A method and system for identifying a formation type of a wellbore

Abstract

The application discloses a method and system for identifying a drilling formation type, comprising the following steps: according to real-time logging data under a drilling condition in a well to be evaluated, the real-time logging data is pretreated, then key features are extracted from the pretreated real-time logging data, and key feature evaluation data is obtained; and according to the key feature evaluation data, a preset formation identification model is used to predict a formation type. The application can use less types of data information to realize quick identification and high-precision identification of the formation type while drilling through real-time data of a comprehensive logging instrument in the process of drilling, and has practical application significance and value.

Description

Technical Field

[0001] This invention relates to the field of oil drilling engineering technology, and in particular to a method and system for identifying drilling formation types. Background Technology

[0002] While intelligent technology has become a current trend, formation type identification in drilling engineering still relies on manual experience-based methods such as lithology testing and well logging. This not only suffers from significant time lag but also struggles to guarantee the accuracy of formation type identification. Achieving "formation identification while drilling" through formation identification while drilling can provide a foundation for drilling engineering risk warning and drilling parameter optimization, which has important practical significance for oil and gas development.

[0003] Research on existing patented technologies related to well formation identification and classification reveals problems such as low accuracy in experimental formation identification and difficulty in obtaining the required data. For example, well logging data is collected after drilling operations are completed and provided to technicians for formation identification based on the logging curves. Therefore, formation identification based on this data cannot guarantee the real-time nature of drilling-while-drilling formation identification, and its practical application in current drilling operations remains to be investigated. However, with the advent of microcomputers and continuous improvement in computing power, integrated logging technology has been continuously refined. Integrated logging technology is a comprehensive logging operation in oil drilling operations that uses circulating drilling fluid as the carrier of information and employs various detection instruments to record changes in geological, oil and gas, pressure, and rock properties in the drilling fluid with depth. The sensors in the integrated logging instrument collect a set of data every 3 seconds. Each set of data contains nearly a hundred characteristic parameters, including well depth, drilling pressure, torque, rotational speed, inlet and outlet density, conductivity, flow rate, and gas content, accumulating a massive amount of data during the drilling process. Therefore, in drilling formation type identification tasks, the logging data generated by the integrated logging instrument has extremely high research and utilization value.

[0004] In addition, due to the massive amount of drilling data from each well, the dozens of characteristic parameters recorded by the logging instrument have complex correlations and redundant features, which increases the operating cost of formation identification data; at the same time, the accuracy and generalization ability of drilling formation identification urgently need to be improved. Summary of the Invention

[0005] The purpose of this invention is to provide a drilling formation type identification method to solve the problems of low availability of integrated logging instrument data, redundant characteristic parameters, lack of real-time formation identification, and low accuracy of formation identification while drilling in the above-mentioned drilling projects.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a method for identifying drilling formation types, comprising: preprocessing real-time logging data under drilling conditions in the well to be evaluated, then extracting key features from the preprocessed real-time logging data to obtain key feature evaluation data; and predicting the formation type based on the key feature evaluation data using a preset formation identification model.

[0007] Preferably, the formation identification model is constructed using the following steps: collecting historical logging data from multiple drilled wells to construct an original logging time series dataset; labeling the formation type and drilling conditions corresponding to each historical logging data in the original logging time series dataset, and performing preprocessing; extracting key factors by performing dimensionality reduction processing on the preprocessed original logging time series dataset to form a formation type identification dataset; and training and testing the pre-constructed initial model based on the formation type identification dataset to obtain the formation identification model.

[0008] Preferably, the step of training and testing a pre-constructed initial model based on a stratigraphic type identification dataset to obtain the stratigraphic identification model includes: dividing the stratigraphic type identification dataset into a model training set, a model test set, and a generalization test set; and using the model training set, the model test set, and the generalization test set respectively to sequentially perform model training, model accuracy evaluation testing, and model generalization testing on the initial model.

[0009] Preferably, the step of dividing the formation type identification dataset into a model training set, a model test set, and a generalization test set includes: extracting key factor feature data about a specified drilled well from the formation type identification dataset to form the generalization test set; and dividing the key factor feature data of each formation type under drilling conditions according to a preset ratio based on the remaining data in the formation type identification dataset excluding the generalization test set to form the model training set and the model test set.

