Intelligent identification method for rock lithology of ore-bearing layer of sandstone type uranium ore

Abstract

The invention belongs to the field of data processing, and provides an intelligent identification method for rock lithology of an ore-bearing bed of a sandstone-type uranium deposit, and the method comprises the steps: building a random forest regression model and an XGBoost model; obtaining a logging curve sample set and lithologic data, dividing the logging curve sample set into original data and a verification set, and generating a first training set and a first test set; training to obtain a random forest regression model which is verified to be qualified; based on the random forest regression model which is verified to be qualified, obtaining filling data; based on the original data, the filling data and the lithology data, training to obtain a second XGBoost model and a third XGBoost model which are tested to be qualified; and based on the second XGBoost model and the third XGBoost model which are tested to be qualified, obtaining the lithology of the ore-bearing bed of the uranium ore deposit after identification. The lithology identified by the method is more in line with the substantial geological condition, the precision is higher, and the lithology misjudgment caused by the abnormal GR logging data is reduced.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an intelligent identification method for rock properties of ore-bearing layers of sandstone-type uranium deposits. Background Art

[0002] Sandstone-type uranium deposits are a kind of key performance metal and mineral resources. In the existing mineral exploration, geophysical logging methods (hereinafter referred to as logging) are widely used. Logging is also an effective method to understand and obtain underground geological structure and reservoir characteristics information. Logging data is of great value in reservoir evaluation, and can be used to divide stratigraphic structure, sedimentary phase, lithology identification, etc. Among them, lithology identification is of great significance in uranium deposit research. By identifying different lithologies under sedimentation, metamorphism or volcanism, the main host rocks of uranium deposits (such as sandstone, black shale) and their enrichment-related rock types can be clarified. At the same time, lithology identification helps to analyze the structural features that control the formation of uranium deposits, such as fault zones, folds and magmatic activity, thereby revealing the impact of tectonic activity on the enrichment of uranium deposits and providing a basis for determining the potential geological areas of uranium deposits.

[0003] Digital mapping is an important means of lithology identification. However, due to the depth limitation of surface geological mapping, it can only reflect the distribution of shallow lithology, and it is difficult to identify deep lithology. In the covered area, it cannot give the changes of underlying lithology. In order to overcome the shortcomings of surface geological mapping, geophysics, remote sensing and other technologies have been gradually introduced into lithology identification, and great progress has been made. However, at present, it still faces the technical problems of low lithology identification accuracy and serious lithology misjudgment.

[0004] At the same time, due to factors such as poor drilling conditions, instrument failures, differences in logging conditions, and improper storage, some well logging data sections are easily missing or distorted, affecting the indication of lithology and fluids, bringing difficulties to logging evaluation and interpretation, and even abandoning some logging data due to exploration cost considerations. Given that re-logging is limited by cost and wellbore conditions, the need for low-cost reconstruction of missing logging curves is becoming more urgent. In order to solve the problem of missing logging curves without adding additional manpower and economic costs, technicians in this field have tried a variety of methods, such as traditional empirical models based on rock physics, multivariate regression analysis, and physical model inversion to generate artificial logging curves. However, these methods usually oversimplify formation information, rely on subjective experience, and are difficult to effectively reflect the complex nonlinear relationship between logging data. When faced with the challenges of strong formation heterogeneity and complex downhole conditions, traditional linear methods perform poorly.

[0005] Therefore, exploring and establishing more efficient logging curve reconstruction methods and improving the accuracy of lithology identification are crucial to improving the regional logging database and improving the accuracy of geophysical exploration interpretation. Summary of the invention

[0006] In view of the problems existing in the prior art, the present application provides an intelligent identification method for the rock lithology of the ore-bearing strata of sandstone-type uranium deposits, so as to solve the technical problems of poor reconstruction effect of existing logging curves and inaccurate lithology identification.

