filtering, denoising, and normalization can be performed using the methods described in the general embodiments, and this embodiment will not describe these aspects in detail.
[0056] Step S300: The vibration noise factor and preprocessed mud density monitoring data are input into a pre-constructed vibration noise removal model to generate the mud density value of the current layer. The vibration noise removal model is obtained by training and optimizing the parameters of an initial regression model, and the vibration noise factor represents the directional components of the vibration noise.
[0057] For example, the directional components of vibration noise can be determined by performing feature analysis on the vibration noise within the cementing environment of the target wellbore. For instance, three-dimensional data of the vibration noise within the cementing environment can be acquired using a gyroscope or pump sensor, and the directional components in each dimension can be obtained through analysis of the three-dimensional data. These directional components obtained through feature analysis can be XYZ coordinate components or spherical coordinate components, etc.
[0058] It is important to understand that the regression model can be selected from supervised learning-based machine learning models, such as neural network regression models and support vector regression models. Neural network regression models can be RBF neural network models, BP neural network models, etc., which typically include an input layer, hidden layers, and an output layer. This embodiment does not limit the specific network type of the neural network regression model.
[0059] For example, in one specific embodiment, the vibration noise factor includes the X-axis component, Y-axis component and Z-axis component of the vibration noise. Accordingly, the X-axis component of the vibration noise can be called the X-axis vibration noise factor, the Y-axis component of the vibration noise can be called the Y-axis vibration noise factor, and the Z-axis component of the vibration noise can be called the Z-axis vibration noise factor.
[0060] It should be understood that the above embodiments are not the only limitations on the vibration noise factor. For example, in a comparative embodiment, the vibration noise factor includes the radial component, polar component, and azimuth component of the vibration noise. Compared to the radial, polar, and azimuth components of the vibration noise, using the X-axis, Y-axis, and Z-axis components of the vibration noise as input parameters for the regression model makes the regression model easier to calculate, thus reducing the requirements for the computing power and storage space of the terminal device when deploying the regression model.
[0061] For example, in one specific embodiment, the neural network regression model may have multiple hidden layers. Compared to a single hidden layer, increasing the number of hidden layers allows the neural network regression model to capture more complex and hierarchical features in the input parameters, thereby improving the model's accuracy.
[0062] For example, in one specific embodiment, during training and parameter optimization, the weight hyperparameters involved in the neural network regression model are initialized by considering the propagation law of vibration and noise in the cementing operation. In supervised training of the neural network regression model, considering the propagation law of vibration and noise in the cementing operation environment allows for more accurate weight hyperparameters, thereby improving the training efficiency of the neural network regression model.
[0063] For example, in one specific embodiment, when initializing the weight hyperparameters, for any node in the first hidden layer of the neural network regression model, in the weight set composed of the weights between that node and each node in the input layer, at least one weight value is greater than the remaining weight values, that is, at least one dimension of the input parameter is the dominant factor of that hidden layer node.
[0064] For example, in one specific embodiment, the first hidden layer of the neural network regression model includes four nodes, and the feature extraction function relationship of each node in the first hidden layer has the following characteristics:
[0065] 1) The weight set is composed of the weights between the first node of the first hidden layer and each node of the input layer. The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis vibration noise factor.
[0066] 2) The weight set is composed of the weights between the second node of the first hidden layer and each node of the input layer. The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis vibration noise factor.
[0067] 3) The weight set is composed of the weights between the third node of the first hidden layer and each node of the input layer. The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis vibration noise factor.
[0068] 4) The weight set composed of the weights between the fourth node of the first hidden layer and each node of the input layer, the weights between the fourth node of the hidden layer and the first node of the input layer, the weights between the fourth node of the hidden layer and the second node of the input layer, and the weights between the fourth node of the hidden layer and the third node of the input layer are all greater than the weights between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
[0069] To illustrate the network composition of the neural network regression model in the above embodiments, the following section provides an example of a network architecture in a specific application instance. For example... Figure 3 As shown, the input layer of the neural network regression model includes four nodes, the hidden layer includes one hidden layer with four nodes, and the output layer includes two nodes. The final output of the neural network regression model, i.e., the mud density value of the current layer, is obtained through activation functions such as sigmoid.
