Prediction method and device for maximum tripping-in capacity of casing

Abstract

The invention provides a method and device for predicting the maximum tripping-in capacity of a casing, and the method comprises the steps: obtaining a comprehensive logging parameter data set based on obtained original feature data; a friction resistance coefficient data set is obtained through a friction resistance torque model in combination with the comprehensive logging parameter data set; based on a comprehensive logging sample in the friction resistance coefficient data set and a friction resistance coefficient label of the comprehensive logging sample, training the initially constructed casing friction resistance coefficient prediction model to obtain a target casing friction resistance coefficient prediction model; inputting the obtained real-time logging data into a target casing friction system prediction model to obtain a corresponding friction prediction coefficient; and on the basis of all the friction prediction coefficients and the friction torque model, obtaining a predicted value of the maximum running capacity of the casing. By establishing a casing friction coefficient prediction model and combining a friction torque model, advanced prediction of the casing friction and the maximum tripping-in capacity is realized, so that the well cementation quality and efficiency are improved, and the engineering safety and controllability are enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of oil drilling, and in particular to a method and a device for predicting the maximum running capacity of a casing. Background Art

[0002] Casing running is an important drilling engineering technology in the oil industry. This technology can effectively control rock collapse and wellbore instability during drilling. During the casing running process, the maximum running capacity of the casing is a key parameter, which determines the casing running depth and the drilling progress.

[0003] The existing prediction of the maximum capacity of casing running is mainly based on empirical formulas and physical models, but these methods have certain limitations. For example, the casing rigid rod model established by Liu Chunquan and the casing soft rod / rigid rod model established by Zhang Yang both have the same friction coefficient value problem; the casing rigid rod model established by Wang Jianjun has verification problems; the casing soft rod model established by Tang Peiyu and Yu Changlu has relatively ideal assumptions and has not been verified by examples. In short, the existing methods cannot accurately solve the problem of casing running prediction. Summary of the invention

[0004] The present invention provides a method and device for predicting the maximum running capacity of casing. By analyzing and processing historical data, an intelligent prediction model for casing friction coefficient is established, and the casing running capacity analysis model is improved, thereby improving the prediction accuracy and timeliness, and further improving the efficiency of drilling engineering.

[0005] In a first aspect, the present invention provides a method for predicting the maximum running capacity of a casing, comprising:

[0006] Based on the acquired original characteristic data, a comprehensive logging parameter data set is obtained;

[0007] By using the friction torque model and combining the comprehensive logging parameter data set, a friction coefficient data set is obtained;

[0008] Based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set, the initially constructed casing friction coefficient prediction model is trained to obtain a target casing friction coefficient prediction model;

[0009] Inputting the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient;

[0010] Based on all the friction prediction coefficients and the friction torque model, a predicted value of the maximum casing running capacity is obtained.

[0011] Optionally, based on all the friction prediction coefficients and the friction torque model, a prediction value of the maximum running capacity of the casing is obtained, including:

[0012] Input all the friction prediction coefficients into the friction torque model to obtain a curve of hook load variation with well depth during casing running;

[0013] Based on the curve of the change of hook load with well depth during the casing running process, the casing running depth when the hook load is 0 is predicted, and the predicted value of the maximum running capacity of the casing is obtained.

[0014] Optionally, based on the acquired original characteristic data, a comprehensive logging parameter data set is obtained, including:

[0015] Unify the sampling frequency of the acquired original feature data;

[0016] Taking the well depth as an index, all the original feature data after the unified frequency are spliced ​​and processed, and a null value column of the drill string friction coefficient is reserved to obtain the comprehensive logging parameter data set.

[0017] Optionally, a friction coefficient data set is obtained by combining the comprehensive logging parameter data set with the friction torque model, including:

[0018] Through the friction torque model, the casing friction coefficient is inverted for the acquired comprehensive logging parameter data set to obtain the friction coefficient;

[0019] The comprehensive logging parameter data set and the friction coefficient are combined to obtain a friction coefficient data set.

