Method and apparatus for analyzing friction resistance of drill string
Through the combination of neural network and friction resistance calculation theoretical model, the impact of rock chip bed and wellbore trajectories on friction resistance is separated and analyzed, and the problem of low friction resistance calculation accuracy in the existing technology is solved, and more accurate friction resistance analysis is achieved , improves drilling efficiency.
Patent Information
- Application Number
- PCT/CN2023/136653
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-06-02
- Filing Date
- 2023-12-06
- Publication Date
- 2024-12-05
AI Technical Summary
In the calculation of the friction torque of the drill string, the assumption that the conditions are not comprehensive enough, resulting in the calculation accuracy of the friction resistance. Especially in the sliding drilling conditions, the formation of the rock chip bed due to the unclean borehole, which increases the friction of the drill string. Resistance affects drilling efficiency.
Using neural network to establish a bottom-hole drilling pressure torque prediction model, combine the friction resistance calculation theoretical model, train the friction resistance coefficient intelligent analysis model, separate the conventional friction resistance coefficient and the additional friction resistance coefficient, and analyze the coupling effect of the friction between the cutting bed and the wellbore trajectory The impact of resistance.
It improves the accuracy of friction resistance calculation, helps staff clean the wellbore and adjust parameters in a timely manner, and improves drilling efficiency.
Smart Images

Figure CN2023136653_05122024_PF_FP_ABST
Abstract
Description
Drill string friction analysis method and device Technical Field
[0001] The present invention relates to the technical field of oil and gas drilling, and in particular to a drill string friction analysis method and device. Background Art
[0002] This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims. No statement herein is admitted to be prior art by virtue of its inclusion in this section.
[0003] In recent years, with the exploration and development of unconventional oil and gas resources and the efficient development of "three-low" oil and gas fields, extended-reach wells and horizontal wells, the primary well types for unconventional oil and gas extraction, have become widely used. As the length of the horizontal section increases, the friction and torque of the drill string also increase. This is especially true during sliding drilling. Because the drill string does not rotate, it can easily lead to unclean wellbore, forming cuttings beds in highly deviated and horizontal sections, increasing overall drill string friction and affecting drilling efficiency.
[0004] To calculate the friction torque of the drill string, most methods currently use a soft rod model or a rigid rod model. However, due to certain assumptions set during the friction calculation, these methods do not fully consider the factors affecting the friction, resulting in low calculation accuracy.
[0005] Summary of the Invention
[0006] In an embodiment of the present invention, a drill string friction analysis method is proposed to quantitatively analyze the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory, including:
[0007] Obtain theoretical model for friction calculation;
[0008] Determine the current drilling bottom hole weight on bit and bottom hole torque according to the bottom hole weight on bit torque prediction model; the bottom hole weight on bit torque prediction model is first established using a neural network and then trained based on a preset drilling data set;
[0009] According to the current drilling bottom hole pressure and bottom hole torque, the friction coefficient is determined using the friction calculation theoretical model;
[0010] Obtaining friction coefficient influencing parameters, and training a friction coefficient intelligent analysis model established using a neural network based on the friction coefficient and the friction coefficient influencing parameters; the friction coefficient intelligent analysis model is used to analyze the influence of the coupling between the cuttings bed and the wellbore trajectory on the friction coefficient;
[0011] According to the intelligent analysis model of friction coefficient, the conventional friction coefficient and additional friction coefficient are determined; among them, the conventional friction coefficient is the friction coefficient caused by the wellbore wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
[0012] In an embodiment of the present invention, a drill string friction analysis device is provided to quantitatively analyze the effect of the coupling between the cuttings bed and the wellbore trajectory on the friction coefficient, including:
[0013] Theoretical model acquisition module, used to obtain the theoretical model for friction calculation;
[0014] A WOB torque determination module is used to determine the current bottom hole WOB and bottom hole torque based on a bottom hole WOB torque prediction model; the bottom hole WOB torque prediction model is first established using a neural network and then trained based on a preset drilling data set;
[0015] A friction coefficient determination module is used to determine the friction coefficient using a friction calculation theoretical model based on the current drilling bottom hole pressure and bottom hole torque;
[0016] The analysis model training module is used to obtain the friction coefficient influencing parameters and train the friction coefficient intelligent analysis model established by the neural network based on the friction coefficient and the friction coefficient influencing parameters. The friction coefficient intelligent analysis model is used to analyze the influence of the coupling between the cuttings bed and the wellbore trajectory on the friction coefficient.
[0017] The friction coefficient analysis module is used to determine the conventional friction coefficient and the additional friction coefficient based on the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
[0018] In an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, a drill string friction analysis method is implemented.
[0019] An embodiment of the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a drill string friction analysis method is implemented.