[0007] In a first aspect, an embodiment of the present application provides a method for intelligently identifying the rock lithology of a sandstone-type uranium ore-bearing layer, the method comprising: Step S1: Establish random forest regression model and XGBoost model; Step S2: obtaining a well logging curve sample set and lithology data, dividing the well logging curve sample set into original data and a validation set, and generating a first training set and a first test set based on the original data; Step S3: training the random forest regression model based on the first training set and the first test set to obtain a verified random forest regression model; obtaining a predicted filling GR curve based on the verification set and the verified random forest regression model; Step S4: Fill the predicted filling GR curve into the original data to obtain filling data; Step S5: generating a second training set, a second test set, a third training set, and a third test set based on the original data, the filling data, and the lithology data; Step S6: training the XGBoost model based on the second training set, the second test set, the third training set, and the third test set, respectively, to obtain a second XGBoost model that passes the test and a third XGBoost model that passes the test; Step S7: input the original data into the second XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the original identification; input the filled data into the third XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the filled identification.

[0008] In a possible implementation, the hyperparameters of the random forest regression model include the number of trees n_estimators, the maximum depth max_depth, the minimum number of sample splits min_samples_split, the minimum number of leaf node samples min_samples_leaf, and the maximum number of features max_features; The evaluation indicators of the random forest regression model include mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and R-square value of the prediction model fitting accuracy. .

[0009] In a possible implementation, the hyperparameters of the XGBoost model include the number of trees n_estimators, the maximum depth max_depth, the learning rate learning_rate, the subsampling parameter subsample, the proportion of the features of the logging curve samples used to train each tree to all features colsample_bytree, the minimum weight and min_child_weight of a leaf node, and the minimum loss reduction gamma required to split a node; The evaluation indicators of the XGBoost model include accuracy, precision, recall, harmonic mean of precision and recall, and confusion matrix.

[0010] In a possible implementation, step S2 is specifically as follows: Obtain well logging curve sample sets and lithology data; Classifying the logging curve samples in the logging curve sample set according to the data completeness of the logging curve samples in the logging curve sample set, wherein the set of logging curve samples with complete data is the original data, and the set of logging curve samples with missing data is the verification set; The apparent resistivity curve, well diameter curve, well temperature curve, natural potential curve and three-lateral resistance curve of the logging curve samples in the original data are used as the first input data set, and the natural gamma logging curve is used as the first label data set. The data of the first input data set are paired with the corresponding natural gamma logging curve in the first label data set to obtain a first logging data pair set; the first logging data pair set is divided into a first training set and a first test set in a ratio of 8:2.

[0011] In a possible implementation, step S3 is specifically as follows: Define the hyperparameter range of the random forest regression model. Set the number of trees n_estimators to (50, 500), the maximum depth to max_depth to (5, 30), the minimum number of sample splits to min_samples_split to (2, 10), the minimum number of leaf node samples to min_samples_leaf to (1, 5), and the maximum number of features to max_features to (0.5, 1.0), and set the random forest regression model to default; Initialize the Bayesian optimization algorithm, select the random forest default model for the evaluator, the hyperparameter range of the random forest regression model for the search range, 200 for the number of search iterations, 5 for cross validation, negative mean square error for scoring, and 66 for the random number; The first training set is searched using the Bayesian optimization algorithm, and after obtaining the optimal hyperparameters, the performance of the random forest regression model is evaluated using the first test set to obtain a verified and qualified random forest regression model; The validation set is input into the verified random forest regression model to obtain the predicted filling GR curve.

[0012] In a possible implementation, step S4 is specifically as follows: deleting the abnormal part of the natural gamma logging curve in the logging curve sample of the original data, and then adding the part of the predicted filled GR curve corresponding to the abnormal part of the natural gamma logging curve to the natural gamma logging curve from which the abnormal part is deleted, to obtain the filled data.

[0013] In a possible implementation, step S5 is specifically as follows: The apparent resistivity curve, natural gamma curve, well diameter curve, well temperature curve, natural potential curve, and three-lateral resistance curve of the original data and the filled data are used as the second input data set and the third input data set respectively, and the lithology data is used as the label data set; Pairing the data in the second input data set with the corresponding lithology data in the label data set to obtain a second well logging data pair set; Pairing the data in the third input data set with the corresponding lithology data in the label data set to obtain a third well logging data pair set; Dividing the second well logging data pair set and the third well logging data pair set into a second training set, a second test set and a third training set, a third test set in a ratio of 7:3 respectively; The lithology data include mudstone, siltstone, fine sandstone, medium sandstone, coarse sandstone, and conglomerate.