[0020] Optionally, after obtaining the friction coefficient dataset by combining the friction torque model with the comprehensive logging parameter dataset, the method further includes:

[0021] A Pearson correlation analysis algorithm is used to perform correlation analysis on the friction coefficient data set, and according to the analysis result, a preset number of correlation maximum parameter items are selected from the friction coefficient data set in sequence.

[0022] In a second aspect, the present invention provides a device for predicting the maximum running capacity of a casing, comprising:

[0023] A comprehensive logging parameter data set acquisition module is used to obtain a comprehensive logging parameter data set based on the acquired original feature data;

[0024] A friction coefficient data set acquisition module, used to obtain a friction coefficient data set by combining the comprehensive logging parameter data set with a friction torque model;

[0025] A target model determination module is used to train the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model;

[0026] A prediction module, used for inputting the acquired real-time logging data into the target casing friction system prediction model to obtain a corresponding friction prediction coefficient;

[0027] The maximum running capacity determination module is used to obtain the maximum running capacity prediction value of the casing based on all the friction resistance prediction coefficients and the friction resistance torque model.

[0028] Optionally, the maximum downloading capacity determination module includes:

[0029] A variation curve acquisition submodule is used to input all the friction resistance prediction coefficients into the friction resistance torque model to obtain a curve of hook load variation with well depth during casing running;

[0030] The maximum running capacity determination submodule is used to predict the casing running depth when the hook load is 0 based on the curve of the hook load changing with the well depth during the casing running process, and obtain the predicted value of the maximum running capacity of the casing.

[0031] Optionally, the comprehensive logging parameter data set acquisition module includes:

[0032] A frequency unification submodule is used to unify the sampling frequency of the acquired original feature data;

[0033] The comprehensive logging parameter data set acquisition submodule is used to splice all the original feature data after the unified frequency with the well depth as the index, and reserve the null value column of the drill string friction coefficient to obtain the comprehensive logging parameter data set.

[0034] In a third aspect, the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method provided in the first aspect are performed.

[0035] In a fourth aspect, the present application provides a storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps in the method provided in the first aspect are performed.

[0036] It can be seen from the above technical solutions that the present invention has the following advantages:

[0037] The present invention provides a method and device for predicting the maximum running capacity of casing, the method comprising: obtaining a comprehensive logging parameter data set based on the acquired original feature data; obtaining a friction coefficient data set by combining the comprehensive logging parameter data set with a friction torque model; training the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and their friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model; inputting the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient; obtaining the maximum running capacity prediction value of casing based on all the friction prediction coefficients and the friction torque model. By integrating the original feature data with the artificial neural network algorithm, a casing friction coefficient prediction model is established, and combined with the friction torque model, the advanced prediction of casing friction and maximum running capacity is achieved, thereby improving the cementing quality and efficiency, and enhancing the engineering safety and controllability. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 paying creative labor.

[0039] Figure 1 A flowchart of a first embodiment of a method for predicting the maximum running capacity of a casing according to the present invention;

[0040] Figure 2 It is a flowchart of a second embodiment of a method for predicting the maximum running capacity of a casing according to the present invention;

[0041] Figure 3 It is a schematic diagram of the friction torque model;

[0042] Figure 4 This is a schematic diagram of the inversion results of the friction coefficient;

[0043] Figure 5 It is the structural diagram of the casing friction coefficient prediction model;

[0044] Figure 6 This is the flow chart for calculating the curve of hook load changing with well depth;

[0045] Figure 7 The curve of the hook load changing with the well depth in this embodiment;

[0046] Figure 8 The present invention is a structural block diagram of an embodiment of a device for predicting the maximum running capacity of a casing according to the present invention. DETAILED DESCRIPTION

[0047] The embodiment of the present invention provides a method and device for predicting the maximum running capacity of casing. By analyzing and processing historical data, an intelligent prediction model for the casing friction coefficient is established, and the casing running capacity analysis model is improved, thereby improving the prediction accuracy and timeliness, and further improving the efficiency of drilling projects.