[0020] The drill string friction analysis method and device proposed in the embodiment of the present invention can solve the problem that the existing technology uses the method of establishing a soft rod model or a rigid rod model to calculate the friction coefficient, which is not fully considered and leads to low friction calculation accuracy; by obtaining a friction calculation theoretical model; determining the bottom hole drilling pressure and bottom hole torque of the current drilling according to the bottom hole drilling pressure and bottom hole torque prediction model; the bottom hole drilling pressure and bottom hole torque prediction model is first established using a neural network and then obtained after training based on a preset drilling data set; determining the friction coefficient based on the bottom hole drilling pressure and bottom hole torque of the current drilling using the friction calculation theoretical model; obtaining friction coefficient influencing parameters, and training a friction coefficient intelligent analysis model established by using a neural network based on the friction coefficient and the friction coefficient influencing parameters; wherein the friction coefficient intelligent analysis model is used to analyze the influence of the coupling effect of the cuttings bed and the wellbore trajectory on the friction coefficient; and determining the conventional friction coefficient and the additional friction coefficient based on the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory. The embodiment of the present invention considers the effect of the coupling between the cuttings bed and the wellbore trajectory on frictional resistance, and realizes quantitative analysis of the frictional resistance coefficient caused by the coupling between the cuttings bed and the wellbore trajectory, which helps to improve the accuracy of frictional resistance calculation, guide workers to clean the wellbore in time, and improve drilling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0022] FIG1 is a schematic flow chart of a drill string friction analysis method according to an embodiment of the present invention;
[0023] FIG2 is a schematic diagram of the force on the soft rod model according to an embodiment of the present invention;
[0024] FIG3 is a schematic diagram of the structure of a gated recurrent unit recurrent neural network according to an embodiment of the present invention;
[0025] FIG4 is a specific example of a drill string friction analysis method according to an embodiment of the present invention;
[0026] FIG5 is a specific example of a drill string friction analysis method according to an embodiment of the present invention; FIG2;
[0027] FIG6 is a specific example of a drill string friction analysis method according to an embodiment of the present invention; FIG3;
[0028] FIG7 is a specific example of a drill string friction analysis method according to an embodiment of the present invention; FIG4;
[0029] FIG8 is a specific example of a drill string friction analysis method according to an embodiment of the present invention; FIG5;
[0030] FIG9 is a specific example of a drill string friction analysis method according to an embodiment of the present invention; FIG6;
[0031] FIG10 is a diagram showing a specific example of training a friction coefficient intelligent analysis model according to an embodiment of the present invention;
[0032] FIG11 is a diagram showing a specific example of a friction coefficient analysis device according to an embodiment of the present invention;
[0033] FIG12 is a schematic diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0034] To make the purpose, technical solutions and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.
[0035] The term "and / or" herein simply describes an association relationship, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, and the existence of B alone. In addition, the term "at least one" herein refers to any combination of at least two of any one or more of a plurality of items. For example, "at least one of A, B, and C" can represent any one or more elements selected from the set consisting of A, B, and C.
[0036] In the description of this specification, the terms "include", "including", "have", "contain", etc. are all open terms, which mean including but not limited to. The descriptions with reference to the terms "one embodiment", "a specific embodiment", "some embodiments", "for example", etc. mean that the specific features, structures or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. The order of steps involved in each embodiment is used to schematically illustrate the implementation of the present application, and the order of steps therein is not limited and can be appropriately adjusted as needed.
[0037] The principles and spirit of the present invention are explained in detail below with reference to several representative embodiments of the present invention.
[0038] FIG1 is a flow chart of a drill string friction analysis method according to an embodiment of the present invention. As shown in FIG1 , the method includes:
[0039] Step 101, obtaining a friction resistance calculation theoretical model;
[0040] Step 102: Determine the bottom hole weight on bit and bottom hole torque of the current drilling according to a bottom hole weight on bit torque prediction model; the bottom hole weight on bit torque prediction model is first established using a neural network and then trained based on a preset drilling data set;
[0041] Step 103, determining the friction coefficient using a friction calculation theoretical model based on the current drilling bottom hole pressure and bottom hole torque;
[0042] Step 104: Obtain friction coefficient influencing parameters, and train a friction coefficient intelligent analysis model established using a neural network based on the friction coefficient and the friction coefficient influencing parameters; wherein the friction coefficient intelligent analysis model is used to analyze the influence of the coupling between the cuttings bed and the wellbore trajectory on the friction coefficient;
[0043] Step 105 , determining a conventional friction coefficient and an additional friction coefficient based on the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the wellbore wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
[0044] As can be seen from the process shown in Figure 1, the embodiment of the present invention obtains a friction calculation theoretical model based on actual on-site needs and calculation time; uses a neural network to establish a bottom hole drilling pressure torque prediction model, and trains the bottom hole drilling pressure torque prediction model based on a preset drilling data set, so that the bottom hole drilling pressure torque prediction model has the ability to predict the bottom hole drilling pressure and bottom hole torque in real time through the surface hook load, surface torque and related parameters; determines the bottom hole drilling pressure and bottom hole torque of the current drilling based on the trained bottom hole drilling pressure torque prediction model; determines the friction coefficient based on the bottom hole drilling pressure and bottom hole torque of the current drilling using the friction calculation theoretical model; obtains the parameters affecting the friction coefficient, and calculates the friction coefficient based on the bottom hole drilling pressure and bottom hole torque of the current drilling. Based on the friction coefficient and friction coefficient influencing parameters, a friction coefficient intelligent analysis model established using a neural network is trained; wherein, the friction coefficient intelligent analysis model is used to analyze the influence of the coupling effect of the cuttings bed and the wellbore trajectory on the friction coefficient; based on the friction coefficient intelligent analysis model, the conventional friction coefficient and the additional friction coefficient are determined; wherein, the conventional friction coefficient is the friction coefficient caused by the wellbore wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory; in the prior art, the method of establishing a soft rod model or a rigid rod model is used to calculate the friction coefficient, without considering the influence of the coupling effect of the cuttings bed and the wellbore trajectory on friction, resulting in low friction calculation accuracy. The embodiments of the present invention consider the influence of the coupling effect of the cuttings bed and the wellbore trajectory on friction, quantitatively analyze the friction coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory, can improve the accuracy of friction calculation, guide personnel to take cleaning measures and adjust parameters in a timely manner, and facilitate safe, efficient and high-quality drilling.
[0045] In order to explain the above drill string friction analysis method more clearly, each step is described in detail below.
[0046] In one embodiment of the present invention, for step 101, a theoretical model for calculating friction resistance is obtained based on actual on-site needs and calculation time requirements. The theoretical model for calculating friction resistance includes a soft rod model and a rigid rod model. By comparing the soft rod model and the rigid rod model, it can be seen that the soft rod model has a fast calculation convergence speed and the overall accuracy can meet the requirements of the actual well site. Although it does not consider the rigidity of the drill string like the rigid rod model, in actual drilling operations, a drill string with strong rigidity is generally prevented from entering the horizontal section.