[0014] In a possible implementation, step S6 is specifically as follows: The second training set, the second test set, the third training set, and the third test set are preprocessed, and the XGBoost model is set by default to not use the target variable for label encoding, and the evaluation indicator of the XGBoost model is the logistic loss; Define the hyperparameter range of the XGBoost model, set the number of trees n_estimators to (50, 500), the maximum depth max_depth to (3, 10), the learning rate learning_rate to (0.01, 0.3, 'log-uniform'), the subsampling parameter subsample to (0.5, 1.0), the proportion of the features of the logging curve samples used to train each tree to the total features colsample_bytree to (0.5, 1.0), the minimum weight of a leaf node and min_child_weight to (1, 10), and the minimum loss reduction gamma required to split a node to (0, 5); Initialize the XGBoost Bayesian algorithm, select the XGBooost model as the evaluator, the hyperparameter range of the XGBoost model as the search range, 50 as the search iterations, 3 as the cross validation, accuracy as the score, and 45 as the random number; The XGBoost Bayesian algorithm was used to search the second training set and the third training set respectively. After obtaining the optimal hyperparameters, performance evaluation was performed on the second test set and the third test set respectively to obtain a qualified second XGBoost model and a qualified third XGBoost model.

[0015] In a possible implementation, the preprocessing is to use standardization to subject the well logging curve samples to a uniform standard normal distribution.

[0016] In a possible implementation, when evaluating the performance of the random forest regression model, the R-square value of the prediction model fitting accuracy is used. As the first evaluation metric for random forest regression models.

[0017] Based on the above invention content, compared with the prior art, this application has achieved the following technical effects: (1) The present invention obtains a predicted filled GR curve by standardizing all input logging curves and fitting the missing values ​​of GR logging data by a random forest regression model. The predicted filled GR curve is used to fill the abnormal part of the natural gamma ray curve in the original data to obtain filled data. The filled data is used to identify lithology. The identified lithology is more consistent with the actual geological conditions and has higher accuracy, thereby successfully reducing the misjudgment of lithology caused by abnormal GR logging data.

[0018] (2) The present invention improves the computational efficiency of data reconstruction and lithology identification by selecting reasonable sample data, using a random forest regression model for data reconstruction, and using an XGBoost model for lithology identification.

[0019] (3) During the model training process, the present invention uses an advanced Bayesian optimization algorithm to find the optimal hyperparameters, which has obvious advantages over the traditional grid search in terms of low efficiency. In addition, during the parameter adjustment process, multiple evaluation indicators are combined for evaluation, which further improves the accuracy of the hyperparameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 It is a schematic flow chart of a method for intelligently identifying rock properties of a sandstone-type uranium ore-bearing layer provided in an embodiment of the present application; Figure 2 A schematic diagram of filling data provided by an embodiment of the present invention; Figure 3 A comparison chart of the classification results of the original data and the filled data of the Z1 well provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0023] It should be clear that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] See also Figure 1 , is a flow chart of a method for intelligently identifying the rock properties of a sandstone-type uranium ore-bearing layer provided by an embodiment of the present invention. Figure 1 As shown, the method specifically includes: Step S1: Establish a random forest regression model and an XGBoost model.

[0025] The hyperparameters of the random forest regression model include the number of trees n_estimators, the maximum depth max_depth, the minimum number of sample splits min_samples_split, the minimum number of leaf nodes min_samples_leaf, and the maximum number of features max_features.

[0026] The evaluation indicators of the random forest regression model include mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and R-square value of the prediction model fitting accuracy. The smaller the mean absolute error (MAE), the smaller the absolute value of the model's prediction error is on average, and the higher the model's accuracy is. The smaller the mean square error (MSE), the closer the predicted value is to the true value. The smaller the root mean square error (RMSE), the smaller the model's prediction error is, and the closer the predicted value is to the true value. The closer it is to 1, the better the fitting effect of the random forest regression model is and the better the overall performance is.

[0027] The hyperparameters of the XGBoost model include the number of trees n_estimators, the maximum depth max_depth, the learning rate learning_rate, the subsampling parameter subsample, the proportion of the features of the logging curve samples used to train each tree to the total features colsample_bytree, the minimum weight and min_child_weight of a leaf node, and the minimum loss reduction gamma required to split a node.