[0070] Assume a 1i Let i be the output of the i-th node in the hidden layer, where i = 1, 2, 3, 4. Then the outputs of each node in the hidden layer can be represented by the following formula:
[0071]
[0072] In the above formula, x1 represents the input of the first node of the input layer and is the X-axis vibration noise factor; x2 represents the input of the second node of the input layer and is the Y-axis vibration noise factor; x3 represents the input of the third node of the input layer and is the Z-axis vibration noise factor; x4 represents the input of the fourth node of the input layer and is the pre-processed mud density monitoring data. This represents the weight between the i-th node in the hidden layer and the first node in the input layer; This represents the weight between the i-th node in the hidden layer and the second node in the input layer; This represents the weight between the i-th node in the hidden layer and the third node in the input layer; This represents the weight between the i-th node in the hidden layer and the fourth node in the input layer. Where:
[0073] The nonlinear mapping and feature extraction process of the first node in the hidden layer is dominated by the X-axis vibration noise factor, i.e., the output a of the first node in the hidden layer.11 In the middle, weight The initial assignment is greater than the weight. Weight Weight Initialization assignment;
[0074] The nonlinear mapping and feature extraction process of the second node in the hidden layer is dominated by the Y-axis vibration noise factor, i.e., the output a of the second node in the hidden layer. 12 In the middle, weight The initial assignment is greater than the weight. Weight Weight Initialization assignment;
[0075] The nonlinear mapping and feature extraction process of the third node in the hidden layer is dominated by the Z-axis vibration noise factor, i.e., the output a of the third node in the hidden layer. 13 In the middle, weight The initial assignment is greater than the weight. Weight Weight Initialization assignment;
[0076] The nonlinear mapping and feature extraction process of the fourth node in the hidden layer is jointly dominated by the X-axis vibration noise factor, Y-axis vibration noise factor, and Z-axis vibration noise factor, i.e., the output a of the fourth node in the hidden layer. 14 In the middle, weight Weight Weight The initial values assigned are all greater than the weights. The initialization assignment.
[0077] Assume Y m Let m be the output of the m-th node in the output layer, where m = 1, 2. Then the outputs of each node in the output layer can be expressed by the following formula:
[0078]
[0079] In the above formula, This represents the weight between the m-th node in the output layer and the first node in the hidden layer. This represents the weight between the m-th node in the output layer and the second node in the hidden layer. This represents the weight between the m-th node in the output layer and the third node in the hidden layer. This represents the weight between the m-th node in the output layer and the fourth node in the hidden layer.
[0080] The final output of the neural network regression model can be expressed by the following formula:
[0081]
[0082] In the above formula, ρ represents the final output of the vibration noise removal model, i.e., the predicted mud density value, f() represents the sigmoid activation function, b represents the bias, and n represents the number of nodes in the output layer.
[0083] In another specific application example, the neural network regression model includes multiple hidden layers. Based on the above application example, those skilled in the art can derive one or more feasible neural network regression model architectures with multiple hidden layers. These feasible neural network regression model architectures will not be described in detail here.
[0084] Example 2
[0085] See Figure 2 This invention provides a method for collecting cementing mud density based on vibration noise removal, which differs from Embodiment 1 in that: after step S300, it further includes the following steps:
[0086] Step S400: Monitor the difference between the mud density value generated by the vibration noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment. If the difference is greater than the first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration noise removal model are updated. Otherwise, the mud density value generated by the vibration noise removal model is used as the mud density value of the current layer.
[0087] The actual mud density value can be obtained by measuring a portion of the mud using methods such as balance.
[0088] It is known that machine learning models often experience model drift after being deployed and running for a period of time. To ensure the prediction accuracy of the machine learning model, drift correction is necessary. When it is found that the mud density value of the current segment generated by the current wellbore vibration and noise removal model differs significantly from the actual mud density value obtained by methods such as balance, it can be determined that the current wellbore vibration and noise removal model has experienced model drift, and parameter adjustment and optimization of the current wellbore vibration and noise removal model are required.
