Drill bit optimization methods, apparatuses, computing devices, and storage media

Abstract

This invention provides a method, apparatus, computing device, and storage medium for drill bit selection. The method includes: using at least two trained rate of drilling (RDR) prediction models, based on image data of candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, determining the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bits in the formation to be drilled, corresponding to each of the at least two trained RDR prediction models; determining the predicted MRD of the candidate drill bits in the formation to be drilled, based on the preset weights of each of the at least two trained RDR prediction models and the sub-predicted MRD of each of the at least two trained RDR prediction models; and selecting the candidate drill bit as the preferred drill bit when the predicted MRD meets preset RDR conditions. This method helps improve the accuracy of prediction results, thereby improving drilling efficiency and obtaining better drilling results, achieving the goal of increasing drilling speed and efficiency.

Description

Technical Field

[0001] This invention relates to the field of oil drilling engineering technology, and in particular to a method, apparatus, computing device, and storage medium for selecting optimal drill bits. Background Technology

[0002] As the primary equipment in rock drilling machinery, the type of drill bit plays a crucial role in the overall efficiency of drilling operations. Choosing an unsuitable drill bit leads to frequent tripping and replacement, reducing the rate of penetration (ROP), shortening drill bit life, increasing drilling costs, and decreasing drilling efficiency. Statistics show that drill bit costs have risen to 40% of the total cost of a single well. Therefore, selecting appropriate drill bits is an important means of reducing drill bit costs, and optimizing drill bit selection is a key approach to cutting drilling expenses.

[0003] In related technologies, insufficient consideration is given to the structural characteristics of the drill bit itself; merely considering drilling parameters and formation parameters is insufficient to fully reflect the drill bit's rock-breaking effect and mechanical drilling speed. To optimize drill bit operating efficiency, a systematic drill bit selection method is needed. Summary of the Invention

[0004] The main objective of this invention is to provide a drill bit optimization method, apparatus, computing device, and storage medium to achieve drill bit optimization.

[0005] This invention provides a drill bit selection method, comprising: using at least two trained rate of drilling (RDR) prediction models, determining the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bit in the formation to be drilled based on image data of the candidate drill bit, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled; determining the predicted mechanical rate of drilling of the candidate drill bit in the formation to be drilled based on the preset weights of the at least two trained RDR prediction models and the sub-predicted mechanical rates of drilling corresponding to the at least two trained RDR prediction models; and selecting the candidate drill bit as the preferred drill bit when the predicted mechanical rate of drilling meets the preset RDR conditions.

[0006] In one embodiment, at least two trained drilling rate prediction models include: a K-nearest neighbor model, an XGboost model, and a random forest model.

[0007] In one embodiment, using at least two trained rate of drilling (RDR) prediction models, based on image data of candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, a sub-predicted mechanical rate of drilling (MRD) corresponding to each of the at least two trained RDR prediction models is determined in the formation to be drilled. This includes: using each trained RDR prediction model, extracting multiple structural features of the candidate drill bits based on image data of the candidate drill bits; performing dimensionality reduction processing on the multiple structural features using principal component analysis to obtain the principal components of the multiple structural features of the candidate drill bits; and using each trained RDR prediction model, based on the principal components of the structural features of the candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, determining the sub-predicted mechanical rate of drilling (MRD) corresponding to each trained RDR prediction model in the formation to be drilled.

[0008] In one embodiment, the drilling speed prediction model is trained by the following steps: training multiple candidate drilling speed prediction models using a training sample set, and selecting at least two trained candidate drilling speed prediction models that meet preset evaluation conditions as the trained drilling speed prediction models. The training sample set includes multiple training samples, each of which includes: image data of the drill bit used in the drilled well, formation parameters of the drilled formation, engineering parameters of the drilled well, and the actual mechanical drilling speed.

[0009] In one embodiment, the preset evaluation conditions include: the model's determination coefficient is greater than or equal to a first preset threshold; and / or the model's mean square error is less than or equal to a second preset threshold.