[0028] The evaluation indicators of the XGBoost model include accuracy, precision, recall, harmonic mean of precision and recall, and confusion matrix.

[0029] Accuracy: It indicates the proportion of correctly predicted samples to the total samples.

[0030] Precision: It indicates the proportion of samples predicted to be positive that are actually positive.

[0031] Recall: It indicates the proportion of samples that are correctly predicted to be positive among the samples that are actually positive.

[0032] F1-score: represents the harmonic mean of precision and recall, and is suitable for use when precision and recall need to be balanced.

[0033] Confusion Matrix: It shows the detailed distribution of prediction results in matrix form.

[0034] Step S2: Obtain a well logging curve sample set and lithology data, divide the well logging curve sample set into original data and a validation set, and generate a first training set and a first test set based on the original data. The details are as follows: Step S21: Acquire a logging curve sample set and lithology data.

[0035] Step S22: classifying the logging curve samples in the logging curve sample set according to the data completeness of the logging curve samples in the logging curve sample set, wherein the set of logging curve samples with complete data is the original data, and the set of logging curve samples with missing data is the verification set.

[0036] Step S23: using the apparent resistivity curve (RES), well diameter curve (WD), well temperature curve (WT), natural potential curve (SP) and tri-lateral resistance curve (TAR) of the logging curve samples in the original data as the first input data set, and using the natural gamma logging curve (GR) as the first label data set, pairing the data of the first input data set with the corresponding natural gamma logging curve in the first label data set to obtain a first logging data pair set; dividing the first logging data pair set into a first training set and a first test set in a ratio of 8:2.

[0037] The data used in this embodiment are derived from the rock lithology and corresponding logging curve sample sets provided by the northern region of the Songliao Basin. The logging curve sample set consists of three uranium mineralized industrial holes, with well numbers Z1, Z2, and Z3, respectively. The data of the mineralized section of the lower section of the Yaojia Formation were selected as logging curve samples, totaling 3375 sets of logging curve samples. These logging curve samples cover a variety of rock types, including mudstone, siltstone, fine sandstone, medium sandstone, coarse sandstone, and conglomerate, among which sandstone is a good uranium reservoir. Based on the above data, the distribution of the first training set, the first test set, and the validation set is shown in Table 1.

[0038] Table 1: Step S3: Based on the first training set and the first test set, the random forest regression model is trained to obtain a verified random forest regression model; based on the verification set and the verified random forest regression model, a predicted filling GR curve is obtained. Specifically including: Step S31: Define the hyperparameter range of the random forest regression model. Set the number of trees n_estimators to (50, 500), the maximum depth to max_depth to (5, 30), the minimum sample split number to min_samples_split to (2, 10), the minimum number of leaf node samples to min_samples_leaf to (1, 5), and the maximum number of features to max_features to (0.5, 1.0), and set the random forest regression model to default.

[0039] Step S32: Initialize the Bayesian optimization algorithm, select the random forest default model for the estimator, select the hyperparameter range of the random forest regression model for the search range (search_spaces), select 200 for the number of search iterations (n_iter), select 5 for cross validation (cv), select negative mean square error (neg_mean_squared_eeror) for scoring, and select 66 for the random number.

[0040] Step S33: Use the Bayesian optimization algorithm to search the first training set, and after obtaining the optimal hyperparameters, use the first test set to perform performance evaluation on the random forest regression model to obtain a verified and qualified random forest regression model.

[0041] When evaluating the performance of the random forest regression model, the R-square value of the prediction model fitting accuracy is used. As the first evaluation index of the random forest regression model. In this embodiment, when the logging curve samples of the random forest regression model of the Z1 well are adjusted to the optimal hyperparameters, the evaluation index values ​​of the random forest regression model are shown in Table 2.

[0042] Table 2: Step S34: input the verification set into the verified random forest regression model to obtain a predicted filling GR curve.

[0043] Step S4: Fill the predicted filling GR curve into the original data to obtain filling data. Specifically: The abnormal part of the natural gamma logging curve (GR) in the logging curve sample of the original data is deleted, and then the part of the predicted filled GR curve corresponding to the abnormal part of the natural gamma logging curve (GR) is added to the natural gamma logging curve (GR) with the abnormal part deleted to obtain the filled data.