[0048] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, 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 below 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 creative work are within the scope of protection of the present invention.

[0049] For example, see Figure 1 , Figure 1 The flowchart of the first embodiment of a method for predicting the maximum running capacity of a casing according to the present invention specifically includes:

[0050] S101, obtaining a comprehensive logging parameter data set based on the acquired original characteristic data;

[0051] S102, obtaining a friction coefficient data set by combining the comprehensive logging parameter data set with the friction torque model;

[0052] S103, training the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model;

[0053] S104, inputting the acquired real-time logging data into the target casing friction system prediction model to obtain a corresponding friction prediction coefficient;

[0054] S105, obtaining a predicted value of the maximum running capacity of the casing based on all the friction prediction coefficients and the friction torque model.

[0055] A method for predicting the maximum running capacity of casing provided in an embodiment of the present invention includes: obtaining a comprehensive logging parameter data set based on the acquired original feature data; obtaining a friction coefficient data set by combining the comprehensive logging parameter data set with the friction torque model; training the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and their friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model; inputting the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient; obtaining the maximum running capacity prediction value of casing based on all the friction prediction coefficients and the friction torque model. By integrating the original feature data with the artificial neural network algorithm, a casing friction coefficient prediction model is established, and combined with the friction torque model, the advanced prediction of casing friction and maximum running capacity is achieved, thereby improving the cementing quality and efficiency, and enhancing the engineering safety and controllability.

[0056] For example 2, please refer to Figure 2 , Figure 2 The flowchart of the second embodiment of the method for predicting the maximum running capacity of casing according to the present invention specifically includes:

[0057] S201, unifying the sampling frequency of the acquired original feature data;

[0058] In the embodiment of the present invention, the original feature data is obtained from the time database. The data collected in the time database runs through the entire drilling process, including starting, running, drilling, short tripping, casing, cementing, electrical measurement, and complex accident handling. When obtaining the original feature data, the casing working conditions are judged according to the changes in field data in combination with manual experience. The judgment conditions are: the drilling pressure is 0, the displacement pulse remains unchanged, the big hook height changes, and the drill bit position is 0. After determining the working conditions, the data such as the running depth and the ground hook load are extracted, and then the depth domain conversion is performed as part of the input parameters of the casing friction coefficient inversion model.

[0059] Then, based on Python programming language and combined with the differential equation of string mechanics, the finite difference method was used to establish Figure 3 The friction torque model shown in Figure 1. The model input parameters are friction coefficient, well depth, well inclination, azimuth, wellbore structure, casing structure and drilling fluid density.

[0060] Specifically, the solution process using the finite difference method is as follows:

[0061] The differential equation of the overall force of the drill string is known to be:

[0062]

[0063]

[0064] The calculation formula of the wellbore axis curvature is:

[0065]

[0066] The calculation formula of the contact force between the drill string and the well wall is:

[0067]

[0068]

[0069]

[0070]

[0071] Then, using the finite difference numerical solution method, the overall force model of the drill string can be written as follows:

[0072] When pressure is used as positive:

[0073]

[0074]

[0075] Among them, among them, In the equation, the negative sign represents pulling, and the positive sign represents running; F is the axial pressure on the drill string, N; S is the well depth, m; q is the weight of the drill string per unit length, N / m; α is the well inclination, rad; EI is the bending stiffness of the drill string, N·m2; n t is the contact force between the drill string and the well wall, N / m; μ1 is the axial friction coefficient of the drill string; k b is the curvature of the wellbore axis, m-1; F i and F i+1 are the axial forces of the drill string near the ground and near the drill bit in the i-th section, N; M Ti and M Ti+1 are the torques at both ends of the drill string in the i-th section, N·m; and are the wellbore curvatures at both ends of the i-th section, m-1; q i , EI i 、n ti , Δs i and D bi They are the linear weight (N / m), bending stiffness (N·m2), contact force with the wellbore wall (N / m), length (m) and outer diameter (m) of the i-th section of drill string.