[0010] In one embodiment, the preset weights of at least two trained rate of drilling (RPD) prediction models are determined through the following steps: acquiring historical drilling data for multiple drilled wells, wherein the historical drilling data includes: drill bit type and image data of the drill bit used in the drilled wells, formation type and formation parameters of the drilled formations, engineering parameters of the drilled wells, and the actual mechanical rate of drilling; for each drilled well, using at least two trained RPD prediction models, determining the sub-predicted mechanical rate of drilling corresponding to each of the at least two trained RPD prediction models based on the image data of the drill bit used, the formation parameters of the drilled formations, and the engineering parameters of the drilled wells; for drilled wells with the same drill bit type but different formation types, or for drilled wells with the same formation type but different drill bit types, determining the preset weights of at least two trained RPD prediction models based on the sub-predicted mechanical rate of drilling corresponding to each of the at least two trained RPD prediction models and the actual mechanical rate of drilling of the drilled well through regression analysis.

[0011] In one embodiment, when there are multiple candidate drill bits, the preset drilling speed condition includes: predicting the maximum mechanical drilling speed among the multiple candidate drill bits; or the minimum difference between the predicted mechanical drilling speed and the target mechanical drilling speed or the difference being less than a preset difference threshold.

[0012] This invention provides a drill bit selection device, comprising: a first prediction module, used to determine, using at least two trained rate of drilling (RDR) prediction models, the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bit corresponding to each of the at least two trained RDR prediction models in the formation to be drilled, based on image data of the candidate drill bit, formation parameters of the formation to be drilled, and engineering parameters of the formation to be drilled; a second prediction module, used to determine the predicted mechanical rate of drilling (MRD) of the candidate drill bit in the formation to be drilled, based on the preset weights of each of the at least two trained RDR prediction models and the sub-predicted MRD corresponding to each of the at least two trained RDR prediction models; and a drill bit selection module, used to select the candidate drill bit as the preferred drill bit when the predicted MRD meets preset RDR conditions.

[0013] The present invention provides a computing device, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the drill bit optimization method described above.

[0014] The present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the drill bit optimization method described above.

[0015] The drill bit selection method of the present invention utilizes multiple drilling speed prediction models, comprehensively considering drill bit characteristics, magnetic layer parameters, and engineering parameters, to predict the mechanical drilling speed of candidate drill bits. This method helps to improve the accuracy of the prediction results, thereby improving drilling efficiency and obtaining better drilling results, and achieving the goal of increasing drilling speed and efficiency. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0017] Figure 1 This is a flowchart of a drill bit selection method according to an exemplary embodiment of this application;

[0018] Figure 2 This is a flowchart of a drill bit selection method according to a specific embodiment of this application;

[0019] Figure 3 This is a schematic diagram of the determination coefficients of at least two candidate drilling speed prediction models according to a specific embodiment of this application. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] Example 1

[0022] This embodiment provides a preferred method for drill bits. Figure 1 This is a flowchart of a drill bit optimization method according to an exemplary embodiment of this application. Figure 1 As shown, the method in this embodiment may include:

[0023] S100: Using at least two trained drilling rate prediction models, based on the image data of the candidate drill bit, the formation parameters of the formation to be drilled, and the engineering parameters of the well to be drilled, determine the sub-predicted mechanical drilling rate of the candidate drill bit corresponding to each of the at least two trained drilling rate prediction models in the formation to be drilled.

[0024] S200: Determine the predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled based on the preset weights of at least two trained drilling rate prediction models and the sub-predicted mechanical drilling rates corresponding to the at least two trained drilling rate prediction models.

[0025] S300: When the predicted mechanical drilling speed meets the preset drilling speed conditions, the alternative drill bit will be selected as the preferred drill bit.

[0026] The above method first uses a single drilling speed prediction model to obtain a sub-predicted mechanical drilling speed based on the image data of the candidate drill bit, the formation parameters of the formation to be drilled, and the engineering parameters of the well to be drilled. Then, based on the preset weights of at least two drilling speed prediction models, the predicted mechanical drilling speed for drilling with the candidate drill bit is determined. Based on the predicted mechanical drilling speed, it is decided whether to determine the candidate drill bit as the preferred drill bit.