[0089] Through the monitoring process described in the above embodiments, timely correction of the drift of the vibration noise removal model is achieved, thereby ensuring the accuracy of mud density acquisition and guaranteeing cementing quality.
[0090] For example, the degree of difference can be characterized by the difference between the mud density value generated by the vibration noise removal model and the actual mud density value of the corresponding layer obtained by measurement.
[0091] It should be noted that, Figure 1 and Figure 2 This is a flowchart illustrating a cementing mud density acquisition method based on vibration noise removal. It should be understood that, although... Figure 1 and Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 and Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
[0092] Example 3
[0093] See Figure 4 This invention provides a cementing mud density acquisition device 400 based on vibration noise removal, comprising an acquisition module 410, a preprocessing module 420, and a prediction module 430, wherein:
[0094] The acquisition module 410 is used to acquire vibration and noise data in the current section of the target wellbore environment, as well as mud density monitoring data in the current section.
[0095] The preprocessing module 420 is used to preprocess the mud density monitoring data of the current layer.
[0096] The prediction module 430 is used to input the pre-processed mud density monitoring data and vibration noise factor into a pre-built vibration noise removal model to generate the mud density value of the current segment.
[0097] Among them, the vibration noise factor is the component of vibration noise in each direction. The vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
[0098] In one specific embodiment, the vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
[0099] In one specific embodiment, preprocessing includes filtering, denoising, and normalization.
[0100] In one specific embodiment, the regression model is a neural network regression model.
[0101] In one specific embodiment, when initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to its input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
[0102] In one specific embodiment, the hidden layer connected to the input layer in the neural network regression model includes four nodes, wherein:
[0103] The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise.
[0104] The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise.
[0105] The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise.
[0106] The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
[0107] In one specific embodiment, the cementing mud density acquisition device based on vibration and noise removal further includes a monitoring and correction module. The monitoring and correction module is used to monitor the degree of difference between the mud density value generated by the vibration and noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment. If the degree of difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration and noise removal model are updated. Otherwise, the mud density value generated by the vibration and noise removal model is used as the mud density value of the current layer.
[0108] The modules in the above embodiments can be implemented using a general-purpose computing device. A general-purpose computing device typically includes a memory and a processor. The memory stores instructions, and when the processor executes the instructions, the general-purpose computing device performs the various steps or program modules of the embodiments of the present invention. For example, the cementing mud density acquisition device based on vibration noise removal includes a processor and a memory. The acquisition module 410, preprocessing module 420, and prediction module 430, etc., are all stored as program units in the memory, and the processor executes the program modules stored in the memory to achieve the corresponding functions.
[0109] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters enables a method for acquiring cementing mud density based on vibration and noise removal.
[0110] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0111] Example 4
[0112] This invention provides a method for controlling cementing mud density based on vibration and noise removal, comprising the following implementation steps:
[0113] Step SS1: Collect vibration and noise data and mud density monitoring data of the current section of the target wellbore environment;
[0114] Step SS2 involves preprocessing the mud density monitoring data for the current layer.
[0115] Step SS3: Input the pre-processed mud density monitoring data and vibration noise factor into the pre-built vibration noise removal model to generate the mud density value of the current layer.
[0116] Step SS4: If the difference between the mud density value of the current segment generated by removing the vibration and noise model and the mud density design value of the current segment cementing operation exceeds the second threshold, then a control command is generated based on the difference value to change the mud density in the current segment.
[0117] Among them, the vibration noise factor is the component of vibration noise in each direction. The vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
[0118] In one specific embodiment, the vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
[0119] In one specific embodiment, preprocessing includes filtering, denoising, and normalization.
[0120] In one specific embodiment, the regression model is a neural network regression model.
[0121] In one specific embodiment, when initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to the input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
[0122] In one specific embodiment, the hidden layer connected to the input layer in the neural network regression model includes four nodes, wherein:
[0123] The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise.
[0124] The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise.
[0125] The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise.
[0126] The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
[0127] In one specific embodiment, the cementing mud density control method based on vibration noise removal further includes: monitoring the degree of difference between the mud density value generated by the vibration noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment; if the degree of difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration noise removal model are updated; otherwise, the mud density value generated by the vibration noise removal model is used as the mud density value of the current layer.
