Gradient lifting decision tree-based lithofacies classification prediction method, device and system
Through the lithophago classification prediction method based on gradient enhancement decision trees, the lithophago classification model is constructed using logging characteristics, which solves the problems of low manual identification efficiency and low accuracy in the existing technology, and achieves more efficient and accurate lithophago classification.
Patent Information
- Application Number
- CN202311210017.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-19
- Publication Date
- 2025-03-21
AI Technical Summary
In the prior art, artificial lithophase identification and classification methods based on drilling centering data and logging data have problems of low efficiency and scarcity of core data, resulting in low accuracy of classification results.
The lithophago classification prediction method based on the gradient enhancement decision tree is adopted. By obtaining the logging characteristics of the well to be predicted, and using multiple sets of sample data to train the gradient enhancement decision tree model, a lithophago classification model is constructed to realize automatic lithophago classification prediction.
The efficiency and accuracy of lithophagocytic classification are improved, which is more efficient and accurate than manual recognition methods, and the training data can be continuously updated to improve the accuracy of the classification model.
Smart Images
Figure CN119669911A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of lithofacies classification, in particular to a lithofacies classification prediction method, device and system based on a gradient boosting decision tree. Background Art
[0002] The detailed characterization of delta plain-delta front sedimentary microfacies is a necessary means to carry out efficient development of clastic oil and gas. However, the detailed characterization of clastic lithofacies is mainly based on the manual identification of drilling coring data and logging data. This working method has two defects: on the one hand, the cost of drilling coring is expensive and the core data is scarce; on the other hand, it is difficult to obtain a unified result and the efficiency is low if lithofacies are divided based on personal experience. Therefore, how to use limited logging data to carry out efficient prediction of delta lithofacies has become a technical problem.
[0003] At present, machine learning-based methods have been applied in the field of development geology, including factor analysis, Bayesian discriminant method, data mining and unsupervised machine learning and artificial neural network (ANN), etc. However, these methods are mainly used for reservoir evaluation and do not involve delta plain-delta front lithofacies classification prediction. Summary of the invention
[0004] The present invention provides a lithofacies classification prediction method, device and system based on a gradient boosting decision tree, which overcomes the deficiencies of the above-mentioned prior art and can effectively solve the problems of low efficiency of the existing artificial lithofacies identification and classification methods based on drilling coring data and logging data and low accuracy of classification results due to scarcity of core data.
[0005] One of the technical solutions of the present invention is achieved by the following measures: a petrographic classification prediction method based on a gradient boosting decision tree, comprising: Obtaining logging characteristics of the well to be predicted; Inputting the logging characteristics of the well to be predicted into the lithofacies classification model, wherein the lithofacies classification model is obtained by training the gradient boosting decision tree model using multiple sets of sample data, each set of sample data in the multiple sets of sample data includes: the logging characteristics of the historical well core sample and the identification information used to identify the category of the historical well core sample; The lithofacies classification model outputs the lithofacies classification results of the well to be predicted.
[0006] The following are further optimizations and / or improvements to the above technical solutions: The above logging characteristics include natural gamma, natural unit, acoustic time difference, wellbore diameter, deep induction resistivity, and eight lateral resistivity.
[0007] The construction process of the above lithofacies classification model includes: Constructing a sample data set, wherein the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; Divide the sample data set into a training data subset and a test data subset; The gradient boosting decision tree model is trained using the training data subset to obtain the lithofacies classification model; The lithofacies classification model is tested using a test data subset, and the classification accuracy of the lithofacies classification model is evaluated based on the test results; In response to the classification accuracy evaluation result meeting the set value, the lithofacies classification model is output; in response to the classification accuracy evaluation result not meeting the set value, the gradient boosting decision tree model parameters are adjusted, and the gradient boosting decision tree model is retrained using the training data subset.
[0008] The above-mentioned true lithofacies types include delta plain distributary channel lithofacies type, delta front underwater distributary channel lithofacies type, estuary bar lithofacies type, front sheet sandstone facies type, and pro-delta mudstone facies type.