[0027] In this embodiment, when predicting the mechanical drilling rate of the drill bit, the characteristics of the drill bit itself, as well as formation parameters and engineering parameters, are comprehensively considered, which helps to improve the accuracy of the prediction results of a single drilling rate prediction model. At the same time, using at least two drilling rate prediction models to predict the mechanical drilling rate, and then determining the predicted mechanical drilling rate of the candidate drill bit based on the results obtained from each drilling rate prediction model, can further improve the accuracy of the mechanical drilling rate prediction results. This helps the construction personnel select the truly suitable drill bit from the candidate drill bits, which is conducive to improving drilling efficiency.

[0028] In this embodiment, the drilling speed prediction model may include, for example, random forest, XGboost, extreme tree regression, neural network, support vector machine, K-nearest neighbor algorithm, etc. The number of models may be two, three or more, and those skilled in the art can choose according to their needs.

[0029] Formation parameters of the formation to be drilled may include, for example, porosity, permeability, oil saturation, and effective thickness of the oil layer. Drilling parameters may include, for example, well depth, drilling pressure, rotation speed, displacement, and mud density.

[0030] In one example, using at least two trained rate of drilling (RDR) prediction models, and based on image data of candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bits corresponding to the at least two trained RDR prediction models in the formation to be drilled can be determined. This can include: using each trained RDR prediction model, extracting multiple structural features of the candidate drill bits based on image data of the candidate drill bits; using principal component analysis to reduce the dimensionality of the multiple structural features to obtain the principal components of the multiple structural features of the candidate drill bits; and using each trained RDR prediction model, based on the principal components of the structural features of the candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, determining the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bits corresponding to that trained RDR prediction model in the formation to be drilled.

[0031] In one example, determining the predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled, based on the preset weights of at least two trained drilling rate prediction models and the sub-predicted mechanical drilling rates corresponding to the at least two trained drilling rate prediction models, may include: weighting and summing the sub-predicted mechanical drilling rates corresponding to the at least two trained drilling rate prediction models according to preset weights, and using the sum as the predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled.

[0032] Preset drilling speed conditions may include, for example, minimizing the difference between the predicted mechanical drilling speed and the target mechanical drilling speed, or setting the difference to a preset threshold. When there are multiple candidate drill bits, the preset drilling speed conditions may also include setting the predicted mechanical drilling speed to the maximum among the multiple candidate drill bits.

[0033] In one example, at least two trained drilling rate prediction models may include: a K-nearest neighbor model, an XGBoost model, and a random forest model. In one example, the preset weights for the K-nearest neighbor model, the random forest model, and the XGBoost model may be 0.3, 0.5, and 0.2, respectively. Other preset weights may be set for these three drilling rate prediction models as needed.

[0034] In one example, the preset weights of at least two trained rate of drilling (RPD) prediction models can be determined through the following steps: Obtain historical drilling data for multiple drilled wells, including: drill bit type and image data of the drill bit used in the drilled wells, formation type and formation parameters of the drilled formation, engineering parameters of the drilled wells, and the actual mechanical rate of drilling (MRD); for each drilled well, using at least two trained RPD prediction models, determine the sub-predicted MRD corresponding to each of the at least two trained RPD prediction models based on the image data of the drill bit used, the formation parameters of the drilled formation, and the engineering parameters of the drilled wells; for drilled wells with the same drill bit type but different formation types, or for drilled wells with the same formation type but different drill bit types, determine the preset weights of at least two trained RPD prediction models based on regression analysis, according to the sub-predicted MRD corresponding to each of the at least two trained RPD prediction models and the actual MRD of the drilled well.

[0035] In this example, drill bit types may include, for example, roller cone bits, PDC (Polycrystalline Diamond Compact bit) bits, and hybrid bits. Formation types may be classified according to formation depth, formation lithology, or other formation characteristics; no specific limitations are made here.

[0036] In one example, the drilling speed prediction model can be trained by the following steps: training multiple candidate drilling speed prediction models using a training sample set, and selecting at least two trained candidate drilling speed prediction models that meet the preset evaluation conditions as the trained drilling speed prediction models. The training sample set includes multiple training samples, each of which includes: image data of the drill bit used in the drilled well, formation parameters of the drilled formation, engineering parameters of the drilled well, and the actual mechanical drilling speed of the well.

[0037] Preset evaluation criteria may include: the model's determination coefficient being greater than or equal to a first preset threshold; and / or the model's mean square error being less than or equal to a second preset threshold. Of course, other evaluation metrics can also be used to evaluate the trained candidate drilling speed prediction model, thereby determining the trained drilling speed prediction model.

