A mechanical drilling speed prediction method, device, storage medium and equipment

Abstract

The present invention discloses a mechanical drilling rate prediction method, device, storage medium and equipment, which obtains characteristic data and theoretical mechanical drilling rate of a target wellbore, processes the characteristic data based on a pre-built machine learning model, obtains the drilling rate ratio of the target wellbore, and obtains the predicted value of the mechanical drilling rate based on the theoretical mechanical drilling rate and the drilling rate ratio. The method can effectively solve the problem of the difficulty of drilling rate prediction by integrating the theoretical mechanical drilling rate and the machine learning model, and at the same time improves the accuracy of mechanical drilling rate prediction, which is conducive to providing a reliable basis for decision makers and realizing real-time optimization of drilling operations.

Description

Technical Field

[0001] The present invention relates to the field of exploration technology, and in particular to a mechanical drilling rate prediction method, device, storage medium and equipment. Background Art

[0002] As the oilfield exploration and development field is expanding from conventional oil and gas resources to unconventional oil and gas resources such as low permeability, deep, ultra-deep, deep sea, shale oil, shale gas, etc., it faces challenges such as resource deterioration, exploration diversification, development complexity, and environmental harshness. Drilling is a key link in the discovery, exploration and exploitation of oil and gas resources. As the geological environment encountered is becoming more and more complex and the drilling conditions are becoming more and more harsh, it is urgent to develop a new generation of revolutionary drilling technology. Intelligent drilling is a major development trend in the field of drilling engineering in the future. Intelligent drilling technology is a revolutionary drilling technology that integrates theories and technologies such as big data, artificial intelligence, information engineering, and downhole control engineering. Through the application of ground automated drilling rigs, downhole intelligent actuators, intelligent monitoring and decision-making technologies, etc., it can realize advanced detection, closed-loop control, precise guidance and intelligent decision-making of drilling operations, which can greatly improve drilling efficiency and reservoir drilling rate, significantly increase the single well production and recovery rate of complex oil and gas reservoirs, and reduce drilling costs.

[0003] Mechanical drilling rate prediction is one of the important research directions in the field of intelligent drilling. As an important technical indicator of drilling operations, accurate prediction of mechanical drilling rate can scientifically guide drilling parameter design and optimization, drilling speed-up tool selection, drilling cycle and cost prediction, and assist in optimizing wellbore trajectory and wellbore structure. Among the existing methods, a theoretical drilling rate prediction method based on the classic mechanical specific energy model is proposed. This method assumes that the rock breaking of the drill bit is an energy balance process during drilling. The drilling rate is calculated by detecting parameters such as mechanical specific energy, drilling pressure, rotation speed, and torque. However, the drilling process is a dynamic process, and there are many factors that affect the drilling rate. Any slight change in any parameter will affect the mechanical drilling rate. This method considers limited parameters and has a low model dimension. The nonlinear mapping ability is weak and it cannot accurately predict the dynamically changing drilling rate. Summary of the invention

[0004] The technical problem to be solved by the present invention is: how to solve the problem of great difficulty and low precision in drilling speed prediction, so as to achieve accurate prediction of mechanical drilling speed.

[0005] In order to solve the above technical problems, the present invention provides a method for predicting mechanical drilling speed, which comprises:

[0006] Acquiring characteristic data and theoretical mechanical drilling rate of the target wellbore, wherein the characteristic data includes formation property parameters, drilling parameters, fluid parameters and drill bit parameters;

[0007] Processing the characteristic data based on a pre-built machine learning model to obtain a drilling speed ratio value of the target wellbore;

[0008] A predicted value of the mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and the drilling speed ratio.

[0009] In some embodiments, the machine learning model is constructed by the following steps:

[0010] Acquire multiple groups of characteristic data of the target wellbore as training sample sets;

[0011] Based on each set of sample data in the training sample set, training the machine learning model;

[0012] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, the trained machine learning model is obtained.

[0013] In some embodiments, the formation property parameters include: layer position, lithology, gamma value, acoustic time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; the drilling parameters include: hook load, drilling pressure, rotation speed, displacement and pump pressure; the fluid parameters include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force; the drill bit parameters include drill bit model, drill bit wear and drill bit diameter.

[0014] In some embodiments, obtaining the theoretical mechanical penetration rate includes:

[0015] Obtaining drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore;

[0016] The theoretical mechanical drilling speed is determined by using a mechanical specific energy model based on the drilling mechanical parameters, the screw action parameters, the impactor action parameters, the drill bit action parameters, the hydraulic action parameters and the mechanical specific energy.

[0017] In some embodiments, the mechanical specific energy model comprises:

[0018]

[0019] Among them, E MMF represents mechanical specific energy, and the drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotating disk speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM The impactor action parameters include the impactor speed T SF , is the impactor mass mFT and impact velocity v FT The drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; The hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0020] In some embodiments, obtaining the mechanical specific energy of the target wellbore includes:

[0021] Obtaining the compressive strength and rock breaking energy efficiency of the target wellbore;

[0022] The mechanical specific energy of the target wellbore is determined based on the compressive strength and the rock-breaking energy efficiency.

[0023] In some embodiments, the machine learning model is constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm, or an extreme gradient boosting algorithm.

[0024] Another aspect of the present invention provides a mechanical drilling rate prediction device, comprising:

[0025] A data acquisition module, wherein the data acquisition module is used to acquire characteristic data of a target wellbore and a theoretical mechanical drilling rate, wherein the characteristic data includes formation property parameters, drilling parameters, fluid parameters and drill bit parameters;

[0026] A ratio determination module, the ratio determination module is used to process the characteristic data based on a pre-built machine learning model to obtain a drilling speed ratio of the target wellbore;

[0027] A drilling speed prediction module is used to obtain a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio.

[0028] Another aspect of the present invention provides a storage medium, in which a computer program is stored. When the computer program is executed, the mechanical drilling rate prediction method as described above can be implemented.

[0029] Another aspect of the present invention provides a device, comprising: a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the mechanical drilling rate prediction method as described above can be implemented.