[0009] The above combined test results evaluate the classification accuracy of the gradient boosting decision tree model, including: The test results of each sample data in the test data subset are compared with the corresponding real lithofacies type to generate a classification accuracy confusion matrix; The classification accuracy of the lithofacies classification model was evaluated based on the classification accuracy confusion matrix; The accuracy of the classification result is compared with the set value. In response to the classification accuracy evaluation result meeting the set value, the lithofacies classification model is output. In response to the classification accuracy evaluation result not meeting the set value, the parameters of the gradient boosting decision tree model are adjusted, and the gradient boosting decision tree model is retrained using the training data subset.
[0010] The second technical solution of the present invention is achieved by the following measures: a lithology classification prediction device based on a gradient boosting decision tree, comprising: A feature extraction unit, which obtains the logging features of the well to be predicted; A classification prediction unit inputs the logging characteristics of the well to be predicted into a lithofacies classification model, wherein the lithofacies classification model is obtained by training a gradient boosting decision tree model using multiple sets of sample data, and each set of sample data in the multiple sets of sample data includes: the logging characteristics of the historical well core sample and identification information used to identify the category of the historical well core sample; The result output unit, the lithofacies classification model outputs the lithofacies classification result of the well to be predicted.
[0011] The following are further optimizations and / or improvements to the above technical solutions: The above also includes a model building unit, including: A sample acquisition module constructs a sample data set, wherein the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; A subset partitioning module divides the sample data set into a training data subset and a test data subset; A training module, which trains a gradient boosting decision tree model using a subset of training data to obtain a lithofacies classification model; The testing module uses the test data subset to test the lithofacies classification model and evaluates the classification accuracy of the lithofacies classification model based on the test results; The evaluation module outputs a lithofacies classification model in response to the classification accuracy evaluation result meeting a set value, and adjusts the gradient boosting decision tree model parameters in response to the classification accuracy evaluation result not meeting the set value, and retrains the gradient boosting decision tree model using a training data subset.
[0012] The third technical solution of the present invention is achieved by the following measures: a lithology classification prediction system based on a gradient boosting decision tree, comprising: A lithofacies classification prediction device based on a gradient boosting decision tree, wherein the lithofacies classification prediction device based on a gradient boosting decision tree is the lithofacies classification prediction device based on a gradient boosting decision tree described in the above technical solution; An interactive device is used for an operator to communicate with the lithofacies classification prediction device, so that the operator provides the lithofacies classification prediction device with an interactive interface for the person to be predicted through an interactive unit.
[0013] The present invention combines the gradient boosting decision tree model with limited logging data to obtain a lithofacies classification model through machine learning, and then uses the lithofacies classification model to realize automatic lithofacies classification prediction. Compared with the method of dividing lithofacies based on manual experience, the present invention has high efficiency and accuracy, and can continuously update the sample data used for training, thereby continuously updating the lithofacies classification model and improving the accuracy of lithofacies classification. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Attached Figure 1 The figure is a flow chart of the method of the present invention.
[0015] Attached Figure 2 This is a frequency statistics diagram of the petrographic classification sample data of Example 2 of the present invention.
[0016] Attached Figure 3 Schematic diagram of the classification accuracy confusion matrix of Example 2 of the present invention.
[0017] Attached Figure 4 This is a schematic diagram for comparing the results of Example 2 of the present invention.
[0018] Attached Figure 5The figure is a schematic diagram of the structure of a device of the present invention.
[0019] Attached Figure 6 This is a schematic diagram of another device structure of the present invention.
[0020] Attached Figure 7 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION
[0021] The present invention is not limited by the following embodiments, and specific implementation methods can be determined based on the technical solution of the present invention and actual conditions.
[0022] The present invention will be further described below in conjunction with embodiments and drawings: Embodiment 1: As attached Figure 1 As shown, the embodiment of the present invention discloses a lithology classification prediction method based on a gradient boosting decision tree, comprising: Step S110, obtaining the well logging characteristics of the well to be predicted.
[0023] In the above steps, the logging characteristics include natural gamma, natural unit, acoustic time difference, well diameter, deep induction resistivity, and eight lateral resistivity.