[0047] In specific implementation, due to the complex modeling process of the finite element model and multiple boundary conditions, it is often difficult to converge under well site noise, and the finite element model tends to analyze the friction resistance at a certain point. Therefore, the soft rod model is selected as the theoretical model for friction resistance calculation. The force diagram of the soft rod model is shown in Figure 2. The relationship between the tension at the upper end of the drill string and the tension at the lower end of the drill string is: T i =T i+1 +Wcosα±F μ
[0048] The relationship between the torque on the upper end of the drill string and the torque on the lower end of the drill string is: M i =M i+1 +M μ
[0049] The friction coefficient of the drill string includes the axial friction coefficient and the circumferential friction coefficient. The relationship between the axial friction coefficient and the axial friction force of the drill string, and the relationship between the circumferential friction coefficient and the circumferential torque of the drill string are: F μ =μ a NM μ =μ t Nr
[0050] The calculation formula of the support force of the well wall on the drill string is as follows:
[0051] Where N is the support force of the well wall on the drill string, that is, the contact normal pressure between the drill string and the well wall, the unit is N; T i ,T i+1 They are the tension at the upper and lower ends of the drill string, respectively, in N; T is the axial force of the drill string, and T can be taken when calculating the support force of the well wall on the drill string. i ,T i+1 Either one or the average of the two, in N; W is the buoyant weight of the drill string in the drilling fluid, in N; M i ,M i+1 are the torques on the upper and lower ends of the drill string, in N·m; μ a ,μ tare the axial friction coefficient and circumferential friction coefficient of the drill string, respectively, which are dimensionless quantities; r is the radius of the drill string, which changes with the drilling times, and the unit is mm; α is the well inclination, and the unit is °; Δα is the well inclination change at the upper and lower ends of the drill string, and the unit is °; is the azimuth, in degrees; F is the azimuth change at the upper and lower ends of the drill string, in degrees; μ is the axial friction force on the drill string, in N; M μ It is the circumferential torque on the drill string, in N·m.
[0052] In one embodiment of the present invention, the preset drilling data set in step 2 includes: an input data set and an output data set; wherein the input data set includes one or any combination of the following: surface hook load, surface torque, drill string outer diameter, average well inclination, drill string linear weight, drilling fluid density, well inclination difference between two ends of the drill string, and azimuth difference between two ends of the drill string; and the output data set includes bottomhole bit pressure and bottomhole torque.
[0053] In one embodiment of the present invention, for step 102, in order to extract relevant features of the time series data during the drilling process, the bottom hole weight on bit torque prediction model is established using a gated recurrent unit recurrent neural network. The selection is based on the fact that compared with recurrent neural networks such as LSTM (long short-term memory network), its structure is simpler, the computational efficiency is higher, and it can effectively extract the time series features in the data; the structure of the gated recurrent unit recurrent neural network is shown in Figure 3, where σ and tanh are activation functions; in the figure, z t is the update gate, which is used to control the extent to which the information of the previous moment is brought into the current state. t The larger the value, the more state information from the previous moment is brought in; t is the reset gate, which is used to control the degree of ignoring the state information of the previous moment, r t The smaller the value, the more state information is ignored; by updating the gate and resetting the gate structure, the gated recurrent unit recurrent neural network can retain some of the previous memory h t-1 , combined with the new input x t And get the activation feature The feature h at this moment can be output t , in the embodiment of the present invention, are the bottom hole weight on bit and bottom hole torque at this moment.
[0054] In one embodiment of the present invention, the surface hook load, surface torque, drill string outer diameter, average well inclination, drill string linear weight, drilling fluid density, the difference in well inclination between the two ends of the drill string, and the difference in azimuth between the two ends of the drill string are used as inputs of the bottom hole weight-on-bit torque prediction model, and the bottom hole weight-on-bit and bottom hole torque are used as outputs of the bottom hole weight-on-bit torque prediction model to train the bottom hole weight-on-bit torque prediction model.
[0055] Regarding step 102, referring to FIG4, the detailed process of determining the bottom hole weight on bit and bottom hole torque of the current drilling according to the bottom hole weight on bit torque prediction model is as follows:
[0056] Step 401 , inputting the current drilling input data set into a bottom hole weight on bit torque prediction model;
[0057] Step 402: Determine the bottom hole weight on bit and bottom hole torque of the current drilling.
[0058] In one embodiment of the present invention, the surface hook load, surface torque, drill string outer diameter, average well inclination, drill string linear weight, drilling fluid density, well inclination difference between the two ends of the drill string, and azimuth difference between the two ends of the drill string in the input data set of the current drilling are input into a bottom hole weight on bit torque prediction model to determine the bottom hole weight on bit and bottom hole torque of the current drilling.
[0059] Regarding step 103, referring to FIG5 , the detailed process of determining the friction coefficient using the friction calculation theoretical model according to the current drilling bottom hole pressure and bottom hole torque is as follows:
[0060] Step 501, setting an initial friction coefficient and determining it as the current friction coefficient;
[0061] Step 502: Repeat the following steps until the difference between the calculated ground hook load and the actual ground hook load is less than or equal to a first preset error, or the difference between the calculated ground torque and the actual ground torque is less than or equal to a second preset error, and output the current friction coefficient as the determined friction coefficient:
[0062] Step 503, calculating the surface hook load and surface torque based on the current drilling bottom hole weight, bottom hole torque and current friction coefficient using a friction calculation theoretical model;
[0063] Step 504: When the difference between the calculated ground hook load and the actual ground hook load is greater than a first preset error or the difference between the calculated ground torque and the actual ground torque is greater than a second preset error, the friction coefficient is modified and the modified friction coefficient is used as the current friction coefficient.