[0030] Compared with the prior art, one or more embodiments of the above scheme may have the following advantages or beneficial effects:

[0031] The mechanical drilling rate prediction method, device, storage medium and equipment provided by the present invention are applied to obtain the characteristic data and theoretical mechanical drilling rate of the target wellbore, process the characteristic data based on a pre-built machine learning model, obtain the drilling rate ratio of the target wellbore, and obtain the predicted value of the mechanical drilling rate based on the theoretical mechanical drilling rate and the drilling rate ratio. This method can effectively solve the problem of the difficulty of drilling rate prediction by integrating the theoretical mechanical drilling rate and the machine learning model, while improving the accuracy of mechanical drilling rate prediction, which is conducive to providing a reliable basis for decision makers and realizing real-time optimization of drilling operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The scope of the present disclosure may be better understood by reading the following detailed description of exemplary embodiments in conjunction with the accompanying drawings. The drawings included are:

[0033] Figure 1 A schematic flow chart of a mechanical drilling rate prediction method provided by an embodiment of the present invention is shown;

[0034] Figure 2 A schematic diagram of the process of training a machine learning model and predicting mechanical drilling speed provided by an embodiment of the present invention is shown;

[0035] Figure 3 A schematic diagram of predicting mechanical drilling speed based on a machine learning model and a theoretical mechanical drilling speed provided by an embodiment of the present invention is shown;

[0036] Figure 4 A schematic diagram of a mechanical drilling rate prediction method provided in the second embodiment of the present invention is shown;

[0037] Figure 5 A schematic diagram showing the comparison between the theoretical mechanical drilling speed determined based on the improved mechanical specific energy model and the measured mechanical drilling speed;

[0038] Figure 6 A schematic diagram showing the comparison between the mechanical drilling speed predicted by the method based on the deep fusion of theory and machine learning model and the measured mechanical drilling speed;

[0039] Figure 7 A schematic diagram of rock compressive strength provided by a specific embodiment of the present invention is shown;

[0040] Figure 8 A schematic diagram of mechanical drilling speed predicted by a method based on deep fusion of theory and machine learning model provided by a specific embodiment of the present invention is shown;

[0041] Fig. 9 A schematic diagram showing a comparison between the mechanical drilling rate after the drilling parameters are optimized and the measured mechanical drilling rate provided by a specific embodiment of the present invention;

[0042] Fig.10A schematic structural diagram of a mechanical drilling speed prediction device provided by an embodiment of the present invention is shown;

[0043] Fig.11 A schematic diagram of a device structure provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions and advantages of the present invention clearer, the implementation method of the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments, so that the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0045] Mechanical drilling rate prediction is one of the important research directions in the field of intelligent drilling. As an important technical indicator of drilling operations, accurate prediction of mechanical drilling rate can scientifically guide drilling parameter design and optimization, drilling speed-up tool selection, drilling cycle and cost prediction, and assist in optimizing wellbore trajectory and wellbore structure. Among the existing methods, a theoretical drilling rate prediction method based on the classical mechanical specific energy model is proposed. This method assumes that during drilling, rock breaking by the drill bit is an energy balance process. The drilling rate is calculated by detecting parameters such as mechanical specific energy, drilling pressure, rotation speed, and torque. The classical mechanical specific energy model can be expressed as follows:

[0046]

[0047] E MMF is the mechanical specific energy, WOB is the drilling pressure, N is the rotary speed, T is the drill bit torque, ROP is the mechanical drilling speed, A b is the cross-sectional area of ​​the drill bit.

[0048] However, the drilling process is a dynamic process, and there are many factors that affect the drilling speed. Any slight change in any parameter will affect the mechanical drilling speed. The existing method of predicting the drilling speed based on the classic mechanical specific energy model considers limited parameters and has a low model dimension. The nonlinear mapping ability is weak and cannot accurately predict the dynamically changing drilling speed.

[0049] In view of this, the present application provides a mechanical drilling speed prediction method, device, storage medium and equipment, which obtains the characteristic data and theoretical mechanical drilling speed of the target wellbore, processes the characteristic data based on a pre-built machine learning model, obtains the drilling speed ratio of the target wellbore, and obtains the predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio. This method can effectively solve the problem of the difficulty of drilling speed prediction by integrating the theoretical mechanical drilling speed and the machine learning model, and at the same time improves the accuracy of mechanical drilling speed prediction, which is conducive to providing a reliable basis for decision makers and realizing real-time optimization of drilling operations.

[0050] Embodiment 1

[0051] See also Figure 1 As shown, Figure 1 A schematic flow chart of a mechanical drilling rate prediction method provided by an embodiment of the present invention is shown, which includes:

[0052] Step S101: Acquire characteristic data and theoretical mechanical drilling rate of a target wellbore, wherein the characteristic data includes formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0053] Step S102: Process the feature data based on the pre-built machine learning model to obtain the drilling speed ratio value of the target wellbore.

[0054] Step S103: obtaining a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio.

[0055] In the embodiment of the present invention, step S101 may specifically include acquiring characteristic data of the target wellbore by adopting conventional technical means in the art.

[0056] Among them, the characteristic data may include formation property parameters, drilling parameters, fluid parameters and drill bit parameters; in some embodiments, the formation property parameters may include layer position, lithology, gamma value, acoustic wave time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; drilling parameters may include hook load, drilling pressure, rotation speed, displacement and pump pressure; fluid parameters may include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force, wherein, as a specific example, the first static shear force can be the static shear force of the fluid standing for 10 seconds during the drilling process, and the second static shear force can be the static shear force of the fluid standing for 10 minutes during the drilling process; drill bit parameters may include drill bit model, drill bit wear and drill bit diameter.

[0057] In some embodiments, obtaining the theoretical mechanical penetration rate may be:

[0058] Obtain the drilling pressure WOB, rotary table speed N, drill bit torque T, and drill bit cross-sectional area A during the drilling process b and mechanical specific energy E MMF ;

[0059] According to the parameters obtained above, the theoretical mechanical penetration rate ROP is determined using the classical mechanical specific energy model, where the classical mechanical specific energy model can be expressed by the following expression:

[0060]

[0061] In other embodiments, the theoretical mechanical penetration rate may also be determined based on an improved mechanical specific energy model to improve the accuracy of mechanical penetration rate prediction, and details may be found in the description of the following second embodiment.

[0062] In an embodiment of the present invention, the machine learning model can be constructed by the following steps:

[0063] Acquire multiple groups of characteristic data of target wellbore as training sample sets;

[0064] Based on each set of sample data in the training sample set, the machine learning model is trained;

[0065] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, a trained machine learning model is obtained.