[0076] S202, using the well depth as an index, performing splicing processing on all the original feature data after the unified frequency, and reserving a null value column of the drill string friction coefficient, to obtain the comprehensive logging parameter data set;

[0077] S203, performing casing friction coefficient inversion on the acquired comprehensive logging parameter data set through a friction torque model to obtain a friction coefficient;

[0078] The embodiment of the present invention uses the binary search method to invert the friction coefficient. The specific process is as follows: first, the friction coefficient is initialized between 0 and 1. The axial force at the end of the casing at the bottom of the hole is 0 and is brought into the overall force model of the casing. The axial force at the upper end of the drill string (close to the ground) is calculated section by section from the bottom of the hole using the axial force transfer equation until the calculated value of the large hook load is obtained on the ground, and compared with the large hook load measured on the ground. If the calculated value and the measured value meet the given error, it means that the friction coefficient is relatively accurate, and the cycle ends. Otherwise, a new friction coefficient is reinitialized using the binary search method, and it is calculated section by section from the drill bit to the ground again. The error between the calculated value and the measured value is compared again, and the calculation is repeated until the accurate casing friction coefficient is found. Figure 4 shown.

[0079] S204, combining the comprehensive logging parameter data set and the friction coefficient to obtain a friction coefficient data set;

[0080] S205, performing a correlation analysis on the friction coefficient data set using a Pearson correlation analysis algorithm, and selecting a preset number of maximum correlation parameter items from the friction coefficient data set in sequence according to the analysis result;

[0081] In the embodiment of the present invention, the original characteristic data includes the data in the well inclination data table, engineering parameter table, drilling fluid performance table, drill bit record table and cuttings description record table provided at the drilling site. In order to make the subsequent casing friction coefficient prediction model have better prediction effect and higher calculation efficiency, it is necessary to clean the original characteristic data and optimize the input data of the casing friction coefficient prediction model, including:

[0082] (1) Unify the sampling frequency of the original feature data: Since the sampling frequency of the engineering parameter table is 1m, the sampling frequency of other original feature data needs to be unified to 1m. For the well inclination data, the cubic spline curve interpolation method is used to interpolate the well trajectory data: First, the entire trajectory is divided into a three-dimensional straight line part and a spatial arc part by calculating the wellbore curvature of each point. The processing of the three-dimensional straight line part is relatively simple, that is, the well inclination angle and azimuth angle at every meter are consistent. For the three-dimensional arc segment, the well inclination data is interpolated using the cubic spline curve: the original data with a sampling frequency of 30 meters is converted to a sampling frequency of 1 meter, so that it can be directly used for subsequent friction torque calculation. The drilling fluid performance data is mainly measured and recorded regularly by the field staff, and the sampling frequency is tens of meters. Since the drilling fluid performance data will not change significantly, it is assumed that the drilling fluid data at every meter in a certain section is the same as the data of the measurement point in the section. Similarly, the drill bit record data and the cuttings description record data are also regularly measured and recorded by the field staff, and it is assumed that the data at every meter in a certain section is the same as the data of the measurement point in the section.

[0083] (2) Data splicing of original feature data: All original feature data with unified frequency are spliced ​​with well depth as index, and a null value column for drill string friction coefficient is reserved. The above data processing flow is implemented through programming to form the final comprehensive logging parameter data set.

[0084] (3) Friction coefficient inversion: The friction torque model is used to invert the casing friction coefficient based on the comprehensive path parameter data set, and the friction coefficient at each meter along the well depth is finally obtained. The friction coefficient data set is then merged with the comprehensive logging parameter data set to obtain the friction coefficient data set. The inversion effect is shown in Figure 2. Figure 3 The schematic diagram of the inversion results of the friction coefficient is shown in Figure 2.