[0064] In one embodiment of the present invention, the surface hook load is used as the tension at the upper end of the drill string, the surface torque is used as the torque at the upper end of the drill string, and the bottomhole weight on bit and bottomhole torque of the current drilling process are used as the tension and torque at the lower end of the drill string. A soft rod model is used to calculate the surface hook load and surface torque, and the friction coefficient is inverted. The determined friction coefficient includes the axial friction coefficient and the circumferential friction coefficient. If no drilling accident occurs and a workover is performed, the friction coefficient of the completed well section is assumed to remain unchanged. The friction coefficient of each well section, along with its corresponding formation parameters, construction parameters, drill string performance parameters, drilling fluid parameters, wellbore trajectory parameters, cuttings bed parameters, and drilling operation parameters, is collected and organized to form a drill string friction coefficient database.
[0065] During specific implementation, the axial friction coefficient and the circumferential friction coefficient can be iteratively calculated using the soft rod model respectively; the steps for calculating the axial friction coefficient are as follows: first, an initial axial friction coefficient is set and determined as the current axial friction coefficient; the following steps are repeated until the difference between the calculated ground hook load and the actual ground hook load is less than or equal to a first preset error, and the current axial friction coefficient is output as the determined axial friction coefficient: based on the current drilling bottom hole pressure and the current axial friction coefficient, the tension at the upper end of the drill string, that is, the ground hook load, is calculated using the soft rod model; when the difference between the calculated ground hook load and the actual ground hook load is greater than the first preset error, the axial friction coefficient is modified, and the modified axial friction coefficient is used as the current axial friction coefficient. The steps for calculating the circumferential friction coefficient are as follows: first, an initial circumferential friction coefficient is set and determined as the current circumferential friction coefficient; the following steps are repeated until the difference between the calculated surface torque and the actual surface torque is less than or equal to a second preset error, and the current circumferential friction coefficient is output as the determined circumferential friction coefficient: based on the current bottom hole torque of drilling and the current circumferential friction coefficient, the upper end torque of the drill string, i.e., the surface torque, is calculated using a soft rod model; when the difference between the calculated surface torque and the actual surface torque is greater than the second preset error, the circumferential friction coefficient is modified, and the modified circumferential friction coefficient is used as the current circumferential friction coefficient.
[0066] Regarding step 104, referring to FIG6 , the friction coefficient influencing parameters are obtained. The detailed process of training the friction coefficient intelligent analysis model established by the neural network based on the friction coefficient and the friction coefficient influencing parameters is as follows:
[0067] Step 601, obtaining friction coefficient influencing parameters; wherein the friction coefficient influencing parameters include: formation parameters, drill string performance parameters, drilling fluid parameters, drilling operation parameters, cuttings bed parameters and wellbore trajectory parameters;
[0068] Step 602: Connect the first sub-neural network and the second sub-neural network in parallel to establish an intelligent analysis model for the friction coefficient; wherein the first sub-neural network and the second sub-neural network are both preset BP neural networks;
[0069] Step 603: training a friction coefficient intelligent analysis model based on the friction coefficient and the friction coefficient influencing parameters.
[0070] In one embodiment of the present invention, referring to FIG7 , the first sub-neural network and the second sub-neural network built based on the BP neural network both select a structure of 1 input layer, 1 hidden layer, and 1 output layer. The first sub-neural network and the second sub-neural network are connected in parallel to establish an intelligent analysis model of the friction coefficient, and the grid search method is used to optimize the number of hidden layer neurons.
[0071] In one embodiment of the present invention, parameters influencing the friction coefficient are obtained, and the parameters influencing the friction coefficient include: formation parameters (formation lithology, formation elastic coefficient, formation compressive strength), drill string performance parameters (drill string stiffness, drill string wall thickness, drill string outer diameter), drilling fluid parameters (drilling fluid type, density, viscosity, static shear force, solid content), drilling operation parameters (drill string speed, drilling fluid displacement and flow rate, riser pressure, drill tool assembly, drilling method), cuttings bed parameters (wellbore curvature, well inclination, azimuth), and wellbore trajectory parameters (wellbore size, cuttings bed height, annular cuttings concentration, drill string eccentricity); based on the friction coefficient and the parameters influencing the friction coefficient, an intelligent analysis model for the friction coefficient is trained; wherein the friction coefficient includes the axial friction coefficient and the circumferential friction coefficient.
[0072] Referring to Figure 8, based on the friction coefficient and friction coefficient influencing parameters, the detailed process of training the friction coefficient intelligent analysis model is as follows:
[0073] Step 801, taking formation parameters, drill string performance parameters, drilling fluid parameters and drilling operation parameters as inputs of the first sub-neural network;
[0074] Step 802, using the cuttings bed parameters and the wellbore trajectory parameters as inputs to the second sub-neural network;
[0075] Step 803: Using the friction coefficient as an output label for training the friction coefficient intelligent analysis model;
[0076] Step 804: training a friction coefficient intelligent analysis model.
[0077] In one embodiment of the present invention, formation parameters (formation lithology, formation elastic coefficient, formation compressive strength), drill string performance parameters (drill string stiffness, drill string wall thickness, drill string outer diameter), drilling fluid parameters (drilling fluid type, density, viscosity, static shear force, solid content), and drilling operation parameters (drill string speed, drilling fluid displacement and flow rate, riser pressure, drill tool assembly, drilling method) are used as inputs of the first sub-neural network; cuttings bed parameters (wellbore curvature, well inclination, azimuth) and wellbore trajectory parameters (wellbore size, cuttings bed height, annular cuttings concentration, drill string eccentricity) are used as inputs of the second sub-neural network; and friction coefficients, including axial friction coefficient and circumferential friction coefficient, are used as output labels for training the friction coefficient intelligent analysis model to train the friction coefficient intelligent analysis model.