[0038] The drill bit selection method of the present invention utilizes multiple drilling speed prediction models, comprehensively considering drill bit characteristics, magnetic layer parameters, and engineering parameters, to predict the mechanical drilling speed of candidate drill bits. This method helps to improve the accuracy of the prediction results, thereby improving drilling efficiency and obtaining better drilling results, and achieving the goal of increasing drilling speed and efficiency.

[0039] Example 2

[0040] This embodiment provides a drill bit selection device, comprising: a first prediction module, used to determine, using at least two trained drilling rate prediction models, the sub-predicted mechanical drilling rate (MRDR) of the candidate drill bit corresponding to each of the at least two trained MDR prediction models in the formation to be drilled, based on image data of the candidate drill bit, formation parameters of the formation to be drilled, and engineering parameters of the formation to be drilled; a second prediction module, used to determine the predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled, based on the preset weights of each of the at least two trained MDR prediction models and the sub-predicted mechanical drilling rates corresponding to each of the at least two trained MDR prediction models; and a drill bit selection module, used to select the candidate drill bit as the preferred drill bit when the predicted mechanical drilling rate meets preset drilling rate conditions.

[0041] In one example, the drill bit selection device may further include a processor and a memory, the processor being used to execute a first prediction module, a second prediction module, and a drill bit selection module stored in the memory.

[0042] In one example, at least two trained drilling rate prediction models may include: a K-nearest neighbor model, an XGboost model, and a random forest model.

[0043] In one example, the first prediction module can also be used to: extract multiple structural features of candidate drill bits based on image data of candidate drill bits using each trained drilling rate prediction model; perform dimensionality reduction on the multiple structural features using principal component analysis to obtain the principal components of the multiple structural features of candidate drill bits; and determine the sub-predicted mechanical drilling rate of the candidate drill bits corresponding to each trained drilling rate prediction model in the formation to be drilled, based on the principal components of the structural features of candidate drill bits, formation parameters of the formation to be drilled, and engineering parameters of the formation to be drilled.

[0044] In one example, the drill bit selection device may further include: a model training module, used to train multiple candidate drilling speed prediction models using a training sample set, and select at least two trained candidate drilling speed prediction models that meet preset evaluation conditions as trained drilling speed prediction models, wherein the training sample set includes multiple training samples, each training sample including: image data of the drill bit used in the drilled well, formation parameters of the drilled formation, engineering parameters of the drilled well, and the actual mechanical drilling speed of the well.

[0045] In one example, the preset evaluation criteria may include: the model's determination coefficient is greater than or equal to a first preset threshold; and / or the model's mean square error is less than or equal to a second preset threshold.

[0046] In one example, the model training module can also be used to: acquire drilling history data for multiple drilled wells, wherein the drilling history data includes: the drill bit type and image data of the drill bit used in the drilled wells, the formation type and formation parameters of the drilled formations, the drilling engineering parameters, and the actual mechanical drilling rate of the drilled wells; for each drilled well, using at least two trained drilling rate prediction models, based on the image data of the drill bit used, the formation parameters of the drilled formations, and the drilling engineering parameters, determine the sub-predicted mechanical drilling rate corresponding to each of the at least two trained drilling rate prediction models for that drilled well; for drilled wells with the same drill bit type but different formation types, or for drilled wells with the same formation type but different drill bit types, through regression analysis, determine the preset weights of each of the at least two trained drilling rate prediction models based on the sub-predicted mechanical drilling rate corresponding to each of the at least two trained drilling rate prediction models and the actual mechanical drilling rate of the drilled well.

[0047] In one example, when there are multiple candidate drill bits, the preset drilling speed conditions may include: predicting the highest mechanical drilling speed among multiple candidate drill bits; or the smallest difference between the predicted mechanical drilling speed and the target mechanical drilling speed or the difference being less than a preset difference threshold.

[0048] Example 3

[0049] A trend is defined as the direction of change of one parameter relative to another relative to a reference point. In trend analysis, what matters is not the current value of any single parameter, but rather its trend relative to other parameters. Machine learning is increasingly popular in drilling parameter trend analysis, with its significant advantage being the ability to overcome human limitations by simultaneously analyzing the trends of multiple parameters. Machine learning methods also provide automated solutions to these limitations, while artificial intelligence can leverage human knowledge to reduce the occurrence of stuck drill pipe.