[0010] Preferably, the step of training the initial model using the model training set includes: randomly sampling the model training set to construct n different training sample datasets, and based on these, building n different decision tree models; connecting the outputs of the multiple decision tree models in parallel with a result analysis module to form the initial model, wherein the result analysis module is an average value calculation module or a voting classification module; and training the initial model according to the model training set.

[0011] Preferably, the preprocessing step of the labeled original logging time series dataset includes: deleting logging data under non-drilling conditions; outlier removal, wherein the 3σ principle is used to remove outliers from the real-time logging data datasets of different formation types under drilling conditions.

[0012] Preferably, the step of extracting key factors by performing dimensionality reduction processing on the preprocessed original well logging time series dataset to form a formation type identification dataset includes: using principal component analysis algorithm to perform principal component analysis on the historical well logging data with formation type labels in the preprocessed original well logging time series dataset to achieve dimensionality reduction, thereby extracting formation-sensitive features in the well logging data as key features.

[0013] Preferably, the step of labeling the formation type and drilling condition corresponding to each historical logging data in the original logging time series dataset includes: labeling each historical logging data in the original logging time series dataset with a corresponding formation type label and drilling condition label based on the drilling daily report, logging daily report and formation stratification information.

[0014] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method described above.

[0015] In addition, embodiments of the present invention provide a system for identifying drilling formation types, including: a field data acquisition module configured to preprocess real-time logging data of the depth to be evaluated in the well to be evaluated, and then extract key features from the preprocessed real-time logging data to obtain key feature evaluation data; and an actual prediction module configured to predict the formation type of the depth to be evaluated based on the key feature evaluation data and using a preset formation identification model.

[0016] Compared with the prior art, one or more embodiments of the above solutions may have the following advantages or beneficial effects:

[0017] This invention proposes a method and system for identifying drilling formation types. The method and system employ a drilling data preprocessing approach based on time-series data segmentation and recombination and the 3-sigma outlier criterion to segment, process outliers, and classify labels on the original integrated logging data. This significantly improves the utilization rate of logging data and lays a solid foundation for subsequent data mining. The invention also utilizes a PCA-based method for selecting dominant feature parameters for drilling formation identification, reducing the dimensionality of formation identification data, solving the data redundancy problem, and ensuring the running speed of the random forest formation type identification model. Furthermore, the drilling formation type identification method based on the random forest algorithm in this invention can achieve rapid and high-precision identification of formation types during drilling using less data from the integrated logging instrument in real time, making it more practically significant and valuable.

[0018] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the description, claims, and drawings. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0020] Figure 1 This is a schematic diagram illustrating the steps of a method for identifying well formation types according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram illustrating the construction process of the formation identification model in the method for identifying drilling formation types according to an embodiment of this application.

[0022] Figure 3 This is a schematic diagram of the formation identification model in the method for identifying drilling formation types according to an embodiment of this application.

[0023] Figure 4 This is a schematic diagram illustrating the specific process of a method for identifying drilling formation types according to an embodiment of this application.

[0024] Figure 5 This is a schematic diagram of the system for identifying drilling formation types according to an embodiment of this application. Detailed Implementation

[0025] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples, so that the process of how the present invention uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly. It should be noted that, as long as there is no conflict, the various embodiments and features in the various embodiments of the present invention can be combined with each other, and the resulting technical solutions are all within the protection scope of the present invention.

[0026] Furthermore, the steps illustrated in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in a different order than that shown here.

[0027] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments. Unless the context clearly indicates otherwise, the singular forms “a” and “an” as used herein are also intended to include the plural. It should also be understood that the terms “comprising” and / or “including” as used herein specify the presence of the stated features, integers, steps, operations, units, and / or components, without excluding the presence or addition of one or more other features, integers, steps, operations, units, components, and / or combinations thereof.

[0028] Figure 1 This is a schematic diagram illustrating the steps of a method for identifying well formation types according to an embodiment of this application. Figure 4 This is a schematic flowchart illustrating a specific process in the method for identifying drilling formation types according to an embodiment of this application. The following is in conjunction with... Figure 1 and Figure 4 The specific steps of the method for identifying drilling formation types (also known as the "formation type identification method") described in the embodiments of the present invention will be explained.