[0024] Step S120, inputting the logging characteristics of the well to be predicted into the lithofacies classification model, wherein the lithofacies classification model is obtained by training a gradient boosting decision tree model using multiple groups of sample data, and each group of sample data in the multiple groups of sample data includes: the logging characteristics of the historical well core sample and identification information used to identify the category of the historical well core.
[0025] In the above steps, the construction process of the lithofacies classification model includes: (1) Constructing a sample data set, where the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; Since typical well lithofacies include delta plain distributary channel lithofacies type, delta front underwater distributary channel lithofacies type, estuary bar lithofacies type, front sheet sandstone facies type, and prodelta mudstone facies type, the actual lithofacies types of the sample data in this step include the above five types.
[0026] (2) Divide the sample data set into a training data subset and a test data subset; In this step, the train_test_split module can be used to divide the training data subset and the test data subset, where the training data subset can account for 80% of the sample data set and the test data subset can account for 20% of the sample data set.
[0027] (3) Using the training data subset to train the gradient boosting decision tree model to obtain the lithofacies classification model; In this step, the gradient boosting decision tree model is trained using a subset of the training data. During training, the GridSearchCV-grid parameter adjustment module can be used for grid search and cross-validation to find the most accurate parameters within the specified parameter range, and the remaining learning parameters are set according to the final selected hyperparameters. For example, the number of iterations in the Boosting workflow can be set to 500 times, the maximum depth of each learner can be set to 3, the minimum number of classification samples can be set to 2, the learning rate can be set to 0.1, and the loss function can be set to variance. The principle of the gradient boosting decision tree model (GBDT model) used during training is as follows: At the beginning, a function value that can continuously reduce the loss is estimated. The decision tree consists of a single root node, represented by a constant γ, as shown in formula (1) f 0 (x) = arigmin γ (1) For m (m=1, 2, 3, ...) iterations of the computational experiment, perform the following steps: Step (1) calculates and applies the negative gradient of the loss function in this model and estimates the residual according to formula (2): γ im = (2) Step (2) calculates the value of the corresponding leaf node in the regression tree; Step (3) uses linear retrieval to estimate the value of the leaf node area according to formula (3): γ im = arigmin y (3) Step (4) The regression tree is updated using formula (4): (4) The gradient decision tree is finally output as a function of x, as shown in formula (5): (5) (4) Testing the lithofacies classification model using a test data subset; (5) Compare the test results of each sample data in the test data subset with the corresponding real lithofacies type to generate a classification accuracy confusion matrix; (6) Evaluate the classification accuracy of the lithofacies classification model based on the classification accuracy confusion matrix; (7) Compare the accuracy of the classification result with the set value. In response to the classification accuracy evaluation result satisfying the set value, output the lithofacies classification model. In response to the classification accuracy evaluation result not satisfying the set value, adjust the parameters of the gradient boosting decision tree model and retrain the gradient boosting decision tree model using the training data subset.
[0028] Step S130: the lithofacies classification model outputs the lithofacies classification result of the well to be predicted.
[0029] The present invention discloses a lithofacies classification prediction method based on a gradient boosting decision tree. The lithofacies classification model is obtained by machine learning using limited well logging data in combination with the gradient boosting decision tree model, and then the lithofacies classification model is used to realize automatic lithofacies classification prediction. Compared with the method of dividing lithofacies based on manual experience, the method has high efficiency and accuracy, and the sample data used for training can be continuously updated, thereby continuously updating the lithofacies classification model and improving the accuracy of lithofacies classification.