[0050] The method in this embodiment uses change point detection to identify drilling rig activity and related drilling parameters, thereby capturing the changing trends of these parameters. Inherent in these parameter trends are various factors that influence their readings, such as wellbore geometry, bottom hole assembly (BHA), dog leg severity (DLS), formation characteristics, pump displacement, and tubing rotation. A key factor in successfully identifying stuck drilling is the ability to perform critical and continuous monitoring of trends in multiple drilling and tripping parameters to detect anomalies early.

[0051] This embodiment provides a specific example of a drill bit optimization method. Figure 2 This is a flowchart of a drill bit selection method according to a specific embodiment of this application.

[0052] The preferred drill bit method in this embodiment may include the following steps:

[0053] (1) Collect images of different types of drill bits and related usage records, such as drill bit factory data, well depth, diameter, drilling pressure, rotation speed, drilling speed, displacement, formation parameters, engineering parameters, actual mechanical drilling speed, etc., and construct the correlation between images of different types of drill bits and drill bit usage records to form a training sample set.

[0054] (2) Process the drill bit image, extract multiple structural features of the drill bit, and obtain the principal components of multiple structural features through principal component analysis (PCA (Principal Component Analysis): k-dimensional features reconstructed on the basis of the original n-dimensional features, where k is less than n. For example, the 15 structural features of the drill bit image are reduced to 2). Then, add the principal components to the training sample set.

[0055] An improved edge protocol for drill bit images is employed using a Laplacian filter with α=1, and a structuring element is applied to clearly expanding, eroding, and closing boundary edges. A set of one-dimensional intensity signals is then generated to compute distance classifiers. These classifiers are used to describe the most important information in the original two-dimensional image.

[0056] First-order surface metrics are extracted from the drill bit image. These metrics, chosen for their regional variation, typically affect the relationship between the regions and their major and minor axes. First, the RGB image is converted to a binary class or array of 0s and 1s using a data class transformation technique. The surface metrics obtained for each binary bit image are compared with new metrics used as a reference image. These metrics include: image pixel area, perimeter, convex area, major axis length, minor axis length, area ratio, eccentricity, equivalent region diameter, major axis deviation angle, and image pixel area ratio.

[0057] Second-order statistical measures are extracted from drill bit images. The gray-level co-occurrence matrix (GLCM) is used to represent these measures, including the representation of image texture features (by calculating the frequency of pixel occurrences in the image). Features of the GLCM include energy, homogeneity, contrast, entropy, and correlation.

[0058] (3) Train multiple drilling speed prediction models using a training sample set, using mechanical drilling speed as the output data, and select at least two models that meet the preset evaluation conditions, assigning corresponding preset weights to the selected models. Drilling speed prediction models include, for example, support vector machines, neural networks, XGboost, random forests, linear regression, etc. Different models have different prediction effects and their own strengths. Therefore, some evaluation indicators can be used to evaluate each model and select drilling speed prediction models that meet the prediction requirements. In one example, a neural network model, an XGboost model, and a random forest model can be selected, and the weights set for the three models can be, for example, 0.3, 0.5, and 0.2, respectively. The preset evaluation conditions can include a decision coefficient greater than a certain threshold and / or a mean squared error less than a certain threshold. Multiple drilling speed prediction models can be trained using a training sample set through cross-validation. For example, 70% of the training samples can be selected as the training set, 15% as the validation set, and 15% as the test set.

[0059] Geological data, engineering parameter data, and relevant drill bit parameter data from 50 wells in Block A of the Northwest Oilfield were used for model training. Random Forest, XGBoost, Limit Tree Regression, Neural Network, Support Vector Machine Regression, and K-Nearest Neighbor Regression algorithms were selected for calculation and comparison. Among them, R... 2 The coefficient represents the decision coefficient of the trained model, which is between (0, 1). The larger the value, the better the model's prediction performance. Figure 3 The diagram shows the determination coefficients of at least two candidate drilling speed prediction models according to a specific embodiment of this application. The results show that the random forest model, XGboost model and K-nearest neighbor algorithm model have relatively high prediction accuracy. Therefore, the above three algorithm models can be selected for mechanical drilling speed prediction of candidate drill bits.