[0085] (4) Correlation analysis of friction coefficient data set: In order to simplify the casing friction coefficient prediction model and reduce the risk of overfitting, thereby increasing the generalization ability of the casing friction coefficient prediction model. The Pearson correlation analysis algorithm is used to perform correlation analysis on the friction coefficient data set, and according to the analysis results, the top 15 parameters with the largest correlation are selected in turn. The top 15 parameters of the embodiment of the present invention are standard well depth, well inclination, azimuth, borehole curvature, closure distance, closure azimuth, drilling fluid density, viscosity, yield strength, plastic viscosity, 600 rpm reading, drilling fluid system, well diameter, drill string outer diameter and lithology. Among them, the Pearson correlation calculation expression is:

[0086]

[0087] Where X and Y are variables; Cov(X,Y) is the covariance of X and Y; D(X) is the variance of X; D(Y) is the variance of Y; ρ XYis the Pearson correlation coefficient of X and Y.

[0088] (5) Normalization of friction coefficient data set: In order to eliminate the scale differences of each feature in the data set and thus avoid the occurrence of gradient problems, the embodiment of the present invention normalizes the data to the maximum and minimum values, thereby eliminating the influence of the dimension and accelerating the training speed of the casing friction coefficient prediction model.

[0089] (6) Data division and model evaluation of the friction coefficient data set: All the data of the friction coefficient data set are divided according to the well depth index in a ratio of 8:2, of which 4727 data are used as the training set and 1182 data are used as the test set. Then RMSE is selected as the model to evaluate the hook load prediction effect. The evaluation expression of RMSE is:

[0090]

[0091] S206, training the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model;

[0092] In the embodiment of the present invention, please refer to Figure 5 , Figure 5 The figure is a structural diagram of the casing friction coefficient prediction model. To ensure the regression performance of the casing friction coefficient prediction model, the number of hidden layers of the casing friction coefficient prediction model is set to 2. The number of hidden layer neurons is reasonably set by the grid search method to ensure the prediction accuracy of the casing friction coefficient prediction model and avoid model overfitting. A prediction model with a topology of 14-128-128-1 is established. The activation function is Tanh, and the learning rate is 0.1. On this basis, the casing friction coefficient prediction model is trained based on the friction coefficient data set to realize the intelligent prediction of the friction coefficient.

[0093] In the specific training process, let y(t) represent the true value of the friction coefficient, f(t) represent the neural network predicted value of the friction coefficient, and e(t) represent the difference between the true value and the predicted value, then:

[0094] y(t)=f(t)+e(t);

[0095] The neural network model achieves accurate prediction by minimizing the back propagation method. Taking a single hidden layer neural network as an example, the model is expressed as follows:

[0096]

[0097] Among them, E is the loss function of the neural network, that is, the objective function of the optimization problem.

[0098] S207, inputting the acquired real-time logging data into the target casing friction system prediction model to obtain a corresponding friction prediction coefficient;

[0099] S208, inputting all the friction prediction coefficients into the friction torque model to obtain a curve of hook load variation with well depth during casing running;

[0100] S209, based on the curve of the change of hook load with well depth during the casing running process, predict the casing running depth when the hook load is 0, and obtain the predicted value of the maximum running capacity of the casing.

[0101] In the embodiment of the present invention, the predicted friction coefficient is substituted into the friction torque model under the sliding lowering condition, and the bottom hole boundary condition torque and drilling pressure are set to 0, so that the wellhead axial force (hook load) can be inversely calculated.

[0102] See also Figure 6 , Figure 6 The following is a flow chart of the calculation of the curve of the hook load changing with the well depth. In the specific implementation, the calculation is cyclically calculated from the wellhead to the well bottom, and the following is obtained: Figure 7 The curve of the hook load changing with the depth of the casing during the casing running process is shown. The maximum casing running depth is determined by the curve change trend and the point where the maximum hook load is equal to 0. The specific steps include: first, collecting the parameters required in the casing rigid rod model, that is, the friction torque model. In addition to the friction coefficient, it is also necessary to determine the trajectory, length, weight, bending stiffness, and pipe string structure and other related parameters of the casing. Determine the initial casing running depth and initial hook load value, and then substitute the initial casing running depth and hook load value into the casing rigid rod model for calculation. The model calculates the corresponding hook load value at the current depth based on the casing friction coefficient and the casing length, weight, bending stiffness and other parameters. Determine whether the calculated hook load value is 0. If the hook load value is 0, it means that the casing has been completely run. If the hook load value is 0, it can be concluded that the current depth is the predicted casing running depth. If the hook load value is not 0, it means that the casing has not been completely run, and the next step of iterative calculation is required. Repeat the iterative calculation until the hook load value is close enough to 0 or reaches the preset accuracy requirement. At this time, the casing running depth when the hook load is 0 can be accurately predicted, and the maximum running capacity of this embodiment is 3150m.