[0128] On the other hand, embodiments of the present invention also provide a machine-readable storage medium having a program stored thereon, which, when executed by a processor, implements the above-described cementing mud density acquisition method based on vibration and noise removal or the above-described cementing mud density control method based on vibration and noise removal.
[0129] In another aspect, embodiments of the present invention provide a processor for running a program, wherein the program executes the above-described cementing mud density acquisition method based on vibration and noise removal or the above-described cementing mud density control method based on vibration and noise removal.
[0130] In one specific embodiment, a computer device is also provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor A01, a network interface A02, a display screen A04, an input device A05, and a memory (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A06. The non-volatile storage medium A06 stores an operating system B01 and a computer program B02. The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 stored in the non-volatile storage medium A06. The network interface A02 is used for communication with external terminals via a network connection. When the computer program is executed by the processor A01, it implements a method for collecting cementing mud density based on vibration noise removal. The display screen A04 can be an LCD screen or an e-ink display screen. The input device A05 can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device casing, or an external keyboard, touchpad, or mouse.
[0131] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0132] In one embodiment, the cementing mud density acquisition device 400 based on vibration noise removal provided by the present invention can be implemented as a computer program, which can be implemented in the form of, for example... Figure 5 The computer device shown runs on this device. The computer device's memory can store the various program modules that make up the cementing mud density acquisition device 400 based on vibration noise removal, for example, Figure 4The data acquisition module 410, preprocessing module 420, and prediction module 430 are shown. The computer program, comprised of these modules, causes the processor to execute the steps in the vibration-noise removal-based cementing mud density acquisition method of the various embodiments of the present invention described in this specification.
[0133] Figure 5 The computer device shown can be used as follows Figure 4 The module execution steps in the cementing mud density acquisition device 400 based on vibration noise removal shown can be performed by the computer equipment through the acquisition module 410 to execute step S100, through the preprocessing module 420 to execute step S200, and through the prediction module 430 to execute step S300.
[0134] In another aspect, embodiments of the present invention also provide a cementing mud density control system based on vibration and noise removal. The system includes the aforementioned computer equipment and control equipment. The control equipment is used to receive the mud density value of the current layer output by the computer equipment and determine whether the difference between the mud density value of the current layer output by the computer equipment and the mud density design value for cementing operations in the current layer exceeds a second threshold. If so, a control command is generated based on the difference value to change the mud density in the current layer.
[0135] For example, in a specific embodiment, the computer equipment and the control equipment can be an integrated device or separate devices. For instance, the computer equipment for implementing the cementing mud density acquisition method based on vibration noise removal is a local device, and the control equipment is a host computer device.
[0136] In another aspect, embodiments of the present invention also provide a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: collecting vibration noise and mud density monitoring data of the current section environment of the target wellbore; preprocessing the mud density monitoring data of the current section; inputting the preprocessed mud density monitoring data and vibration noise factor into a pre-constructed vibration noise removal model to generate the mud density value of the current section; wherein, the vibration noise factor is the directional component of vibration noise, and the vibration noise removal model is obtained by training and parameter optimization of an initial regression model.
[0137] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0138] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0139] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for acquiring cementing mud density based on vibration noise removal, characterized in that, include: Collect vibration and noise data and mud density monitoring data of the current section of the target wellbore environment; Preprocess the mud density monitoring data of the current layer; The pre-processed mud density monitoring data and vibration noise factor are input into a pre-built vibration noise removal model to generate the mud density value of the current layer. The vibration noise factor refers to the directional components of the vibration noise, and the vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
2. The method for acquiring cementing mud density based on vibration noise removal according to claim 1, characterized in that, The vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
3. The method for acquiring cementing mud density based on vibration and noise removal according to claim 1, characterized in that, Preprocessing includes filtering, denoising, and normalization.
4. The method for acquiring cementing mud density based on vibration noise removal according to claim 1 or 2, characterized in that, The regression model is a neural network regression model.