[0066] Among them, the drilling speed ratio is the ratio of the measured drilling speed to the theoretical mechanical drilling speed calculated based on the classic mechanical specific energy model. The drilling speed ratio can combine the many factors that cause the difference between the theoretical mechanical drilling speed and the measured mechanical drilling speed into one value. By using the machine learning model to train the drilling speed ratio, the difference between the measured mechanical drilling speed and the theoretical mechanical drilling speed is gradually learned during the training process. Based on the trained machine learning model, combined with the theoretical mechanical drilling speed, the predicted mechanical drilling speed value that integrates the theoretical and machine learning models is obtained.

[0067] As a specific example, see Figure 2 As shown, Figure 2 A flow chart of machine learning model training and mechanical drilling rate prediction provided by an embodiment of the present invention is shown. During the drilling process, raw data is collected by utilizing wellhead and downhole sensors, and the required parameters are remotely transmitted to the human-computer interaction terminal as a data source. After the data source is cleaned, it is stored in the machine learning training database as a training set data source. The training set data source may include multiple groups of feature data. For each group of feature data, it can be read and preprocessed by a computer. The data preprocessing process may include clearing zero data to meet the training requirements of the machine learning model; feature engineering is used to extract feature data from the preprocessed data, and the extracted feature data is input into the machine learning model for model training, testing, analysis and other operations. Among them, whether a trained machine learning model is obtained can be determined by judging whether the drilling rate ratio output by the machine learning model meets the accuracy requirements. The accuracy can be determined by the following expression:

[0068] As an example, it can be determined whether a trained machine learning model is obtained by judging whether the accuracy of the drilling speed ratio is greater than a preset threshold. When the accuracy of the drilling speed ratio is less than or equal to the preset threshold, the parameters are adjusted, and the training of the machine learning model is continued until the accuracy of the drilling speed ratio output by the machine learning model is greater than the preset threshold, and a trained machine learning model is obtained. It should be noted that the preset threshold can be flexibly set according to user needs. As a specific example, the preset threshold can be 90%. Based on the trained machine learning model, the drilling data, i.e., the characteristic data of the target wellbore, is processed to obtain the drilling speed ratio prediction result of the target wellbore.

[0069] Also, see Figure 3 As shown, Figure 3 The schematic diagram of predicting the mechanical drilling speed based on the machine learning model and the theoretical mechanical drilling speed provided by the embodiment of the present invention is shown. In the embodiment of the present invention, the theoretical mechanical drilling speed can be calculated based on the classic mechanical specific energy model, and the predicted value of the mechanical drilling speed can be obtained based on the theoretical mechanical drilling speed and the drilling speed ratio, that is, the fused mechanical drilling speed can be calculated. Finally, the predicted value of the mechanical drilling speed can be sent to the terminal, etc., by means of data transmission, so as to further optimize the drilling parameters based on the predicted value of the mechanical drilling speed and achieve the effect of optimizing the drilling parameters in real time.

[0070] In some embodiments, the machine learning model can be constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm, or an extreme gradient boosting algorithm. As a preferred example, a machine learning model can be constructed based on an extreme gradient boosting algorithm XGboost to effectively prevent algorithm overfitting while ensuring the accuracy of the machine learning model.

[0071] In the embodiment of the present invention, step S103 may specifically be to obtain a predicted value of the mechanical drilling speed based on the product of the theoretical mechanical drilling speed and the drilling speed ratio.

[0072] The above is a mechanical drilling rate prediction method provided by an embodiment of the present invention, which obtains the characteristic data and theoretical mechanical drilling rate of the target wellbore, processes the characteristic data based on a pre-built machine learning model, obtains the drilling rate ratio of the target wellbore, and obtains the predicted value of the mechanical drilling rate based on the theoretical mechanical drilling rate and the drilling rate ratio. This method can effectively solve the problem of the difficulty of drilling rate prediction by integrating the theoretical mechanical drilling rate and the machine learning model, while improving the accuracy of mechanical drilling rate prediction, which is conducive to providing a reliable basis for decision makers and realizing real-time optimization of drilling operations.

[0073] In order to improve the prediction accuracy of the mechanical drilling speed, the accuracy of obtaining the theoretical mechanical drilling speed can also be improved. For details, please refer to the description in the second embodiment.

[0074] Embodiment 2

[0075] See also Figure 4 As shown, Figure 4 A schematic flow chart of a mechanical drilling rate prediction method provided in the second embodiment of the present invention is shown, which includes:

[0076] Step S401: Acquire characteristic data of a target wellbore, the characteristic data including formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0077] Step S402: Process the feature data based on the pre-built machine learning model to obtain the drilling speed ratio of the target wellbore.

[0078] Step S403: Obtain drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore.

[0079] Step S404: Based on the drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy, a mechanical specific energy model is used to determine the theoretical mechanical penetration rate.

[0080] Step S405: obtaining a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio.

[0081] In the embodiment of the present invention, step S401 may specifically include acquiring characteristic data of the target wellbore using conventional technical means in the art.

[0082] Among them, the characteristic data may include formation property parameters, drilling parameters, fluid parameters and drill bit parameters; in some embodiments, the formation property parameters may include layer position, lithology, gamma value, acoustic wave time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; drilling parameters may include hook load, drilling pressure, rotation speed, displacement and pump pressure; fluid parameters may include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force, wherein, as a specific example, the first static shear force can be the static shear force of the fluid standing for 10 seconds during the drilling process, and the second static shear force can be the static shear force of the fluid standing for 10 minutes during the drilling process; drill bit parameters may include drill bit model, drill bit wear and drill bit diameter.

[0083] In step S402, the data collected from the target wellbore by the sensor can be used as input data of the machine learning model after preprocessing and feature extraction. The feature data can be processed based on the pre-built machine learning model to obtain the drilling speed ratio of the target wellbore.

[0084] In an embodiment of the present invention, the machine learning model can be constructed by the following steps:

[0085] Acquire multiple groups of characteristic data of target wellbore as training sample sets;

[0086] Based on each set of sample data in the training sample set, the machine learning model is trained;

[0087] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, a trained machine learning model is obtained.