[0044] To illustrate the effect of padding data, see Figure 2 , is a schematic diagram of the filling data of the Z1 well provided in an embodiment of the present invention. Figure 2 As shown in the figure, there are three color waveforms. The red dotted waveform is the abnormal part of the natural gamma logging curve (GR) in the logging curve sample of the original data, the black waveform is the natural gamma logging curve (GR) with the abnormal part deleted, the green waveform on the right is the predicted filled GR curve, and the green waveform on the left is the filled data natural gamma logging curve (GR). Figure 2It can be seen from the figure that in Well Z1, the abnormal value of the natural gamma logging curve has only one obvious peak, but there are different peaks, and the span is large. The highest GR abnormal value is as high as 4033.33API (due to the large abnormal value, Figure 2 The missing data are mainly in the first peak and a few local outliers, corresponding to the well depths of 545.75-546.4m, 551.1-551.6m, 556.65-563.2, 565.5-565.85, 566.2-580.2 to 591.15-593.25, a total of 24.15m. The data filled in the peaks show completely different data volumes, with peaks and local minima. The predicted filled data are consistent with the complex underground geological environment, and the fit is also similar.

[0045] Step S5: Based on the original data, the filling data and the lithology data, a second training set, a second test set and a third training set and a third test set are generated. Specifically: The apparent resistivity curve (RES), natural gamma curve (GR), well diameter curve (WD), well temperature curve (WT), natural potential curve (SP), and three lateral resistance curve (TAR) of the original data and the filling data are respectively used as the second input data set and the third input data set, and the lithology data is used as the label data set; Pairing the data in the second input data set with the corresponding lithology data in the label data set to obtain a second well logging data pair set; Pairing the data in the third input data set with the corresponding lithology data in the label data set to obtain a third well logging data pair set; The second logging data pair set and the third logging data pair set are divided into a second training set, a second test set and a third training set, a third test set in a ratio of 7:3 respectively.

[0046] The lithology data include mudstone, siltstone, fine sandstone, medium sandstone, coarse sandstone, and conglomerate.

[0047] Step S6: Train the XGBoost model based on the second training set, the second test set and the third training set, and the third test set to obtain an XGBoost model that passes the test.

[0048] Step S6: Train the XGBoost model based on the second training set, the second test set, the third training set, and the third test set, respectively, to obtain a second XGBoost model that passes the test and a third XGBoost model that passes the test. Specifically: Step S61: preprocess the second training set, the second test set and the third training set, and the third test set, and set the XGBoost model by default to not use the target variable for label encoding, and the evaluation indicator of the XGBoost model is logical loss.

[0049] The preprocessing is to use standardization to make a uniform standard normal distribution for the logging curve samples.

[0050] Step S62: Define the hyperparameter range of the XGBoost model, set the number of trees n_estimators to the range of (50, 500), the maximum depth max_depth to the range of (3, 10), the learning rate learning_rate to the range of (0.01, 0.3, 'log-uniform'), the subsampling parameter subsample to the range of (0.5, 1.0), the proportion of the features of the logging curve samples used to train each tree to the total features colsample_bytree to the range of (0.5, 1.0), the minimum weight and min_child_weight of a leaf node to the range of (1, 10), and the minimum loss reduction gamma required to split a node to the range of (0, 5).

[0051] Step S63: Initialize the XGBoost Bayesian algorithm, select the XGBooost model for the estimator, select the hyperparameter range of the XGBoost model for the search range (search_spaces), select 50 for the number of search iterations (n_iter), select 3 for cross validation (cv), select accuracy for scoring, and select 45 for the random number.

[0052] Step S64: Use the XGBoost Bayesian algorithm to search the second training set and the third training set respectively, and after obtaining the optimal hyperparameters, perform performance evaluation on the second test set and the third test set respectively to obtain a second XGBoost model that passes the test and a third XGBoost model that passes the test.

[0053] Step S7: input the original data into the second XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the original identification; input the filled data into the third XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the filled identification.