[0060] (4) After training is completed, the image data of the candidate drill bit, the formation parameters of the formation to be drilled and the engineering parameters of the well to be drilled can be input into at least two trained drilling rate prediction models respectively, and at least two sub-predicted mechanical drilling rates of the candidate drill bit can be output. Then, the predicted mechanical drilling rate (Rate of Penetration) of the candidate drill bit can be further determined: ROP = a*A + b*B + c*C, where A, B and C represent three drilling rate prediction models, and a, b and c represent the preset weights of the three drilling rate prediction models A, B and C respectively.

[0061] In actual drilling operations, the impact of different formation conditions and drilling technology conditions on mechanical drilling rate (MRD) is not significant. Drill bits with different structural parameters exhibit varying drilling performance, which affects the accuracy of the DRD prediction model. Therefore, multiple DRD prediction models can be selected to jointly predict the mechanical drilling rate of the drill bit. The preset weights of each DRD prediction model are constrained by formation conditions and drill bit type. These preset weights can be determined based on data regression analysis. For example, data from 50 wells can be used to perform regression analysis according to formation A, formation B, formation C, formation D, and roller cone, PDC, and hybrid drill bits to determine the preset weights of each DRD model. Table 1 shows the model weight coefficients of the K-nearest neighbor model, XGboost model, and random forest model under different formation and drill bit type conditions.

[0062] Table 1

[0063]

[0064] (5) Determine whether the candidate drill bit is the preferred drill bit based on whether the predicted mechanical drilling speed meets the preset drilling speed conditions. When there are multiple candidate drill bits, the candidate drill bit with the highest predicted mechanical drilling speed can be selected as the preferred drill bit.

[0065] The method described in this embodiment has been tested in the drilling design and construction process of a key well in an oilfield exploration block, and the drill bit for well X in this block has been optimized. During the drilling design process, the mechanical drilling rate of the drill bit was predicted based on formation parameters and drilling engineering parameters (such as drilling pressure, rotational speed, and the type of drill bit to be used). Drill bit models with higher mechanical drilling rates were HAT127, HJ537G, and M1365D, etc. The mechanical drilling rate of this well section was improved by 21.5% compared with adjacent wells.

[0066] Example 4

[0067] This embodiment provides a computing device, including a processor and a memory. The memory stores a computer program, which, when executed by the processor, implements the steps of the drill bit optimization method described above.

[0068] In one embodiment, the computing device may include one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0069] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash FLASH RAM). Memory is an example of computer-readable media.

[0070] Example 5

[0071] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the drill bit optimization method described above.

[0072] Computer programs can use any combination of one or more storage media. The storage media can be a readable signal medium or a readable storage medium.

[0073] Readable storage media may include, for example, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media may include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0074] A readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying a readable computer program. This propagated data signal may take various forms, such as electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any storage medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0075] The computer program contained on the storage medium can be transmitted using any suitable medium, such as wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0076] Computer programs for performing the operations of this invention can be written in any combination of one or more programming languages. Programming languages ​​may include object-oriented programming languages—such as Java, C++, etc.—as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The computer program may execute entirely on the user's computing device, partially on the user's device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device may be connected to the user's computing device via any type of network (e.g., including a local area network or a wide area network), or it may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0077] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application. When the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0078] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate.

[0079] It should be understood that the exemplary embodiments described herein can be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein. 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. These embodiments are provided so that the disclosure of this application is thorough and complete, and that the concept of these exemplary embodiments is fully conveyed to those skilled in the art, and should not be construed as limiting the invention.