[0030] Example 2: Using the five lithofacies in Well X1 and Well X2 and their corresponding GR, AC, CAL, SP, RILD, and RFOC, a sample data set was established at a depth interval of 0.125 m to construct a lithofacies classification model, as follows: (I) Through observations of cores, cast thin sections, and scanning electron microscopes from Wells X1 and X2 in the study area, the lithofacies are divided into the following five types: Delta plain distributary channel (a): sandy sediments, with plate-like and trough-like cross-bedding, asymmetric ripples, and scour-fill structures; the sandstone is moderately compacted and moderately cemented, with a strong dissolution surface. Therefore, the lithofacies are characterized by moderate compaction, primary intercrystalline pore formation, no strong consolidation diagenesis, strong dissolution, and considerable secondary pore formation. Well logging characteristics are natural gamma GR (API): 68-72; natural unit SP (mV): 60-90; acoustic wave time difference AC (μs / m): 230-240; well diameter CAL (cm): 21.8-22.0; deep induction resistivity RILD (Ω·m): 15-25; eight-lateral resistivity RFOC (Ω·m): 17-30; Underwater distributary channel at the delta front (b): sand and silt sediments, cross-bedding and scour-fill structures; the sandstone is moderately compacted and cemented, with moderate dissolution surfaces. The lithofacies are characterized by moderate compaction, moderate dissolution and primary intercrystalline pore formation. The logging characteristics are GR (API): 65-85; SP (mV): 60-93; AC (μs / m): 230-245; CAL (cm): 21.8-22.4; RILD (Ω·m): 10-20; RFOC (Ω·m): 10-38; Mouth bar (c): fine sand deposits, trough cross-bedding, medium-thick fine sandstones with moderate compaction, strong consolidation, and moderate dissolution, as well as fine sandstones with strong compaction, moderate consolidation, and weak dissolution, and siltstones with strong compaction, moderate consolidation, and weak dissolution. Therefore, the typical characteristics of the lithofacies are moderate or strong consolidation, resulting in blocked pore throats. Cementation mainly includes the growth of carbonate minerals into pores ("pore type"), and strong matrix cementation ("matrix type"). The logging characteristics are GR (API): 95-105; SP (mV): 92-100; AC (μs / m): 220-240; CAL (cm): 22.0-22.2; RILD (Ω·m): 17-20; RFOC (Ω·m): 18-50; Frontal sheet sand (d): silt sand deposits, trough cross-bedding, thin layer thickness; siltstone has undergone strong compaction and moderate cementation, weak dissolution, as well as strong compaction, weak cementation, and weak dissolution. Therefore, this lithofacies is characterized by strong compaction, few secondary pores, weak cementation, and weak dissolution. Well logging characteristics are GR (API): 92-100; SP (mV): 92-00; AC (μs / m): 222-236; CAL (cm): 22.0-22.2; RILD (Ω·m): 18-22; RFOC (Ω·m): 30-38; Swamp microfacies (e): dark mudstone. Well logging characteristics are GR (API): 75-125; SP (mV): 80-110; AC (μs / m): 225-235; CAL (cm): 21.8-22.0; RILD (Ω·m): 10-20; RFOC (Ω·m): 15-45.
[0031] (II) The five lithofacies of Well X1 and Well X2 and their corresponding GR, AC, CAL, SP, RILD, and RFOC were used to establish a sample data set at a depth interval of 0.125 m. The logging data of different lithofacies were combined into a sample data set. The frequency statistics of lithofacies classification sample data are shown in the attached figure. Figure 2 shown.
[0032] (III) Divide the sample data set into training data subset and test data subset. Use the train_test_split module to select the feature subset Fm, and divide it into the training data subset Ftrain and the test data subset Ttest according to the ratio of 80% training data subset (500 data points) and 20% test data subset (125 data points).
[0033] (iv) Use the training data subset to train the gradient boosting decision tree model to obtain the lithofacies classification model, use the test data subset to test the lithofacies classification model, and compare the test results of each sample data in the test data subset with the corresponding real lithofacies type to generate a classification accuracy confusion matrix, as shown in the attached figure. Figure 3 As shown in the figure, it can be seen that for type a, type b and type e mudstone, high classification accuracies of 100%, 95% and 88% were obtained respectively.
[0034] (V) Well Y lithofacies was selected as a lithofacies prediction and classification verification well, and its logging characteristics were obtained. The well logging characteristics were input into the lithofacies classification model obtained by the above training, and the corresponding lithofacies classification results were output. The lithofacies classification results were compared with the manual classification prediction results. The comparison results are shown in the attached figure. Figure 4 As shown in the figure, it can be seen that the prediction accuracy of the lithofacies classification result obtained by using the lithofacies classification prediction method based on the gradient boosting decision tree of the present invention is higher.