5. The method for acquiring cementing mud density based on vibration and noise removal according to claim 4, characterized in that, When initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to the input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
6. The method for acquiring cementing mud density based on vibration noise removal according to claim 5, characterized in that, The hidden layer connected to the input layer in the neural network regression model includes four nodes, of which: The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise. The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise. The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise. The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
7. The method for acquiring cementing mud density based on vibration noise removal according to claim 1, characterized in that, Also includes: The difference between the mud density value generated by the vibration and noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment is monitored. If the difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration and noise removal model are updated. Otherwise, the mud density value generated by the vibration and noise removal model is used as the mud density value of the current layer.
8. A cementing mud density acquisition device based on vibration noise removal, characterized in that, The device includes: The acquisition module is used to collect vibration and noise data in the current section of the target wellbore, as well as mud density monitoring data for the current section. The preprocessing module is used to preprocess the mud density monitoring data of the current layer. The prediction module is used to input the pre-processed mud density monitoring data and vibration noise factor into a pre-built vibration noise removal model to generate the mud density value of the current layer. The vibration noise factor refers to the directional components of the vibration noise, and the vibration noise removal model is obtained by training and optimizing the parameters of the initial regression model.
9. The cementing mud density acquisition device based on vibration noise removal according to claim 8, characterized in that, The vibration noise factor includes the X-axis component, Y-axis component, and Z-axis component of the vibration noise.
10. The cementing mud density acquisition device based on vibration noise removal according to claim 8, characterized in that, Preprocessing includes filtering, denoising, and normalization.
11. The cementing mud density acquisition device based on vibration noise removal according to claim 8 or 9, characterized in that, The regression model is a neural network regression model.
12. The cementing mud density acquisition device based on vibration noise removal according to claim 11, characterized in that, When initializing the weight hyperparameters involved in the neural network regression model, for any node in the hidden layer connected to the input layer in the neural network regression model, at least one weight value in the weight set composed of the weights between that node and each node in the input layer is greater than the remaining weight values.
13. The cementing mud density acquisition device based on vibration noise removal according to claim 12, characterized in that, The hidden layer connected to the input layer in the neural network regression model includes four nodes, of which: The weight between the first node of the hidden layer and the first node of the input layer is greater than the weight between the first node of the hidden layer and other nodes of the input layer. The input of the first node of the input layer is the X-axis component of the vibration noise. The weight between the second node of the hidden layer and the second node of the input layer is greater than the weight between the second node of the hidden layer and other nodes of the input layer. The input of the second node of the input layer is the Y-axis component of the vibration noise. The weight between the third node of the hidden layer and the third node of the input layer is greater than the weight between the third node of the hidden layer and other nodes of the input layer. The input of the third node of the input layer is the Z-axis component of the vibration noise. The weights between the fourth node of the hidden layer and the first node of the input layer, the second node of the input layer, and the third node of the input layer are all greater than the weight between the fourth node of the hidden layer and the fourth node of the input layer. The input of the fourth node of the input layer is the pre-processed mud density monitoring data.
14. The cementing mud density acquisition device based on vibration noise removal according to claim 8, characterized in that, The device also includes a monitoring and correction module, which is used to monitor the degree of difference between the mud density value generated by the vibration and noise removal model and the actual mud density value of the corresponding layer obtained by actual measurement in the current cementing operation environment. If the degree of difference is greater than a first threshold, the actual mud density value is used as the mud density value of the current layer until the parameters of the vibration and noise removal model are updated. Otherwise, the mud density value generated by the vibration and noise removal model is used as the mud density value of the current layer.
15. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the cementing mud density acquisition method based on vibration noise removal as described in any one of claims 1 to 7.
16. A cementing mud density control system based on vibration noise removal, characterized in that, The system includes the computer device and control device as described in claim 15. The control device is used to receive the mud density value of the current layer output by the computer device, and to determine whether the difference between the mud density value of the current layer output by the computer device and the mud density design value of the cementing operation of the current layer exceeds a second threshold. If so, a control command is generated based on the difference value to change the mud density in the current layer.
17. A machine-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the cementing mud density acquisition method based on vibration noise removal as described in any one of claims 1 to 7.