[0080] While the spirit and principles of the invention have been described with reference to several specific embodiments, it should be understood that the invention is not limited to the disclosed specific embodiments, and the division of aspects does not imply that features in these aspects cannot be combined for benefit; such division is merely for ease of description. The invention is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method for selecting the best drill bit, characterized in that, include: Using at least two trained drilling rate prediction models, based on the image data of the candidate drill bit, the formation parameters of the formation to be drilled, and the engineering parameters of the well to be drilled, determine the sub-predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled, which corresponds to the at least two trained drilling rate prediction models respectively. The predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled is determined based on the preset weights of at least two trained drilling rate prediction models and the sub-predicted mechanical drilling rates corresponding to the at least two trained drilling rate prediction models. When the predicted mechanical drilling speed meets the preset drilling speed conditions, the candidate drill bit is selected as the preferred drill bit. Using at least two trained rate of drilling (RDR) prediction models, and based on image data of the candidate drill bit, formation parameters of the formation to be drilled, and engineering parameters of the well to be drilled, determine the sub-predicted mechanical rate of drilling (MRD) of the candidate drill bit in the formation to be drilled, corresponding to each of the at least two trained RDR prediction models, including: Using each trained drilling rate prediction model, multiple structural features of the candidate drill bits are extracted based on the image data of the candidate drill bits; Principal component analysis was used to reduce the dimensionality of the multiple structural features to obtain the principal components of the multiple structural features of the candidate drill bits. Using each trained drilling rate prediction model, the sub-predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled is determined based on the principal components of the structural characteristics of the candidate drill bit, the formation parameters of the formation to be drilled, and the engineering parameters of the formation to be drilled, respectively. Determine the preset weights for at least two trained drill rate prediction models using the following steps: Acquire drilling history data for multiple drilled wells, wherein the drilling history data includes: drill bit type and image data of the drill bit used in the drilled wells, formation type and formation parameters of the drilled formations, engineering parameters of the drilled wells, and actual mechanical drilling rate of the drilled wells; For each drilled well, using at least two trained drilling rate prediction models, based on the image data of the drill bit used, the formation parameters of the drilled formation, and the engineering parameters of the drilled well, the sub-predicted mechanical drilling rate corresponding to each of the at least two trained drilling rate prediction models is determined for that drilled well. For drilled wells with the same drill bit type but different formation types, or for drilled wells with the same formation type but different drill bit types, regression analysis is used to determine the preset weights of at least two trained drilling rate prediction models based on the sub-predicted mechanical drilling rate of the drilled well corresponding to at least two trained drilling rate prediction models and the actual mechanical drilling rate of the drilled well.

2. The preferred drill bit method according to claim 1, characterized in that, At least two trained drilling rate prediction models include: K-nearest neighbor model, XGboost model, and random forest model.

3. The preferred drill bit method according to claim 1, characterized in that, The drilling rate prediction model is trained using the following steps: Multiple candidate drilling speed prediction models are trained using a training sample set. At least two trained candidate drilling speed prediction models that meet the preset evaluation conditions are selected as the trained drilling speed prediction models. The training sample set includes multiple training samples, each of which includes: image data of the drill bit used in the drilled well, formation parameters of the drilled well, engineering parameters of the drilled well, and the actual mechanical drilling speed of the well.

4. The preferred drill bit method according to claim 3, characterized in that, The preset evaluation conditions include: The model's decision coefficient is greater than or equal to a first preset threshold; and / or The mean square error of the model is less than or equal to the second preset threshold.

5. The preferred drill bit method according to claim 1, characterized in that, When there are multiple candidate drill bits, the preset drilling speed conditions include: Predict the maximum mechanical drilling rate among multiple alternative drill bits; or The difference between the predicted mechanical drilling rate and the target mechanical drilling rate is minimized or less than a preset difference threshold.

6. A drill bit selection apparatus for implementing the drill bit selection method according to any one of claims 1-5, characterized in that, include: The first prediction module is used to determine the sub-predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled, based on the image data of the candidate drill bit, the formation parameters of the formation to be drilled, and the engineering parameters of the formation to be drilled, according to at least two trained drilling rate prediction models. The second prediction module is used to determine the predicted mechanical drilling rate of the candidate drill bit in the formation to be drilled, based on the preset weights of at least two trained drilling rate prediction models and the sub-predicted mechanical drilling rates corresponding to the at least two trained drilling rate prediction models. The drill bit selection module is used to select the candidate drill bit as the preferred drill bit when the predicted mechanical drilling speed meets the preset drilling speed conditions.

7. A computing device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program that, when executed by the processor, implements the steps of the preferred drill bit method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The system contains a computer program that, when executed by a processor, implements the steps of the preferred drill bit method as described in any one of claims 1 to 5.

Images