[0088] Among them, the drilling speed ratio is the ratio of the measured drilling speed to the mechanical drilling speed calculated based on the theoretical model. The drilling speed ratio can combine the many factors that cause the difference between the theoretical mechanical drilling speed and the measured mechanical drilling speed into one value. By using the machine learning model to train the drilling speed ratio, the difference between the measured mechanical drilling speed and the theoretical mechanical drilling speed is gradually learned during the training process. Based on the trained machine learning model, combined with the theoretical mechanical drilling speed, the predicted mechanical drilling speed value that integrates the theory and machine learning model is obtained.

[0089] In step S403, conventional technical means in the art may be used to obtain drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore. MMF Represents mechanical specific energy. Drilling mechanical parameters include WOB, threshold WOB bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0090] In some embodiments, the mechanical specific energy can be determined based on the compressive strength and the rock breaking energy efficiency:

[0091] Obtain the compressive strength CCS of the target wellbore and determine the minimum mechanical specific energy E based on the compressive strength CCS MHFmin ;

[0092] Get rock breaking energy efficiency EFF M , based on the minimum mechanical specific energy E MHFmin and rock breaking energy efficiency EFF M Determine the mechanical specific energy.

[0093] Among them, when drilling in strata of different lithology types, the minimum mechanical specific energy can be approximately equal to the compressive strength CCS of the rock, that is,

[0094] The rock breaking energy efficiency can be expressed as:

[0095] Furthermore, the mechanical specific energy can be expressed as: EFF M =CCS / E MHF (1)

[0096] In the embodiment of the present invention, a mechanical specific energy model can be constructed based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters and hydraulic parameters:

[0097]

[0098] Among them, drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0099] By comprehensively considering the rock breaking energies such as screw energy, hydraulic rock breaking energy, and impactor impact work, and introducing threshold drilling pressure and threshold torque parameters, the constructed model is more intuitive and has better stability. It can effectively expand the scope of application of the mechanical specific energy model and improve the accuracy of determining the mechanical drilling speed.

[0100] Step S404 may specifically be to determine the theoretical mechanical drilling speed based on the drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, and hydraulic action parameters according to the mechanical specific energy and the mechanical specific energy model.

[0101] Specifically, by combining equations (1) and (2), the expression of the theoretical mechanical speed can be obtained as follows:

[0102]

[0103] By integrating multiple parameters, the accuracy and stability of mechanical drilling rate prediction can be greatly improved, and the scope of application of mechanical drilling rate prediction methods can be expanded.

[0104] In the embodiment of the present invention, step S405 can be specifically to obtain the predicted value of the mechanical drilling speed based on the product of the theoretical mechanical drilling speed and the drilling speed ratio. By combining the theoretical mechanical drilling speed obtained based on the improved specific energy model and the drilling speed ratio obtained by the machine learning model, the application scope of the specific energy model is effectively expanded, the drilling speed prediction accuracy is improved, and the defects of the mechanical specific energy model, such as low dimension, weak nonlinear mapping ability, and weak detail reflection ability, can be effectively overcome.

[0105] The above is a mechanical drilling rate prediction method provided in the second embodiment of the present invention, which obtains the characteristic data of the target wellbore, processes the characteristic data based on the pre-built machine learning model, and obtains the drilling rate ratio of the target wellbore; at the same time, determines the theoretical mechanical drilling rate based on the improved mechanical specific energy model; and then obtains the predicted value of the mechanical drilling rate based on the theoretical mechanical drilling rate and the drilling rate ratio. This method can effectively expand the scope of application of the specific energy model and improve the accuracy of determining the theoretical mechanical drilling rate by obtaining the theoretical mechanical drilling rate based on the improved mechanical specific energy model, and further, when the theoretical mechanical drilling rate and the machine learning model are integrated, the accuracy of the mechanical drilling rate prediction can be effectively improved, which is beneficial to provide a reliable basis for decision makers and realize real-time optimization of drilling operations. Specific embodiments

[0107] As a specific example, the Shunbei 5-4H well is used for illustration:

[0108] (1) Data of more than 400 wells drilled in an oil field in western China were collected. The directly collected data included: formation, lithology, hook load, drilling pressure, rotation speed, displacement, pump pressure, drilling fluid density, static shear force (10s), static shear force (10min), filtration loss, total solid phase, funnel viscosity, drill bit model, drill bit wear, drill bit diameter, gamma value, and sonic time difference. At the same time, the directly collected data were used to calculate the indirect parameters: specific water power, jet impact force, elastic modulus, Poisson's ratio, cohesion, internal friction angle, and compressive strength. A training data set was constructed based on the above data.

[0109] (2) With a spacing of 1 meter, the theoretical mechanical drilling speed is calculated based on the improved mechanical specific energy model, and the drilling speed ratio is calculated based on the theoretical mechanical drilling speed and the measured mechanical drilling speed.

[0110] (3) Based on the constructed training data set, the machine learning model is trained. In some embodiments, the XGboost algorithm can be used to construct the machine learning model to accurately obtain the drilling speed prediction value under the fusion algorithm.

[0111] (4) Obtain the target wellbore’s stratigraphic position, lithology, hook load, drilling pressure, rotation speed, displacement, pump pressure, drilling fluid density, static shear force (10 s), static shear force (10 min), filtration loss, total solid content, funnel viscosity, drill bit model, drill bit wear, drill bit diameter, gamma value, sonic time difference, specific water power, jet impact force, elastic modulus, Poisson’s ratio, cohesion, internal friction angle, and compressive strength as characteristic data.

[0112] The theoretical mechanical drilling speed determined by the improved mechanical specific energy model has an accuracy of 86%, which can be found in Figure 5 As shown, Figure 5 A schematic diagram comparing the theoretical mechanical drilling speed determined based on the improved mechanical specific energy model and the measured mechanical drilling speed is shown.

[0113] (5) The deep integration of theory and machine learning model is used to predict the mechanical drilling speed, with an accuracy of 93%. Figure 6 As shown, Figure 6 A schematic diagram showing the comparison between the mechanical drilling speed predicted by the method based on the deep fusion of theory and machine learning models and the measured mechanical drilling speed. The predicted value of the mechanical drilling speed obtained after the fusion model is more accurate.