[0078] In specific implementation, in order to prevent the friction coefficient intelligent analysis model from overfitting, the L2 regularization method is adopted. That is, a regularization term is added to the loss function of the friction coefficient intelligent analysis model to achieve weight attenuation. The process is as follows:
[0079] First, add a regularization term to the loss function of the friction coefficient intelligent analysis model:
[0080] Secondly, the weights are derived according to the loss function of the friction coefficient intelligent analysis model:
[0081] Finally, error backpropagation is performed:
[0082] Where C0 is the loss function of the friction coefficient intelligent analysis model; λ is the regularization parameter; n is the number of samples for training the friction coefficient intelligent analysis model; is the Euclidean norm or L2 norm of the friction coefficient intelligent analysis model's weight parameters; C is the regularized loss function of the friction coefficient intelligent analysis model; η is the learning rate, also known as the step size, for training the friction coefficient intelligent analysis model; w is the weight parameter matrix of the friction coefficient intelligent analysis model; and w' is the weight parameter matrix of the friction coefficient intelligent analysis model obtained after error backpropagation. L2 regularization allows the weights to decay to smaller values, reducing the overfitting problem of the friction coefficient intelligent analysis model to a certain extent.
[0083] Referring to FIG9 , based on the friction coefficient and friction coefficient influencing parameters, the detailed process of training the friction coefficient intelligent analysis model also includes:
[0084] Step 901: obtaining a preset first weight and a preset second weight according to the drilling curvature;
[0085] Step 902: Add a preset first weight to the output of the first sub-neural network, and add a preset second weight to the output of the second sub-neural network;
[0086] Step 903 : Training a friction coefficient intelligent analysis model based on the friction coefficient, the friction coefficient influencing parameter, the preset first weight, and the preset second weight.
[0087] In one embodiment of the present invention, referring to Table 1, a preset first weight and a preset second weight are obtained according to the drilling curvature. First, the position of the drill string is judged according to the drilling curvature, and the conventional friction coefficient weight (i.e., the preset first weight) and the additional friction coefficient weight (i.e., the preset second weight) are determined, wherein the position includes the vertical well section, the inclined well section and the horizontal section; since the vertical well section does not have a cuttings bed and the curvature is 0 by default, the preset first weight K1 of the first sub-neural network is 1, and the preset second weight K2 of the second sub-neural network is 0 in the vertical well section; in the inclined well section, the preset first weight and the preset second weight are both 0.5. During the training of the intelligent analysis model of the friction coefficient, the two weights can be continuously adjusted to minimize the model error as much as possible; in the horizontal section, when the cuttings accumulation is more serious, only a small area of the drill string directly contacts the well wall, so the preset first weight K1 of the first sub-neural network is set to 0.2, and the preset second weight K2 of the second sub-neural network is set to 0.8, and they are adjusted according to the training model error.
[0088] Table 1 Friction coefficient weight comparison table
[0089] In one embodiment of the present invention, for step 105, a conventional friction coefficient and an additional friction coefficient are determined based on an intelligent friction coefficient analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the wellbore wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
[0090] During specific implementation, referring to FIG7 , formation parameters, drill string performance parameters, drilling fluid parameters, and drilling operation parameters are input into the first sub-neural network; cuttings bed parameters and wellbore trajectory parameters are input into the second sub-neural network to determine the conventional friction coefficient caused by the wellbore wall and the additional friction coefficient caused by the coupling between the cuttings bed and the wellbore trajectory.
[0091] In one embodiment of the present invention, after determining the conventional and additional friction coefficients based on the friction coefficient intelligent analysis model, the method further includes determining the relationship between the conventional, additional, and friction coefficients. If the friction coefficient is found to be excessive, it indicates that downhole pressure support, pipe sticking, and other phenomena may occur. Adjustments can be made by adjusting parameters such as the wellbore trajectory, drilling fluid flow rate, solids content, and drill string speed. These adjusted parameters are then input into the friction coefficient intelligent analysis model to obtain an optimized friction coefficient, allowing analysis of the friction reduction effect and trend.
[0092] In specific implementation, the position of the drill string is determined according to the drilling curvature. Referring to Figure 7, the conventional friction coefficient μw and additional friction coefficient μ cb By adding the preset first weight K1 and the preset second weight K2, the relationship between the conventional friction coefficient, the additional friction coefficient and the friction coefficient can be determined. The formula is: μ s =μ w K1+μ cb K2
[0093] Among them, μ s is the friction coefficient; μ w is the conventional friction coefficient caused by the well wall; μ cb is the additional friction coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory; K1 and K2 are the weights of the conventional friction coefficient and the additional friction coefficient, that is, the preset first weight and the preset second weight, which have different values at different positions; the above formula can be used to calculate both the axial friction coefficient and the circumferential friction coefficient. For example, when calculating the axial friction coefficient, μ w is the conventional axial friction coefficient caused by the wellbore wall, μ cb is the additional axial friction coefficient caused by the cuttings bed and the wellbore trajectory.
[0094] FIG10 is a flow chart of training the intelligent analysis model of friction coefficient.
[0095] In one embodiment of the present invention, referring to FIG10 , firstly, a theoretical model for friction calculation is analyzed and optimized, and a soft rod model is selected based on timeliness and accuracy requirements; secondly, a bottom hole WOB torque prediction model is established, and a gated recurrent unit recurrent neural network is selected during the establishment process, and the bottom hole WOB torque prediction model is trained using a preset drilling data set; the bottom hole WOB and torque of the current drilling are determined based on the bottom hole WOB torque prediction model, and the friction coefficient is determined by inversion using the soft rod model; finally, parameters influencing the friction coefficient are obtained, including formation parameters, drill string performance parameters, drilling fluid parameters, drilling construction parameters, cuttings bed parameters, and wellbore trajectory parameters, and a parallel BP neural network is trained based on the friction coefficient and the friction coefficient influencing parameters to obtain an intelligent analysis model for the friction coefficient that takes into account the coupling effect of the cuttings bed parameters and the wellbore trajectory.