[0029] Step S110 involves preprocessing the real-time logging data of the depth to be evaluated under the drilling conditions in the well to be evaluated (at the drilling site) and then extracting key features from the preprocessed real-time logging data to obtain key feature evaluation data.

[0030] In step S110, a comprehensive logging tool is used to obtain real-time logging data for the drilling conditions at the depth to be evaluated. In practical applications, the comprehensive logging tool collects and records data every few seconds (e.g., every 3 seconds). Each data collection records 76 different features to form a single logging data point. The data table consisting of n rows and 76 columns, arranged chronologically, is called real-time logging data. The value of n typically ranges from several million to tens of millions, depending on the time required for drilling and completion.

[0031] Because the drilling logging instrument collects a massive amount of data, which includes many non-drilling conditions, and because of the influence of the environment, measurement methods, and system noise, there are also outliers in the drilling data.

[0032] Therefore, in step S110, it is necessary to first obtain real-time logging data under drilling conditions, and then remove outliers from the real-time logging data of the depth to be evaluated under the current drilling conditions, so as to obtain preprocessed real-time logging data.

[0033] Since each logging data contains multiple pieces of information, and not all of them are highly correlated with formation type identification, step S110 also needs to extract several key features related to formation type identification from the preprocessed real-time logging data to obtain key feature evaluation data, and then proceed to step S120.

[0034] Step S120: Based on the key feature evaluation data obtained in step S110, use a preset stratigraphic identification model to predict the stratigraphic type at the depth to be evaluated.

[0035] In step S120, several key feature evaluation data obtained in step S110 are directly input into the formation identification model, so that the formation identification model directly outputs the formation type of the depth to be evaluated under the current drilling conditions.

[0036] Figure 2 This is a schematic diagram illustrating the construction process of the formation identification model in the method for identifying drilling formation types according to an embodiment of this application. The following is in conjunction with... Figure 2 and Figure 4 The construction process of the stratigraphic identification model described in the embodiments of the present invention will be explained.

[0037] The first step is to collect historical logging data from multiple drilled wells to construct the original logging time series dataset.

[0038] In the first step, for each drilled well, all real-time logging data collected using the integrated logging instrument is collected as historical logging data. Then, an integrated logging time series matrix is ​​established according to the time when the logging data in the integrated logging instrument of each drilled well is read.

[0039] The second step is to label the formation type and drilling conditions corresponding to each historical real-time logging data in the raw logging time series dataset obtained in the first step, and to preprocess these labeled raw logging time series datasets.

[0040] The step of labeling the formation type and drilling condition corresponding to each historical logging data in the original logging time series dataset includes: labeling each historical real-time logging data in the original logging time series dataset with the corresponding formation type label and drilling condition label.

[0041] Based on the drilling daily report, logging daily report, and formation stratification information, real-time logging data under drilling conditions are selected and labeled with formation type so that data can be cut and reorganized during preprocessing to obtain a dataset of real-time logging data for each type of formation at the corresponding drilling time.

[0042] The preprocessing steps for the labeled raw logging time series dataset include: firstly deleting real-time logging data under non-drilling conditions; then removing outliers from the real-time logging data under drilling conditions, resulting in a dataset of real-time logging data for different formation types under drilling conditions, denoted as the dataset to be processed.

[0043] In this embodiment of the invention, the 3σ principle is used to remove outliers from the dataset of real-time logging data of different formation types under drilling conditions.

[0044] The 3σ principle states that the confidence interval for normal data is [μ-3σ, μ+3σ], where μ represents the mean of each drilling parameter and σ represents the standard deviation of each drilling parameter. Assume there are m i-dimensional data points as follows:

[0045] x i =(x1) i x2 i ,……,x m i )

[0046] If x exists j i Make |x j i If -μ|>3σ, then x is considered j i Outliers should be removed, and the values ​​in those locations should be filled using the mean method. The specific formula is as follows:

[0047]

[0048] in, To fill in the value, This refers to data near the original outlier point.

[0049] After obtaining the dataset to be processed, the third step is to construct a stratigraphic type identification dataset.

[0050] The third step involves extracting key factors by performing dimensionality reduction on the preprocessed original logging time series dataset (i.e., the dataset to be processed mentioned above) to form a stratigraphic type identification dataset.