[0114] The mechanical drilling rate prediction method provided in the embodiment of the present invention combines theory and machine learning models to predict the mechanical drilling rate, thereby realizing the combination of dynamic physical models and artificial intelligence algorithms, which is conducive to improving the accuracy of prediction. In addition, by obtaining the mechanical drilling rate of the formation to be drilled from the real-time monitoring data of logging, the mechanical drilling rate of the lower formation can be predicted in advance, thereby scientifically guiding the design and optimization of drilling parameters, optimizing drilling speed-up tools, reducing drilling cycles and costs, and assisting in optimizing wellbore trajectories and wellbore structures. For details, please refer to the description of the following specific examples:

[0115] Shunbei 5-12H well is a horizontal development well located on the northern edge of the Shuntuoguole low uplift in Xinjiang. The well has a designed depth of 8321.88m and was drilled in the middle and lower Yingshan Formation of the Ordovician. The main purpose is to evaluate the reservoir development characteristics and oil and gas content of the lower section of the Yingshan Formation in the Shunbei No. 5 fault zone.

[0116] The compressive strength of rock is the most important factor affecting the mechanical drilling speed. Therefore, the logging data was used to calculate the compressive strength of the rock formation drilled in Shunbei 5-12H Well. The results are as follows: Figure 7 As shown, it was found that the compressive strength of the rock in the 4700-5300 meter section of the second well section varied greatly. The volcanic rocks and basalts of the Permian Aqia Group were encountered. The rocks are strong and abrasive. If no speed-up measures are taken, the mechanical drilling speed may be low, causing wear of the drill bit and increasing the number of trips and drilling.

[0117] The mechanical drilling speed of this well section is predicted based on the fusion of the improved mechanical specific energy model and the machine learning model. The mechanical drilling speed prediction results are as follows: Figure 8 As shown in the figure, the curve corresponding to the predicted mechanical drilling speed obtained by integrating the theoretical mechanical drilling speed based on the machine learning model is basically consistent with the curve corresponding to the measured mechanical drilling speed. The fusion prediction result is more accurate than the theoretical mechanical drilling speed obtained based on the improved mechanical specific energy model, which is improved from 81% to 92%. At the same time, it was observed that the highest mechanical drilling speed in this section was 15m / s, and the average mechanical drilling speed was 4m / hr. The mechanical drilling speed was lower at 4800-5200 meters, corresponding to Figure 7 The section with medium to high rock compressive strength. In view of the low mechanical drilling speed in this section, it is recommended to use "special-shaped tooth PDC drill bit + rotary impact screw + large-size drilling tool" to increase the speed when drilling in this section in subsequent wells. The rotary impact drilling technology can improve the drilling pressure transmission efficiency of hard formations. The hard formation rock is more prone to brittle fracture under the action of rotary impact, which can improve the stress state of the drill teeth, better protect the cutting teeth, and improve the drilling efficiency. The stroke of the rotary impact screw drilling tool is 3 to 13 mm, the impact frequency is 3 to 4 times the motor speed, the maximum displacement is 75L / S, and the operating life is 150h, which can meet the technical requirements of large displacement and long life. At the same time, according to the rock characteristics of the Acha Group volcanic rocks and basalts, it is recommended to match the high-efficiency PDC drill bit and design the drill bit with a deep inner cone structure to improve the stability during drilling and reduce the wear of the drill bit caused by lateral and axial vibrations in the well; at the same time, the back tilt angle is increased to improve the impact resistance of the drill bit. The main blade adopts a double-row cutting structure, and the inner cone and nose are designed with an annular shock-absorbing belt to further protect the drill bit and improve the impact resistance of the drill bit. Through in-depth research on market drill bits, it is recommended to use the Jiangzuan KS1952DGR type 5-blade high-efficiency PDC drill bit equipped with 19mm main cutting teeth and 16mm secondary cutting teeth composite pieces. The recommended composite drilling parameters are: drilling pressure 140KN, speed 50rpm+screw, displacement 45L / S. The parameters before and after optimization are shown in Table 1. Table 1 shows the comparison table before and after optimization of drilling parameters provided in the specific embodiment of the present invention. After optimization, the mechanical drilling speed in this section was significantly improved by 33.1%, such as Fig. 9 shown.

[0118] Actual parameters After optimization Drilling Tool Assembly Hybrid PDC drill bit + rotary table Special-shaped tooth PDC drill bit + rotary impact screw + large-size drill rod Drilling pressure 150KN 140KN Speed 50RPM 50RPM+screw density 1.25g / cm3 1.2g / cm3 Displacement 36L / S 45L / S

[0119] Table 1 Comparison of drilling parameters before and after optimization

[0120] Embodiment 3

[0121] See also Fig.10 As shown, Fig.10 A schematic structural diagram of a mechanical drilling speed prediction device is provided, which includes:

[0122] A data acquisition module 10 is used to acquire characteristic data and theoretical mechanical drilling rate of a target wellbore, wherein the characteristic data includes formation property parameters, drilling parameters, fluid parameters and drill bit parameters;

[0123] A ratio determination module 11, the ratio determination module 11 is used to process the characteristic data based on a pre-built machine learning model to obtain the drilling speed ratio of the target wellbore;

[0124] The drilling speed prediction module 12 is used to obtain a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio.

[0125] Among them, the characteristic data may include formation property parameters, drilling parameters, fluid parameters and drill bit parameters; in some embodiments, the formation property parameters may include layer position, lithology, gamma value, acoustic wave time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; drilling parameters may include hook load, drilling pressure, rotation speed, displacement and pump pressure; fluid parameters may include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force, wherein, as a specific example, the first static shear force can be the static shear force of the fluid standing for 10 seconds during the drilling process, and the second static shear force can be the static shear force of the fluid standing for 10 minutes during the drilling process; drill bit parameters may include drill bit model, drill bit wear and drill bit diameter.

[0126] In some embodiments, obtaining the theoretical mechanical penetration rate may be:

[0127] Obtain the drilling pressure WOB, rotary table speed N, drill bit torque T, and drill bit cross-sectional area A during the drilling process b and mechanical specific energy E MMF ;

[0128] According to the parameters obtained above, the theoretical mechanical penetration rate ROP is determined using the classical mechanical specific energy model, where the classical mechanical specific energy model can be expressed by the following expression:

[0129]

[0130] In other embodiments, the theoretical mechanical penetration rate may also be determined based on an improved mechanical specific energy model to improve the accuracy of mechanical penetration rate prediction.

[0131] In some embodiments, the theoretical mechanical penetration rate can be obtained based on the improved mechanical specific energy model:

[0132] Obtain drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore;

[0133] The theoretical mechanical penetration rate is determined by using a mechanical specific energy model based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy.