[0096] It should be noted that although the operations of the method of the present invention are described in a specific order in the above embodiments and drawings, this does not require or imply that these operations must be performed in this specific order, or that all illustrated operations must be performed to achieve the desired results. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.
[0097] The implementation of the drill string friction analysis device can be referenced to the implementation of the aforementioned method, and any repetitive details will not be repeated. The terms "module" or "unit" used below may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented using software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.
[0098] Based on the same inventive concept, the present invention also proposes a drill string friction analysis device, as shown in FIG11 , which includes:
[0099] Theoretical model acquisition module 1101 is used to obtain a theoretical model for friction calculation;
[0100] A weight-on-bit torque determination module 1102 is configured to determine the current bottomhole weight-on-bit and bottomhole torque based on a bottomhole weight-on-bit torque prediction model; the bottomhole weight-on-bit torque prediction model is first established using a neural network and then trained based on a preset drilling data set;
[0101] The friction coefficient determination module 1103 is used to determine the friction coefficient based on the bottom hole pressure and bottom hole torque of the current drilling using a friction calculation theoretical model;
[0102] The analysis model training module 1104 is used to obtain friction coefficient influencing parameters and train a friction coefficient intelligent analysis model established using a neural network based on the friction coefficient and the friction coefficient influencing parameters. The friction coefficient intelligent analysis model is used to analyze the influence of the coupling between the cuttings bed and the wellbore trajectory on the friction coefficient.
[0103] The friction coefficient analysis module 1105 is used to determine the conventional friction coefficient and the additional friction coefficient based on the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
[0104] In one embodiment of the present invention, the bottom hole weight-on-bit torque prediction model is established using a gated recurrent unit recurrent neural network.
[0105] In one embodiment of the present invention, the preset drilling data set includes: an input data set and an output data set;
[0106] The input data set includes one or any combination of surface hook load, surface torque, drill string outer diameter, average well inclination, drill string linear weight, drilling fluid density, well inclination difference between two ends of the drill string, and azimuth difference between two ends of the drill string;
[0107] The output data sets include bottomhole weight on bit and bottomhole torque.
[0108] In one embodiment of the present invention, the weight-on-bit torque determination module 1102 is specifically configured to:
[0109] Inputting the current drilling input data set into the bottom hole weight-on-bit torque prediction model;
[0110] Determine the bottom hole weight on bit and bottom hole torque for the current drilling.
[0111] In one embodiment of the present invention, the friction coefficient determination module 1103 is specifically configured to:
[0112] Set the initial friction coefficient and determine it as the current friction coefficient;
[0113] Repeat the following steps until the difference between the calculated ground hook load and the actual ground hook load is less than or equal to the first preset error, or the difference between the calculated ground torque and the actual ground torque is less than or equal to the second preset error, and output the current friction coefficient as the determined friction coefficient:
[0114] According to the current drilling bottom hole pressure, bottom hole torque and current friction coefficient, the surface hook load and surface torque are calculated using the friction calculation theoretical model;
[0115] When the difference between the calculated ground hook load and the actual ground hook load is greater than the first preset error or the difference between the calculated ground torque and the actual ground torque is greater than the second preset error, the friction coefficient is modified and the modified friction coefficient is used as the current friction coefficient.
[0116] In one embodiment of the present invention, the analysis model training module 1104 is specifically used to:
[0117] Obtaining friction coefficient influencing parameters; wherein the friction coefficient influencing parameters include: formation parameters, drill string performance parameters, drilling fluid parameters, drilling operation parameters, cuttings bed parameters and wellbore trajectory parameters;
[0118] The first sub-neural network and the second sub-neural network are connected in parallel to establish an intelligent analysis model of the friction coefficient; wherein the first sub-neural network and the second sub-neural network are both preset BP neural networks;
[0119] According to the friction coefficient and the parameters affecting the friction coefficient, the friction coefficient intelligent analysis model is trained.
[0120] In one embodiment of the present invention, the analysis model training module 1104 is specifically used to:
[0121] Taking formation parameters, drill string performance parameters, drilling fluid parameters and drilling operation parameters as inputs of the first sub-neural network;
[0122] Cuttings bed parameters and wellbore trajectory parameters are used as inputs to the second sub-neural network;
[0123] Use the friction coefficient as the output label for training the friction coefficient intelligent analysis model;
[0124] Train the intelligent analysis model of friction coefficient.
[0125] In one embodiment of the present invention, the analysis model training module 1104 is further configured to:
[0126] Obtaining a preset first weight and a preset second weight according to the drilling curvature;
[0127] Adding a preset first weight to the output of the first sub-neural network, and adding a preset second weight to the output of the second sub-neural network;
[0128] According to the friction coefficient and the friction coefficient influencing parameters, the preset first weight and the preset second weight, the friction coefficient intelligent analysis model is trained.
[0129] In one embodiment of the present invention, the friction coefficient analysis module 1105 is further configured to:
[0130] Determine the relationship between conventional friction coefficient, additional friction coefficient and friction coefficient.
[0131] It should be noted that while the detailed description above mentions several modules of the drill string friction analysis device, this division is merely exemplary and not mandatory. In practice, depending on embodiments of the present invention, the features and functions of two or more modules described above may be embodied in a single module. Conversely, the features and functions of a single module described above may be further divided and embodied by multiple modules.
[0132] Based on the aforementioned inventive concept, as shown in FIG12 , the present invention further proposes a computer device 1200, comprising a memory 1210, a processor 1220, and a computer program 1230 stored in the memory 1210 and executable on the processor 1220. When the processor 1220 executes the computer program 1230, the aforementioned drill string friction analysis method is implemented.
[0133] Based on the aforementioned inventive concept, the present invention proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the aforementioned drill string friction analysis method is implemented.