[0035] Embodiment 3: As attached Figure 5 As shown, the embodiment of the present invention discloses a lithology classification prediction device based on a gradient boosting decision tree, comprising: A feature extraction unit, which obtains the logging features of the well to be predicted; A classification prediction unit inputs the logging characteristics of the well to be predicted into a lithofacies classification model, wherein the lithofacies classification model is obtained by training a gradient boosting decision tree model using multiple sets of sample data, and each set of sample data in the multiple sets of sample data includes: the logging characteristics of the historical well core sample and identification information used to identify the category of the historical well core sample; The result output unit, the lithofacies classification model outputs the lithofacies classification result of the well to be predicted.
[0036] Embodiment 4: As attached Figure 6 As shown, the embodiment of the present invention discloses a lithology classification prediction device based on a gradient boosting decision tree, which also includes a model building unit, including: A sample acquisition module constructs a sample data set, wherein the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; A subset partitioning module divides the sample data set into a training data subset and a test data subset; A training module, which trains a gradient boosting decision tree model using a subset of training data to obtain a lithofacies classification model; The testing module uses the test data subset to test the lithofacies classification model and evaluates the classification accuracy of the lithofacies classification model based on the test results; The evaluation module outputs a lithofacies classification model in response to the classification accuracy evaluation result meeting a set value, and adjusts the gradient boosting decision tree model parameters in response to the classification accuracy evaluation result not meeting the set value, and retrains the gradient boosting decision tree model using a training data subset.
[0037] Embodiment 5: As attached Figure 7 As shown, the embodiment of the present invention discloses a lithology classification prediction system based on a gradient boosting decision tree, comprising: A lithofacies classification prediction device based on a gradient boosting decision tree, wherein the lithofacies classification prediction device based on a gradient boosting decision tree is the lithofacies classification prediction device based on a gradient boosting decision tree described in the above embodiment; An interactive device is used for an operator to communicate with the lithofacies classification prediction device, so that the operator provides the lithofacies classification prediction device with an interactive interface for the person to be predicted through an interactive unit.
[0038] The above-mentioned interactive devices include computers, tablet computers, etc., which provide network pages for operators to enter information, view results, and modify parameters.
[0039] Embodiment 6: The embodiment of the present invention discloses a storage medium, on which a computer program readable by a computer is stored, and the computer program is configured to perform facies classification prediction based on a gradient boosting decision tree when running.
[0040] The above storage medium may include, but is not limited to: a USB flash drive, a read-only memory, a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.
[0041] Embodiment 7: The embodiment of the present invention discloses an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the computer program is loaded and executed by the processor to realize facies classification prediction based on a gradient boosting decision tree.
[0042] The processor may be a central processing unit (CPU), a general purpose processor, a digital signal processor (DSP), an ASIC, an FPGA or other programmable logic device, a transistor logic device, a hardware component or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. It may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like. The memory may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory, a mobile hard disk, a magnetic disk or an optical disk.
[0043] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.
[0044] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0045] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0046] The above technical features constitute the best embodiment of the present invention, which has strong adaptability and best implementation effect. Non-essential technical features can be added or reduced according to actual needs to meet the requirements of different situations.
Claims
1. A petrographic classification prediction method based on gradient boosting decision tree, characterized in that: include: Obtaining logging characteristics of the well to be predicted; Inputting the logging characteristics of the well to be predicted into the lithofacies classification model, wherein the lithofacies classification model is obtained by training the gradient boosting decision tree model using multiple sets of sample data, each set of sample data in the multiple sets of sample data includes: the logging characteristics of the historical well core sample and the identification information used to identify the category of the historical well core sample; The lithofacies classification model outputs the lithofacies classification results of the well to be predicted.
2. The method for petrographic classification prediction based on gradient boosting decision tree according to claim 1, characterized in that: The logging characteristics include natural gamma, natural unit, acoustic time difference, well diameter, deep induction resistivity, and eight lateral resistivities.