[0051] In the third step, principal component analysis (PCA) is used to perform PCA on the historical real-time logging data with formation type labels in the preprocessed original logging time series dataset to achieve dimensionality reduction, thereby extracting multiple formation-sensitive features from the logging data as key feature terms.

[0052] In this embodiment of the invention, based on the principal component analysis (PCA) method, feature dimensionality reduction is performed on each real-time logging data in the dataset to be processed: Since each real-time logging data collected by the existing integrated logging instrument contains 76-dimensional feature parameters, and the correlation between different feature parameters is complex and redundant, if all feature parameters are directly used as input to the subsequent formation type identification model without processing, it will increase the model complexity and model running time. Therefore, in this embodiment of the invention, principal component analysis can be used to select strongly correlated feature parameters for formation identification to form a formation type identification dataset.

[0053] Optimization of dominant characteristic parameters for drilling formation identification based on principal component analysis (PCA) algorithm.

[0054] For well formation type identification, Principal Component Analysis (PCA) is used to reduce the dimensionality of characteristic parameters. PCA originates from matrix operations, matrix diagonalization, and spectral decomposition. Its core idea is to explain most of the information of the original variables through linear combinations of the original characteristic parameters, thus achieving dimensionality reduction and simplifying the problem. Taking a two-dimensional space as an example, when the coordinate axes are rotated until they are parallel to the major and minor axes of an ellipse, the variable representing the major axis describes the main changes in the data, while the variable representing the minor axis describes the secondary changes. The major axis variable is used to represent most of the information contained in the data, while the information represented by the minor axis is ignored (discarding the secondary dimension), thus completing the dimensionality reduction. For a multi-dimensional dataset for well formation type identification after outlier processing (preprocessed original logging time series dataset), the PCA algorithm can be used to extract the dominant characteristic parameters for well formation identification.

[0055] For the raw logging time series dataset after data preprocessing, principal component analysis algorithm is used to extract the dominant feature parameters, and a general mathematical model can be established.

[0056]

[0057] F = b1Y1 + b2Y2 + ... + b m Y m

[0058] Among them, X i Y is the i-th feature parameter in the list of each well logging data in the original well logging time series dataset; j The j-th principal component (key feature or key factor feature) extracted from the original well logging time series dataset; a ij The factor loadings are the key parameters; the larger the factor loadings, the stronger the correlation between the feature parameters and the principal components. j Let F represent the contribution rate of the j-th principal component, and let F represent the drilling formation identification prediction score.

[0059] Based on the ranking of the coefficients of each characteristic parameter in the calculation of the drilling formation identification prediction score, the dominant parameter set with drilling formation identification as the target can be selected as the input parameter set of the formation identification model.

[0060] After completing the construction of the stratigraphic type identification dataset, we proceed to the fourth step.

[0061] The fourth step involves training and testing the pre-built initial model based on the stratigraphic type identification dataset obtained in the third step to obtain the stratigraphic identification model.

[0062] Specifically, in the fourth step, the stratigraphic type identification dataset is first divided into a model training set, a model test set, and a generalization test set. Then, the model training set, the model test set, and the generalization test set are used to train the pre-built initial model, evaluate the model accuracy, and perform model generalization tests in sequence, thereby obtaining a stratigraphic identification model that can be put into practical application.

[0063] Furthermore, the step of dividing the stratigraphic type identification dataset into a model training set, a model test set, and a generalization test set includes:

[0064] Extract key feature data about specified (a small portion) drilled wells from the formation type identification dataset to form a generalization test set;

[0065] Based on the remaining data in the stratigraphic type identification dataset excluding the generalization test set, the key factor feature data of each stratigraphic type under drilling conditions are divided according to a preset ratio to form a model training set and a model test set.

[0066] For example: In the above formation type identification dataset, the key factor feature data is used as the feature variable and the input variable of the model, and the marked formation type digital label is used as the output variable of the model. Two wells are reserved from the existing historical well data set of 15 wells as the model generalization ability test wells. All the key factor feature data of the remaining 13 wells is divided into a model training set and a model test set according to a preset ratio (e.g., 8:2). An initial model based on random forest is built, the model parameters are set, and the model is trained. The trained model is evaluated for model accuracy on the 13-well test set. Then, the model after completing the model accuracy evaluation is tested on the data of the above two reserved wells to test the model generalization ability, and the input of the feature variable of the formation identification model while drilling is adjusted according to the generalization accuracy, aiming to achieve fast and accurate formation type identification while drilling in the drilling engineering.