[0134] The drill string friction analysis method and device proposed in the embodiment of the present invention can solve the problem that the existing technology uses the method of establishing a soft rod model or a rigid rod model to calculate the friction coefficient, which is not fully considered and leads to low friction calculation accuracy; by obtaining a friction calculation theoretical model; determining the bottom hole drilling pressure and bottom hole torque of the current drilling according to the bottom hole drilling pressure and bottom hole torque prediction model; the bottom hole drilling pressure and bottom hole torque prediction model is first established using a neural network and then obtained after training based on a preset drilling data set; determining the friction coefficient based on the bottom hole drilling pressure and bottom hole torque of the current drilling using the friction calculation theoretical model; obtaining friction coefficient influencing parameters, and training a friction coefficient intelligent analysis model established by using a neural network based on the friction coefficient and the friction coefficient influencing parameters; wherein the friction coefficient intelligent analysis model is used to analyze the influence of the coupling effect of the cuttings bed and the wellbore trajectory on the friction coefficient; and determining the conventional friction coefficient and the additional friction coefficient based on the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory. The embodiments of the present invention take into account the influence of the coupling effect of the cuttings bed and the wellbore trajectory on the frictional resistance, solve the problem that the influence of the coupling effect of the cuttings bed and the wellbore trajectory on the frictional resistance is difficult to quantify and calculate, and realize the quantitative analysis of the frictional resistance coefficient caused by the coupling effect of the cuttings bed and the wellbore trajectory, which helps to improve the accuracy of frictional resistance calculation, guide the staff to clean the wellbore and adjust the parameters in a timely manner, and improve the drilling efficiency.
[0135] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, apparatus, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0136] The present invention is described with reference to the flowcharts and / or block diagrams of the methods and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowcharts and / or one or more blocks in the block diagrams.
[0137] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.
[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.
[0139] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed by the present invention, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.
Claims
1. A drill string friction analysis method, It is characterized in that include: Obtain the theoretical model for friction calculation; Determine the bottom hole drilling pressure and bottom hole torque of the current drilling according to the bottom hole drilling pressure torque prediction model; the bottom hole drilling pressure torque prediction model is first established by using a neural network and then obtained after training according to a preset drilling data set; According to the current drilling bottom hole pressure and bottom hole torque, the friction coefficient is determined by using the friction calculation theoretical model; Obtaining friction coefficient influencing parameters, and training a friction coefficient intelligent analysis model established by using a neural network according to the friction coefficient and the friction coefficient influencing parameters; wherein the friction coefficient intelligent analysis model is used to analyze the influence of the coupling effect between the cuttings bed and the wellbore trajectory on the friction coefficient; According to the intelligent analysis model of friction coefficient, the conventional friction coefficient and the additional friction coefficient are determined; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
2. The method according to claim 1, It is characterized in that The bottom hole pressure torque prediction model is established by using a gated recurrent unit recurrent neural network.
3. The method according to claim 1, It is characterized in that The preset drilling data set includes: an input data set and an output data set; The input data set includes one or any combination of surface hook load, surface torque, drill string outer diameter, average well inclination, drill string line weight, drilling fluid density, well inclination difference between two ends of the drill string, and azimuth difference between two ends of the drill string; The output data sets include bottom hole weight on bit and bottom hole torque.
4. The method according to claim 3, It is characterized in that According to the bottom hole drilling pressure torque prediction model, the bottom hole drilling pressure and bottom hole torque of the current drilling are determined, including: Inputting the current drilling input data set into the bottom hole weight on bit torque prediction model; Determine the bottom hole weight on bit and bottom hole torque of the current drilling.
5. The method according to claim 1, It is characterized in that According to the current drilling bottom hole pressure and bottom hole torque, the friction coefficient is determined using the friction calculation theoretical model, including: Set the initial friction coefficient and determine it as the current friction coefficient; Repeat the following steps until the difference between the calculated ground hook load and the actual ground hook load is less than or equal to the first preset error, or the difference between the calculated ground torque and the actual ground torque is less than or equal to the second preset error, and output the current friction coefficient as the determined friction coefficient: According to the current drilling bottom hole pressure, bottom hole torque and current friction coefficient, the surface hook load and surface torque are calculated using the friction calculation theoretical model; When the difference between the calculated ground hook load and the actual ground hook load is greater than the first preset error or the difference between the calculated ground torque and the actual ground torque is greater than the second preset error, the friction coefficient is modified and the modified friction coefficient is used as the current friction coefficient.
6. The method according to claim 1, It is characterized in that Obtain the friction coefficient influencing parameters, and train the friction coefficient intelligent analysis model established by using a neural network according to the friction coefficient and the friction coefficient influencing parameters, including: Obtaining friction coefficient influencing parameters; wherein the friction coefficient influencing parameters include: formation parameters, drill string performance parameters, drilling fluid parameters, drilling construction parameters, cuttings bed parameters and wellbore trajectory parameters; The first sub-neural network and the second sub-neural network are connected in parallel to establish an intelligent analysis model of friction coefficient; wherein the first sub-neural network and the second sub-neural network are both preset BP neural networks; According to the friction coefficient and the parameters affecting the friction coefficient, the intelligent analysis model of the friction coefficient is trained.
7. The method according to claim 6, It is characterized in that According to the friction coefficient and the parameters affecting the friction coefficient, the friction coefficient intelligent analysis model is trained, including: Taking formation parameters, drill string performance parameters, drilling fluid parameters and drilling operation parameters as inputs of the first sub-neural network; The cuttings bed parameters and the wellbore trajectory parameters are used as inputs of the second sub-neural network; Use the friction coefficient as the output label for training the friction coefficient intelligent analysis model; Train the intelligent analysis model of friction coefficient.