3. The petrographic classification prediction method based on gradient boosting decision tree according to claim 1 or 2, characterized in that: The process of constructing the lithofacies classification model includes: Constructing a sample data set, wherein the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; Divide the sample data set into a training data subset and a test data subset; The gradient boosting decision tree model is trained using the training data subset to obtain the lithofacies classification model; The lithofacies classification model is tested using a test data subset, and the classification accuracy of the lithofacies classification model is evaluated based on the test results; In response to the classification accuracy evaluation result meeting the set value, the lithofacies classification model is output; in response to the classification accuracy evaluation result not meeting the set value, the gradient boosting decision tree model parameters are adjusted, and the gradient boosting decision tree model is retrained using the training data subset.
4. The method for petrographic classification prediction based on gradient boosting decision tree according to claim 3 is characterized in that: The real lithofacies types include delta plain distributary channel lithofacies types, delta front underwater distributary channel lithofacies types, estuary bar lithofacies types, front sheet sandstone lithofacies types, and pro-delta mudstone lithofacies types.
5. The method for petrographic classification prediction based on gradient boosting decision tree according to claim 3 or 4, characterized in that: The classification accuracy evaluation of the gradient boosting decision tree model based on the test results includes: The test results of each sample data in the test data subset are compared with the corresponding real lithofacies type to generate a classification accuracy confusion matrix; The classification accuracy of the lithofacies classification model was evaluated based on the classification accuracy confusion matrix; The accuracy of the classification result is compared with the set value. In response to the classification accuracy evaluation result meeting the set value, the lithofacies classification model is output. In response to the classification accuracy evaluation result not meeting the set value, the parameters of the gradient boosting decision tree model are adjusted, and the gradient boosting decision tree model is retrained using the training data subset.
6. A lithology classification prediction device based on a gradient boosting decision tree using the method according to any one of claims 1 to 5, characterized in that: include: A feature extraction unit, which obtains the logging features of the well to be predicted; A classification prediction unit inputs the logging characteristics of the well to be predicted into a lithofacies classification model, wherein the lithofacies classification model is obtained by training a gradient boosting decision tree model using multiple sets of sample data, and each set of sample data in the multiple sets of sample data includes: the logging characteristics of the historical well core sample and identification information used to identify the category of the historical well core sample; The result output unit, the lithofacies classification model outputs the lithofacies classification result of the well to be predicted.
7. The device for predicting facies classification based on gradient boosting decision tree according to claim 6, characterized in that: Also included are model building units, including: A sample acquisition module constructs a sample data set, wherein the sample data set includes multiple real lithofacies types and logging characteristics of historical wells corresponding to each real lithofacies type; A subset partitioning module divides the sample data set into a training data subset and a test data subset; A training module, which trains a gradient boosting decision tree model using a subset of training data to obtain a lithofacies classification model; The testing module uses the test data subset to test the lithofacies classification model and evaluates the classification accuracy of the lithofacies classification model based on the test results; The evaluation module outputs a lithofacies classification model in response to the classification accuracy evaluation result meeting a set value, and adjusts the gradient boosting decision tree model parameters in response to the classification accuracy evaluation result not meeting the set value, and retrains the gradient boosting decision tree model using a training data subset.
8. A lithofacies classification prediction system based on gradient boosting decision tree, characterized in that: include: A petrographic classification prediction device based on a gradient boosting decision tree, wherein the petrographic classification prediction device based on a gradient boosting decision tree is the petrographic classification prediction device based on a gradient boosting decision tree as described in claims 6 to 7; An interactive device is used for an operator to communicate with the lithofacies classification prediction device, so that the operator provides the lithofacies classification prediction device with an interactive interface for the person to be predicted through an interactive unit.
9. A storage medium, characterized in that: The storage medium stores a computer program that can be read by a computer, and the computer program is configured to execute the lithology classification prediction method based on the gradient boosting decision tree as described in any one of claims 1 to 5 when running.
10. An electronic device, characterized in that: It comprises a processor and a memory, wherein a computer program is stored in the memory, and the computer program is loaded and executed by the processor to implement the lithology classification prediction method based on the gradient boosting decision tree as described in any one of claims 1 to 5.