[0067] In one embodiment, the task of identifying the formation type while drilling is carried out based on the random forest algorithm.

[0068] The ensemble learning model is a very important part of machine learning. Ensemble learning is a machine learning method that uses a series of weak learners (or called base models) for learning and integrates the results of each weak learner to obtain a better learning effect than a single learner. The ensemble learning model usually includes: Bagging algorithm and Boosting algorithm. The random forest model is a typical machine learning model of the Bagging algorithm, and its weak learner is a decision tree model. As Figure 3 shown, the random forest model randomly samples in the formation identification dataset to form n different sample datasets, then builds n different decision tree models based on these datasets, and finally obtains the final result according to the average value (for regression models) or voting (for classification models) of these decision tree models.

[0069] To ensure the generalization ability of the drilling formation identification model, when building each tree of the random forest formation identification model, two basic principles are followed: data randomness and feature randomness. Data randomness: Randomly draw data with replacement from all formation type identification datasets as the data for training one of the decision trees. Feature randomness: If the feature dimension of each formation type identification dataset is M, specify a constant k < M, and randomly select k features from M features. When using the random forest formation identification model, the default number of features k taken is the square root of M.

[0070] In the embodiment of the present invention, in the step of training the initial model using the model training set, it includes:

[0071] Randomly sample the model training set to construct n (n is a positive integer) different training sample datasets, and build n different decision tree models based on these datasets.

[0072] The outputs of multiple decision tree models are connected in parallel and then connected to the result analysis module to form an initial model. The result analysis module is either an average value calculation module or a voting classification module.

[0073] The initial model is trained based on the model training set.

[0074] Thus, after completing the initial model training, accuracy testing, and generalization testing, a stratigraphic identification model is obtained.

[0075] Subsequently, a formation identification model was applied to develop a formation type identification method. During the application of the formation identification model, real-time logging data was acquired and key feature parameters were extracted. The real-time data of these key feature parameters were then input into the formation identification model to directly predict the formation type at the current well depth.

[0076] Based on the above-described formation type identification method, this invention provides a computer-readable storage medium. The storage medium stores a computer program, which is executed to run a method for identifying drilling formation types. The computer program is capable of executing computer instructions, which include computer program code. The computer program code can be in the form of source code, object code, executable file, or some intermediate form.

[0077] Computer-readable storage media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0078] It should be noted that the contents of computer-readable storage media may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, the contents may be appropriately increased or decreased according to the requirements of legislation and patent practice. In other jurisdictions, computer-readable storage media may not include electrical carrier signals and telecommunication signals.

[0079] Based on the formation type identification method described in the above embodiments, the present invention also provides a system for identifying drilling formation types (also referred to as a "formation type identification system"). This formation type identification system is used to implement the formation type identification method described above.

[0080] Figure 5 This is a schematic diagram of the system for identifying drilling formation types according to an embodiment of this application. Figure 5 As shown, the stratigraphic type identification system described in this embodiment of the invention includes: a field data acquisition module 501 and an actual prediction module 502.

[0081] The field data acquisition module 501 is implemented according to the method described in step S110 above, and is configured to preprocess the real-time logging data of the depth to be evaluated in the well to be evaluated, and then extract key features from the preprocessed real-time logging data to obtain key feature evaluation data; the actual prediction module 502 is implemented according to the method described in step S120 above, and is configured to predict the formation type of the depth to be evaluated based on the key feature evaluation data and using a preset formation identification model.

[0082] This invention discloses a method and system for identifying drilling formation types. The method and system employ a drilling data preprocessing approach based on time-series data segmentation and recombination and the 3-sigma outlier criterion to segment, process outliers, and classify labels on the original integrated logging data. This significantly improves the utilization rate of logging data and lays a solid foundation for subsequent data mining. The invention also employs a PCA-based method for selecting dominant feature parameters for drilling formation identification, reducing the dimensionality of formation identification data, solving the data redundancy problem, and ensuring the running speed of the random forest formation type identification model. Furthermore, the drilling formation type identification method based on the random forest algorithm in this invention can achieve rapid and high-precision identification of formation types during drilling using less data from the integrated logging instrument in real time, making it more practically significant and valuable.