[0103] A method for predicting the maximum running capacity of casing provided in an embodiment of the present invention includes: obtaining a comprehensive logging parameter data set based on the acquired original feature data; obtaining a friction coefficient data set by combining the comprehensive logging parameter data set with the friction torque model; training the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and their friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model; inputting the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient; obtaining the maximum running capacity prediction value of casing based on all the friction prediction coefficients and the friction torque model. By integrating the original feature data with the artificial neural network algorithm, a casing friction coefficient prediction model is established, and combined with the friction torque model, the advanced prediction of casing friction and maximum running capacity is achieved, thereby improving the cementing quality and efficiency, and enhancing the engineering safety and controllability.

[0104] For example 3, please refer to Figure 8 , Figure 8 The present invention is a structural block diagram of an embodiment of a device for predicting the maximum running capacity of a casing, the device comprising:

[0105] A comprehensive logging parameter data set acquisition module 301 is used to obtain a comprehensive logging parameter data set based on the acquired original feature data;

[0106] A friction coefficient data set acquisition module 302 is used to obtain a friction coefficient data set by combining the comprehensive logging parameter data set with a friction torque model;

[0107] A target model determination module 303 is used to train the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model;

[0108] The prediction module 304 is used to input the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient;

[0109] The maximum running capacity determination module 305 is used to obtain the maximum running capacity prediction value of the casing based on all the friction prediction coefficients and the friction torque model.

[0110] In an optional embodiment, the maximum loading capacity determination module 305 includes:

[0111] A variation curve acquisition submodule is used to input all the friction resistance prediction coefficients into the friction resistance torque model to obtain a curve of hook load variation with well depth during casing running;

[0112] The maximum running capacity determination submodule is used to predict the casing running depth when the hook load is 0 based on the curve of the hook load changing with the well depth during the casing running process, and obtain the predicted value of the maximum running capacity of the casing.

[0113] In an optional embodiment, the comprehensive logging parameter data set acquisition module 301 includes:

[0114] A frequency unification submodule is used to unify the sampling frequency of the acquired original feature data;

[0115] The comprehensive logging parameter data set acquisition submodule is used to splice all the original feature data after the unified frequency with the well depth as the index, and reserve the null value column of the drill string friction coefficient to obtain the comprehensive logging parameter data set.

[0116] In an optional embodiment, the friction coefficient data set acquisition module 301 includes:

[0117] The friction coefficient determination submodule is used to perform casing friction coefficient inversion on the acquired comprehensive logging parameter data set through a friction torque model to obtain the friction coefficient;

[0118] The data set acquisition submodule is used to merge the comprehensive logging parameter data set with the friction coefficient to obtain a friction coefficient data set.

[0119] In an optional embodiment, it also includes:

[0120] The screening module is used to perform correlation analysis on the friction coefficient data set using the Pearson correlation analysis algorithm, and select the maximum correlation parameter items of a preset number of items from the friction coefficient data set in turn according to the analysis result.

[0121] Embodiment 4, the embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of a method for predicting the maximum casing running capacity of an embodiment.

[0122] Embodiment 5, the embodiment of the present invention further provides a computer storage medium, on which a computer program is stored, and when the computer program is executed by the processor, the steps of a method for predicting the maximum running capacity of casing in the embodiment are implemented.

[0123] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0124] In the several embodiments provided in the present application, it should be understood that the methods, devices, electronic devices and storage media disclosed in the present invention can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as 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.