[0134] Among them, conventional technical means in this field can be used to obtain the drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore. MMF Represents mechanical specific energy. Drilling mechanical parameters include WOB, threshold WOB bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0135] In some embodiments, the mechanical specific energy can be determined based on the compressive strength and the rock breaking energy efficiency:

[0136] Obtain the compressive strength CCS of the target wellbore and determine the minimum mechanical specific energy E based on the compressive strength CCS MHFmin ;

[0137] Get rock breaking energy efficiency EFF M , based on the minimum mechanical specific energy E MHFmin and rock breaking energy efficiency EFF M Determine the mechanical specific energy.

[0138] Among them, when drilling in strata of different lithology types, the minimum mechanical specific energy can be approximately equal to the compressive strength CCS of the rock, that is,

[0139] The rock breaking energy efficiency can be expressed as:

[0140] Furthermore, the mechanical specific energy can be expressed as: EFF M =CCS / E MHF (1)

[0141] In the embodiment of the present invention, a mechanical specific energy model can be constructed based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters and hydraulic parameters:

[0142]

[0143] Among them, drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0144] By comprehensively considering rock breaking energies such as screw energy, hydraulic rock breaking energy, and impactor impact work, and introducing threshold drilling pressure and threshold torque parameters, the constructed model is more intuitive and has better stability. It can effectively expand the scope of application of the specific energy model and improve the accuracy of determining the drilling speed.

[0145] Based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters, the theoretical mechanical drilling speed is determined according to the mechanical specific energy and the mechanical specific energy model.

[0146] Specifically, by combining equations (1) and (2), the expression of the theoretical mechanical speed can be obtained as follows:

[0147]

[0148] By integrating multiple parameters, the accuracy and stability of mechanical drilling speed prediction can be greatly improved, and the scope of application of mechanical speed prediction methods can be expanded.

[0149] The ratio determination module 11 can use the data collected from the target wellbore by the sensor as input data of the machine learning model after preprocessing and feature extraction, and process the feature data based on the pre-built machine learning model to obtain the drilling speed ratio of the target wellbore.

[0150] In an embodiment of the present invention, the machine learning model can be constructed by the following steps:

[0151] Acquire multiple groups of characteristic data of target wellbore as training sample sets;

[0152] Based on each set of sample data in the training sample set, the machine learning model is trained;

[0153] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, a trained machine learning model is obtained.

[0154] Among them, the drilling speed ratio is the ratio of the measured drilling speed to the theoretical mechanical drilling speed calculated based on the theoretical model. The drilling speed ratio can combine the many factors that cause the difference between the theoretical mechanical drilling speed and the measured mechanical drilling speed into one value. By using the machine learning model to train the drilling speed ratio, the difference between the measured mechanical drilling speed and the theoretical mechanical drilling speed is gradually learned during the training process. Based on the trained machine learning model, combined with the theoretical mechanical drilling speed, the predicted mechanical drilling speed value that integrates the theory and machine learning model is obtained.

[0155] In some embodiments, the machine learning model can be constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm, or an extreme gradient boosting algorithm. As a preferred example, a machine learning model can be constructed based on an extreme gradient boosting algorithm XGboost to effectively prevent algorithm overfitting while ensuring the accuracy of the machine learning model.

[0156] The drilling speed prediction module 12 can obtain a predicted value of the mechanical drilling speed based on the product of the theoretical mechanical drilling speed and the drilling speed ratio.

[0157] The above is a mechanical drilling speed prediction device provided by an embodiment of the present invention. By acquiring the characteristic data and theoretical mechanical drilling speed of the target wellbore, the characteristic data is processed based on a pre-built machine learning model to obtain the drilling speed ratio of the target wellbore, and the predicted value of the mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and the drilling speed ratio. The device can effectively solve the problem of the difficulty of drilling speed prediction by integrating the theoretical mechanical drilling speed and the machine learning model, while improving the accuracy of mechanical drilling speed prediction, which is conducive to providing a reliable basis for decision makers and realizing real-time optimization of drilling operations.

[0158] Embodiment 4

[0159] In another aspect of the present application, a storage medium is provided, in which a computer program is stored. When the computer program is executed, the method for predicting the mechanical drilling rate of the first embodiment can be implemented:

[0160] The characteristic data and theoretical mechanical drilling rate of the target wellbore are obtained, and the characteristic data include formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0161] The feature data is processed based on the pre-built machine learning model to obtain the drilling speed ratio value of the target wellbore.

[0162] The predicted value of mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and drilling speed ratio.

[0163] Alternatively, when the computer program is executed, the method for predicting the mechanical drilling speed in the second embodiment can be implemented:

[0164] Acquire characteristic data of the target wellbore, the characteristic data including formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0165] The feature data is processed based on the pre-built machine learning model to obtain the drilling speed ratio value of the target wellbore.

[0166] Obtain the drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore.

[0167] The theoretical mechanical penetration rate is determined by using a mechanical specific energy model based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy.

[0168] The predicted value of mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and drilling speed ratio.

[0169] Among them, the characteristic data may include formation property parameters, drilling parameters, fluid parameters and drill bit parameters; in some embodiments, the formation property parameters may include layer position, lithology, gamma value, acoustic wave time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; drilling parameters may include hook load, drilling pressure, rotation speed, displacement and pump pressure; fluid parameters may include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force, wherein, as a specific example, the first static shear force can be the static shear force of the fluid standing for 10 seconds during the drilling process, and the second static shear force can be the static shear force of the fluid standing for 10 minutes during the drilling process; drill bit parameters may include drill bit model, drill bit wear and drill bit diameter.

[0170] In some embodiments, obtaining the theoretical mechanical penetration rate may be:

[0171] Obtain the drilling pressure WOB, rotary table speed N, drill bit torque T, and drill bit cross-sectional area A during the drilling process b and mechanical specific energy E MMF ;

[0172] According to the parameters obtained above, the theoretical mechanical penetration rate ROP is determined using the classical mechanical specific energy model, where the classical mechanical specific energy model can be expressed by the following expression:

[0173]

[0174] In other embodiments, obtaining the theoretical mechanical penetration rate may be:

[0175] Obtain the compressive strength CCS of the target wellbore and determine the minimum mechanical specific energy E based on the compressive strength CCS MHFmin ;

[0176] Get rock breaking energy efficiency EFF M , based on the minimum mechanical specific energy E MHFmin and rock breaking energy efficiency EFF M Determine the mechanical specific energy.