8. The method according to claim 6, It is characterized in that According to the friction coefficient and the parameters affecting the friction coefficient, the friction coefficient intelligent analysis model is trained, which also includes: Obtaining a preset first weight and a preset second weight according to the drilling curvature; Adding a preset first weight to the output of the first sub-neural network, and adding a preset second weight to the output of the second sub-neural network; According to the friction coefficient and the friction coefficient influencing parameters, the preset first weight and the preset second weight, the friction coefficient intelligent analysis model is trained.
9. The method according to claim 1, It is characterized in that After determining the conventional friction coefficient and the additional friction coefficient according to the friction coefficient intelligent analysis model, it also includes: Determine the relationship between conventional friction coefficient, additional friction coefficient and friction coefficient.
10. A drill string friction analysis device, It is characterized in that include: Theoretical model acquisition module, used to obtain the friction calculation theoretical model; A drilling pressure torque determination module is used to determine the bottom hole drilling pressure and bottom hole torque of the current drilling according to a bottom hole drilling pressure torque prediction model; the bottom hole drilling pressure torque prediction model is first established using a neural network and then obtained after training based on a preset drilling data set; The friction coefficient determination module is used to determine the friction coefficient using a friction calculation theoretical model according to the bottom hole drilling pressure and bottom hole torque of the current drilling; The analysis model training module is used to obtain the friction coefficient influencing parameters, and train the friction coefficient intelligent analysis model established by using a neural network according to the friction coefficient and the friction coefficient influencing parameters; wherein the friction coefficient intelligent analysis model is used to analyze the influence of the coupling effect between the cuttings bed and the wellbore trajectory on the friction coefficient; The friction coefficient analysis module is used to determine the conventional friction coefficient and the additional friction coefficient according to the friction coefficient intelligent analysis model; wherein the conventional friction coefficient is the friction coefficient caused by the well wall, and the additional friction coefficient is the friction coefficient caused by the coupling effect between the cuttings bed and the wellbore trajectory.
11. The device according to claim 10, It is characterized in that The bottom hole pressure torque prediction model is established by using a gated recurrent unit recurrent neural network.
12. The device according to claim 10, It is characterized in that The preset drilling data set includes: an input data set and an output data set; The input data set includes one or any combination of surface hook load, surface torque, drill string outer diameter, average well inclination, drill string line weight, drilling fluid density, well inclination difference between two ends of the drill string, and azimuth difference between two ends of the drill string; The output data sets include bottom hole weight on bit and bottom hole torque.
13. The device according to claim 12, It is characterized in that The drilling pressure torque determination module is specifically used for: Inputting the current drilling input data set into the bottom hole weight on bit torque prediction model; Determine the bottom hole weight on bit and bottom hole torque of the current drilling.
14. The device according to claim 10, It is characterized in that The friction coefficient determination module is specifically used for: Set the initial friction coefficient and determine it as the current friction coefficient; Repeat the following steps until the difference between the calculated ground hook load and the actual ground hook load is less than or equal to the first preset error, or the difference between the calculated ground torque and the actual ground torque is less than or equal to the second preset error, and output the current friction coefficient as the determined friction coefficient: According to the current drilling bottom hole pressure, bottom hole torque and current friction coefficient, the surface hook load and surface torque are calculated using the friction calculation theoretical model; When the difference between the calculated ground hook load and the actual ground hook load is greater than the first preset error or the difference between the calculated ground torque and the actual ground torque is greater than the second preset error, the friction coefficient is modified and the modified friction coefficient is used as the current friction coefficient.
15. The device according to claim 10, It is characterized in that The analysis model training module is specifically used for: Obtaining friction coefficient influencing parameters; wherein the friction coefficient influencing parameters include: formation parameters, drill string performance parameters, drilling fluid parameters, drilling construction parameters, cuttings bed parameters and wellbore trajectory parameters; The first sub-neural network and the second sub-neural network are connected in parallel to establish an intelligent analysis model of friction coefficient; wherein the first sub-neural network and the second sub-neural network are both preset BP neural networks; According to the friction coefficient and the parameters affecting the friction coefficient, the intelligent analysis model of the friction coefficient is trained.
16. The device according to claim 15, It is characterized in that The analysis model training module is specifically used for: Taking formation parameters, drill string performance parameters, drilling fluid parameters and drilling operation parameters as inputs of the first sub-neural network; The cuttings bed parameters and the wellbore trajectory parameters are used as inputs of the second sub-neural network; Use the friction coefficient as the output label for training the friction coefficient intelligent analysis model; Train the intelligent analysis model of friction coefficient.
17. The device according to claim 15, It is characterized in that The analytical model training module is also used to: Obtaining a preset first weight and a preset second weight according to the drilling curvature; Adding a preset first weight to the output of the first sub-neural network, and adding a preset second weight to the output of the second sub-neural network; According to the friction coefficient and the friction coefficient influencing parameters, the preset first weight and the preset second weight, the friction coefficient intelligent analysis model is trained.
18. The device according to claim 10, It is characterized in that The friction coefficient analysis module is also used for: Determine the relationship between conventional friction coefficient, additional friction coefficient and friction coefficient.
19. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.
20. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.
Citation Information
Patent Citations
Frictional resistance prediction method and device
CN109973072A
Method for calculating trip-in friction resistance of horizontal well drilled and polished pipe column
CN110107224A
Method, device and equipment for determining friction resistance
CN111734396A
Method and device for analyzing friction resistance of drill string
CN116362143A
Calibration of drillstring weight for friction factor estimation
US20230059507A1
Cited By
Intelligent drilling guiding method and system for petroleum drilling tool face
CN120119963A
Wall surface blowing porous arrangement design method for boundary layer resistance reduction
CN120257495A
A Design Method for Porous Arrangement of Wall Blowing for Boundary Layer Drag Reduction
CN120257495B
Flow velocity field modeling method and device and readable storage medium thereof
CN120493810A
Well hole friction coefficient segmented inversion method based on genetic algorithm and tubular column mechanics
CN120611637A