[0125] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0126] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0127] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a readable storage medium, including 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 readable storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0128] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the maximum running capacity of casing, characterized in that: include: Based on the acquired original characteristic data, a comprehensive logging parameter data set is obtained; By using the friction torque model and combining the comprehensive logging parameter data set, a friction coefficient data set is obtained; Based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set, the initially constructed casing friction coefficient prediction model is trained to obtain a target casing friction coefficient prediction model; Inputting the acquired real-time logging data into the target casing friction system prediction model to obtain the corresponding friction prediction coefficient; Based on all the friction prediction coefficients and the friction torque model, a predicted value of the maximum casing running capacity is obtained.

2. The method for predicting the maximum running capacity of casing according to claim 1, characterized in that: Based on all the friction prediction coefficients and the friction torque model, the maximum casing running capacity prediction value is obtained, including: Input all the friction prediction coefficients into the friction torque model to obtain a curve of hook load variation with well depth during casing running; Based on the curve of the change of hook load with well depth during the casing running process, the casing running depth when the hook load is 0 is predicted, and the predicted value of the maximum running capacity of the casing is obtained.

3. The method for predicting the maximum running capacity of casing according to claim 1, characterized in that: Based on the acquired original characteristic data, a comprehensive logging parameter data set is obtained, including: Unify the sampling frequency of the acquired original feature data; Taking the well depth as an index, all the original feature data after the unified frequency are spliced ​​and processed, and a null value column of the drill string friction coefficient is reserved to obtain the comprehensive logging parameter data set.

4. The method for predicting the maximum running capacity of casing according to claim 1, characterized in that: Through the friction torque model, combined with the comprehensive logging parameter data set, a friction coefficient data set is obtained, including: Through the friction torque model, the casing friction coefficient is inverted for the acquired comprehensive logging parameter data set to obtain the friction coefficient; The comprehensive logging parameter data set and the friction coefficient are combined to obtain a friction coefficient data set.

5. The method for predicting the maximum running capacity of casing according to claim 3, characterized in that After obtaining the friction coefficient data set by combining the friction torque model with the comprehensive logging parameter data set, the method further includes: A Pearson correlation analysis algorithm is used to perform correlation analysis on the friction coefficient data set, and according to the analysis result, a preset number of correlation maximum parameter items are selected from the friction coefficient data set in sequence.

6. A device for predicting the maximum running capacity of casing, characterized in that: include: A comprehensive logging parameter data set acquisition module is used to obtain a comprehensive logging parameter data set based on the acquired original feature data; A friction coefficient data set acquisition module, used to obtain a friction coefficient data set by combining the comprehensive logging parameter data set with a friction torque model; A target model determination module is used to train the initially constructed casing friction coefficient prediction model based on the comprehensive logging samples and friction coefficient labels in the friction coefficient data set to obtain a target casing friction coefficient prediction model; A prediction module, used for inputting the acquired real-time logging data into the target casing friction system prediction model to obtain a corresponding friction prediction coefficient; The maximum running capacity determination module is used to obtain the maximum running capacity prediction value of the casing based on all the friction resistance prediction coefficients and the friction resistance torque model.

7. The device for predicting the maximum running capacity of casing according to claim 6, characterized in that: The maximum downloading capacity determination module includes: A variation curve acquisition submodule is used to input all the friction resistance prediction coefficients into the friction resistance torque model to obtain a curve of hook load variation with well depth during casing running; The maximum running capacity determination submodule is used to predict the casing running depth when the hook load is 0 based on the curve of the hook load changing with the well depth during the casing running process, and obtain the predicted value of the maximum running capacity of the casing.

8. The device for predicting the maximum running capacity of casing according to claim 6, characterized in that: The comprehensive logging parameter data set acquisition module includes: A frequency unification submodule is used to unify the sampling frequency of the acquired original feature data; The comprehensive logging parameter data set acquisition submodule is used to splice all the original feature data after the unified frequency with the well depth as the index, and reserve the null value column of the drill string friction coefficient to obtain the comprehensive logging parameter data set.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 5 is executed.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is performed.

Images