[0177] Among them, when drilling in strata of different lithology types, the minimum mechanical specific energy can be approximately equal to the compressive strength CCS of the rock, that is,

[0178] The rock breaking energy efficiency can be expressed as:

[0179] Furthermore, the mechanical specific energy can be expressed as: EFF M =CCS / E MHF (1)

[0180] In the embodiment of the present invention, a mechanical specific energy model can be constructed based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters and hydraulic parameters:

[0181]

[0182] Among them, drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0183] Based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters, the theoretical mechanical drilling speed is determined according to the mechanical specific energy and the mechanical specific energy model.

[0184] Specifically, by combining equations (1) and (2), the expression of the theoretical mechanical speed can be obtained as follows:

[0185]

[0186] By integrating multiple parameters, the accuracy and stability of mechanical drilling speed prediction can be greatly improved, and the scope of application of mechanical speed prediction methods can be expanded.

[0187] A machine learning model can be built by following these steps:

[0188] Acquire multiple groups of characteristic data of target wellbore as training sample sets;

[0189] Based on each set of sample data in the training sample set, the machine learning model is trained;

[0190] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, a trained machine learning model is obtained.

[0191] Among them, the drilling speed ratio is the ratio of the measured drilling speed to the theoretical mechanical drilling speed calculated based on the theoretical model. The drilling speed ratio can combine the many factors that cause the difference between the theoretical mechanical drilling speed and the measured mechanical drilling speed into one value. By using the machine learning model to train the drilling speed ratio, the difference between the measured mechanical drilling speed and the theoretical mechanical drilling speed is gradually learned during the training process. Based on the trained machine learning model, combined with the theoretical mechanical drilling speed, the mechanical drilling speed that integrates the theory and the machine learning model is obtained.

[0192] In some embodiments, the machine learning model can be constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm, or an extreme gradient boosting algorithm. As a preferred example, a machine learning model can be constructed based on an extreme gradient boosting algorithm XGboost to effectively prevent algorithm overfitting while ensuring the accuracy of the machine learning model.

[0193] The processes, functions, methods and / or software described above may be recorded, stored or fixed in one or more computer-readable storage media, which include program instructions, which will be implemented by the computer so that the processor executes the program instructions. The medium may also include program instructions, data files, data structures, etc. alone, or a combination thereof. The medium or program instructions may be specifically designed and understood by a person skilled in the art of computer software, or the medium or instructions may be known and available to a person skilled in the art of computer software. Examples of computer-readable media include: magnetic media, such as hard disks, floppy disks, and tapes; optical media, such as CDROM disks and DVDs; magneto-optical media, such as optical disks; and hardware devices, specifically configured to store and execute program instructions, such as read-only memory (ROM), random access memory (RAM), flash memory, etc. Examples of program instructions include machine code (e.g., code generated by a compiler) and files containing high-level code, which may be executed by a computer using an interpreter. The described hardware devices may be configured to act as one or more software modules to perform the operations and methods described above, and vice versa. In addition, the computer-readable storage medium may be distributed in networked computer systems, and computer-readable codes or program instructions may be stored and executed in a decentralized manner.

[0194] Embodiment 5

[0195] In another aspect of the present application, a device is provided, see Fig.11 As shown, Fig.11 A schematic diagram of a device structure provided by an embodiment of the present invention is shown, which includes: a memory 20 and a processor 21. The memory 20 stores a computer program. When the computer program is executed by the processor 21, the mechanical drilling speed prediction method of the first embodiment can be implemented:

[0196] The characteristic data and theoretical mechanical drilling rate of the target wellbore are obtained, and the characteristic data include formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0197] The feature data is processed based on the pre-built machine learning model to obtain the drilling speed ratio value of the target wellbore.

[0198] The predicted value of mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and drilling speed ratio.

[0199] Alternatively, when the computer program is executed, the method for predicting the mechanical drilling speed in the second embodiment can be implemented:

[0200] Acquire characteristic data of the target wellbore, the characteristic data including formation property parameters, drilling parameters, fluid parameters and drill bit parameters.

[0201] The feature data is processed based on the pre-built machine learning model to obtain the drilling speed ratio value of the target wellbore.

[0202] Obtain the drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore.

[0203] The theoretical mechanical penetration rate is determined by using a mechanical specific energy model based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy.

[0204] The predicted value of mechanical drilling speed is obtained based on the theoretical mechanical drilling speed and drilling speed ratio.

[0205] Among them, the characteristic data may include formation property parameters, drilling parameters, fluid parameters and drill bit parameters; in some embodiments, the formation property parameters may include layer position, lithology, gamma value, acoustic wave time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; drilling parameters may include hook load, drilling pressure, rotation speed, displacement and pump pressure; fluid parameters may include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force, wherein, as a specific example, the first static shear force can be the static shear force of the fluid standing for 10 seconds during the drilling process, and the second static shear force can be the static shear force of the fluid standing for 10 minutes during the drilling process; drill bit parameters may include drill bit model, drill bit wear and drill bit diameter.

[0206] In some embodiments, obtaining the theoretical mechanical penetration rate may be:

[0207] Obtain the drilling pressure WOB, rotary table speed N, drill bit torque T, and drill bit cross-sectional area A during the drilling process b and mechanical specific energy E MMF ;

[0208] According to the parameters obtained above, the theoretical mechanical penetration rate ROP is determined using the classical mechanical specific energy model, where the classical mechanical specific energy model can be expressed by the following expression:

[0209]

[0210] In other embodiments, obtaining the theoretical mechanical penetration rate may be:

[0211] Obtain the compressive strength CCS of the target wellbore and determine the minimum mechanical specific energy E based on the compressive strength CCS MHFmin ;

[0212] Get rock breaking energy efficiency EFF M , based on the minimum mechanical specific energy E MHFmin and rock breaking energy efficiency EFF MDetermine the mechanical specific energy.