[0083] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0084] In the description of this invention, unless otherwise stated, "a plurality of" means two or more; the terms "upper," "lower," "left," "right," "inner," "outer," "front end," "rear end," "head," "tail," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0085] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0086] It should be understood that the embodiments disclosed herein are not limited to the specific structures, processing steps, or materials disclosed herein, but should be extended to equivalent substitutions of these features as understood by those skilled in the art. It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.

[0087] The phrase "an embodiment" or "an embodiment" used in this specification means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. Therefore, the phrase "an embodiment" or "an embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment.

[0088] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

[0089] While the embodiments disclosed in this invention are as described above, the content is merely for the purpose of facilitating understanding of the invention and is not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and changes in form and detail of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection of this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for identifying drilling formation types, characterized in that, include: Based on the real-time logging data under drilling conditions in the well to be evaluated, the data is preprocessed, and then key features are extracted from the preprocessed real-time logging data to obtain key feature evaluation data. Based on the key feature evaluation data, the stratigraphic type is predicted using a preset stratigraphic identification model.

2. The method according to claim 1, characterized in that, The stratigraphic identification model is constructed using the following steps: Collect historical logging data from multiple drilled wells to construct a raw logging time series dataset; The formation type and drilling condition corresponding to each historical logging data in the original logging time series dataset are labeled and preprocessed. By performing dimensionality reduction on the preprocessed raw well logging time series dataset, key factors are extracted to form a stratigraphic type identification dataset. Based on the stratigraphic type identification dataset, the pre-constructed initial model is trained and tested to obtain the stratigraphic identification model.

3. The method according to claim 2, characterized in that, The step of training and testing a pre-built initial model based on a stratigraphic type identification dataset to obtain the stratigraphic identification model includes: The stratigraphic type identification dataset is divided into a model training set, a model test set, and a generalization test set. The initial model is trained, its accuracy is evaluated, and its generalization is tested sequentially using the model training set, the model test set, and the generalization test set, respectively.

4. The method according to claim 3, characterized in that, The step of dividing the stratigraphic type identification dataset into a model training set, a model test set, and a generalization test set includes: Extract key factor feature data about a specified drilled well from the formation type identification dataset to form the generalization test set; Based on the remaining data in the formation type identification dataset excluding the generalization test set, the key factor feature data of each formation type under drilling conditions are divided according to a preset ratio to form the model training set and the model test set.

5. The method according to claim 3 or 4, characterized in that, The step of training the initial model using the model training set includes: Randomly sample the training set of the model to construct n different training sample datasets, and build n different decision tree models based on these datasets. The outputs of multiple decision tree models are connected in parallel and then connected to the result analysis module to form the initial model, wherein the result analysis module is an average value calculation module or a voting classification module; The initial model is trained based on the model training set.

6. The method according to any one of claims 2 to 5, characterized in that, The preprocessing steps for the labeled raw logging time series dataset include: Delete logging data from non-drilling conditions; Outlier removal is performed by using the 3σ principle to remove outliers from the real-time logging data datasets of different formation types under drilling conditions.

7. The method according to any one of claims 2 to 6, characterized in that, The steps involved in extracting key factors and forming a stratigraphic type identification dataset by performing dimensionality reduction on the preprocessed raw well logging time series dataset include: Principal component analysis (PCA) is used to perform PCA on historical logging data with formation type labels in the preprocessed raw logging time series dataset to achieve dimensionality reduction, thereby extracting formation-sensitive features from the logging data as key features.

8. The method according to any one of claims 2 to 7, characterized in that, The step of labeling the formation type and drilling condition corresponding to each historical logging data in the original logging time series dataset includes: Based on the drilling daily report, logging daily report, and formation stratification information, each historical logging data in the original logging time series dataset is labeled with the corresponding formation type label and operating condition label.

9. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1 to 8.

10. A system for identifying drilling formation types, characterized in that, include: The field data acquisition module is configured to preprocess the real-time logging data of the depth to be evaluated in the well to be evaluated, and then extract key features from the preprocessed real-time logging data to obtain key feature evaluation data. The actual prediction module is configured to predict the type of strata at the depth to be evaluated based on the key feature evaluation data and using a preset stratum identification model.

Images