[0213] Among them, when drilling in strata of different lithology types, the minimum mechanical specific energy can be approximately equal to the compressive strength CCS of the rock, that is,

[0214] The rock breaking energy efficiency can be expressed as:

[0215] Furthermore, the mechanical specific energy can be expressed as: EFF M =CCS / E MHF (1)

[0216] In the embodiment of the present invention, a mechanical specific energy model can be constructed based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters and hydraulic parameters:

[0217]

[0218] Among them, drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotary table speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM ; The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT ; Drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; Hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

[0219] Based on drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters, the theoretical mechanical drilling speed is determined according to the mechanical specific energy and the mechanical specific energy model.

[0220] Specifically, by combining equations (1) and (2), the expression of the theoretical mechanical speed can be obtained as follows:

[0221]

[0222] By integrating multiple parameters, the accuracy and stability of mechanical drilling speed prediction can be greatly improved, and the scope of application of mechanical speed prediction methods can be expanded.

[0223] A machine learning model can be built by following these steps:

[0224] Acquire multiple groups of characteristic data of target wellbore as training sample sets;

[0225] Based on each set of sample data in the training sample set, the machine learning model is trained;

[0226] When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, a trained machine learning model is obtained.

[0227] Among them, the drilling speed ratio is the ratio of the measured drilling speed to the theoretical mechanical drilling speed calculated based on the theoretical model. The drilling speed ratio can combine the many factors that cause the difference between the theoretical mechanical drilling speed and the measured mechanical drilling speed into one value. By using the machine learning model to train the drilling speed ratio, the difference between the measured mechanical drilling speed and the theoretical mechanical drilling speed is gradually learned during the training process. Based on the trained machine learning model, combined with the theoretical mechanical drilling speed, the mechanical drilling speed that integrates the theory and the machine learning model is obtained.

[0228] In some embodiments, the machine learning model can be constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm, or an extreme gradient boosting algorithm. As a preferred example, a machine learning model can be constructed based on an extreme gradient boosting algorithm XGboost to effectively prevent algorithm overfitting while ensuring the accuracy of the machine learning model.

[0229] It should be noted that the device may include one or more processors 21 and a memory 20, and the processor 21 and the memory 20 may be connected via a bus or other means. The memory 20, as a non-volatile computer-readable storage medium, may be used to store non-volatile software programs, non-volatile computer executable programs and modules. The processor 21 executes various functions and data processing by running the non-volatile software programs, instructions and modules stored in the memory 20, that is, realizing the above-mentioned mechanical drilling rate prediction method.

[0230] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A method for predicting mechanical drilling speed, characterized in that: include: Acquiring characteristic data and theoretical mechanical drilling rate of the target wellbore, wherein the characteristic data includes formation property parameters, drilling parameters, fluid parameters and drill bit parameters; Processing the characteristic data based on a pre-built machine learning model to obtain a drilling speed ratio value of the target wellbore; Obtaining a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio; Wherein, obtaining the theoretical mechanical drilling speed includes: Obtaining drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore; Determine the theoretical mechanical penetration rate using a mechanical specific energy model based on the drilling mechanical parameters, the screw action parameters, the impactor action parameters, the drill bit action parameters, the hydraulic action parameters and the mechanical specific energy; The mechanical specific energy model includes: Among them, E MMF represents mechanical specific energy, and the drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotating disk speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT The drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; The hydraulic action parameters include: drilling fluid displacement Q and water power utilization rate η.

2. The method according to claim 1, characterized in that The machine learning model is constructed by the following steps: Acquire multiple groups of characteristic data of the target wellbore as training sample sets; Based on each set of sample data in the training sample set, training the machine learning model; When the accuracy of the drilling speed ratio value output by the machine learning model is greater than a preset threshold, the trained machine learning model is obtained.

3. The method according to claim 1 or 2, characterized in that: The formation property parameters include: layer position, lithology, gamma value, acoustic time difference, elastic modulus, Poisson's ratio, cohesion, internal friction angle and compressive strength; the drilling parameters include: hook load, drilling pressure, rotation speed, displacement and pump pressure; the fluid parameters include: drilling fluid density, first static shear force, second static shear force, filtration loss, total solid phase content, funnel viscosity, specific water power and jet impact force; the drill bit parameters include drill bit model, drill bit wear and drill bit diameter.

4. The method according to claim 1, characterized in that Obtaining the mechanical specific energy of the target wellbore includes: Obtaining the compressive strength and rock breaking energy efficiency of the target wellbore; The mechanical specific energy of the target wellbore is determined based on the compressive strength and the rock-breaking energy efficiency.

5. The method according to claim 1, characterized in that The machine learning model is constructed based on a K-nearest neighbor algorithm, a support vector machine, a neural network algorithm, a random forest algorithm or an extreme gradient boosting algorithm.

6. A mechanical drilling speed prediction device, characterized in that: include: A data acquisition module, the data acquisition module is used to acquire characteristic data and theoretical mechanical penetration rate of a target wellbore, the characteristic data including formation property parameters, drilling parameters, fluid parameters and drill bit parameters; wherein, acquiring the theoretical mechanical penetration rate includes: acquiring drilling mechanical parameters, screw action parameters, impactor action parameters, drill bit action parameters, hydraulic action parameters and mechanical specific energy of the target wellbore; based on the drilling mechanical parameters, the screw action parameters, the impactor action parameters, the drill bit action parameters, the hydraulic action parameters and the mechanical specific energy, determining the theoretical mechanical penetration rate using a mechanical specific energy model; The mechanical specific energy model includes: Among them, E MMF represents mechanical specific energy, and the drilling mechanical parameters include WOB drilling pressure, threshold drilling pressure W bs , threshold torque T bs and the rotating disk speed N; the screw action parameters include the screw speed K N , screw rated displacement q PDM and screw pressure drop ΔP PDM The impactor action parameters include the impactor speed T SF , is the impactor mass m FT and impact velocity v FT The drill bit action parameters include: drill bit pressure drop ΔP b and drill cross-sectional area A b ; The hydraulic parameters include: drilling fluid displacement Q and water power utilization rate η; A ratio determination module, the ratio determination module is used to process the characteristic data based on a pre-built machine learning model to obtain a drilling speed ratio of the target wellbore; A drilling speed prediction module is used to obtain a predicted value of the mechanical drilling speed based on the theoretical mechanical drilling speed and the drilling speed ratio.

7. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed, the method for predicting the mechanical drilling speed as described in any one of claims 1 to 5 can be implemented.

8. An electronic device, characterized in that: include: A memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the mechanical drilling speed prediction method as described in any one of claims 1 to 5 can be implemented.

Images