[0054] See also Figure 3 , is a comparison chart of the classification results of the original data and the filled data of the Z1 well provided in the embodiment of the present invention, wherein, Figure 3 The middle left picture shows the classification result of the original data of Well Z1. Figure 3The middle right picture shows the classification result of the filling data of Well Z1. Figure 3 As shown in the middle left figure, the lithology recognition accuracy of the original well logging curve to be processed in Well Z1 is 97.3%. Among the input samples, there are 53 mudstones, 32 siltstones, 138 fine sandstones, 143 medium sandstones, and 16 coarse sandstones, with a total of 382 samples. The samples that are correctly predicted on the main team angle line are 51 mudstones, 30 siltstones, 135 fine sandstones, 1418 medium sandstones, and 15 coarse sandstones, with a total of 372 samples. Figure 3 From the middle left figure, it can be concluded that mudstone and fine sandstone perform best in the model, and the other three lithologies perform equally well: siltstone, medium sandstone, and coarse sandstone.

[0055] like Figure 3 As shown in the middle right figure, the lithology recognition accuracy of the Z1 well filling data to be processed is 98.6%. Among the input samples, there are 66 mudstones, 31 siltstones, 144 fine sandstones, 130 medium sandstones, and 11 coarse sandstones, with a total of 382 samples. The samples located on the main team angle line are correctly predicted to be 65 mudstones, 31 siltstones, 143 fine sandstones, 127 medium sandstones, and 11 coarse sandstones, with a total of 377. Figure 3 It can be seen from the middle right figure that coarse sandstone performs best in the model, followed by fine sandstone, medium sandstone and mudstone, and the worst is siltstone.

[0056] In the embodiment of the present application, all input logging curves are standardized and the missing values ​​of GR logging data are fitted by a random forest regression model to obtain a predicted filled GR curve, and the predicted filled GR curve is used to fill the abnormal part of the natural gamma curve in the original data to obtain the filled data. The filled data is used for lithology identification, and the identified lithology is more consistent with the actual geological conditions and has higher accuracy, which successfully reduces the misjudgment of lithology caused by abnormal GR logging data. By selecting reasonable sample data, using a random forest regression model for data reconstruction, and using an XGBoost model for lithology identification, the computational efficiency of data reconstruction and lithology identification is improved. In the process of model training, an advanced Bayesian optimization algorithm is used to find the best hyperparameters, which has obvious advantages over the low efficiency of traditional grid search, and in the process of parameter adjustment, multiple evaluation indicators are combined for evaluation to further improve the accuracy of hyperparameters.

[0057] The above is only a specific implementation of the present application. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. The protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent identification method for rock properties of sandstone-type uranium ore-bearing strata, characterized in that: The method comprises: Step S1: Establish random forest regression model and XGBoost model; Step S2: obtaining a well logging curve sample set and lithology data, dividing the well logging curve sample set into original data and a validation set, and generating a first training set and a first test set based on the original data; Step S3: training the random forest regression model based on the first training set and the first test set to obtain a verified random forest regression model; obtaining a predicted filling GR curve based on the verification set and the verified random forest regression model; Step S4: Fill the predicted filling GR curve into the original data to obtain filling data; Step S5: generating a second training set, a second test set, a third training set, and a third test set based on the original data, the filling data, and the lithology data; Step S6: training the XGBoost model based on the second training set, the second test set, the third training set, and the third test set, respectively, to obtain a second XGBoost model that passes the test and a third XGBoost model that passes the test; Step S7: input the original data into the second XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the original identification; input the filled data into the third XGBoost model that has passed the test to obtain the lithology of the ore-bearing layer of the uranium deposit after the filled identification.

2. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The hyperparameters of the random forest regression model include the number of trees n_estimators, the maximum depth max_depth, the minimum number of sample splits min_samples_split, the minimum number of leaf node samples min_samples_leaf, and the maximum number of features max_features; The evaluation indicators of the random forest regression model include mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and R-square value of the prediction model fitting accuracy. .

3. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The hyperparameters of the XGBoost model include the number of trees n_estimators, the maximum depth max_depth, the learning rate learning_rate, the subsampling parameter subsample, the proportion of the features of the logging curve samples used to train each tree to the total features colsample_bytree, the minimum weight of a leaf node and min_child_weight, and the minimum loss reduction gamma required to split a node; The evaluation indicators of the XGBoost model include accuracy, precision, recall, harmonic mean of precision and recall, and confusion matrix.

4. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The step S2 is specifically as follows: Obtain well logging curve sample sets and lithology data; Classifying the logging curve samples in the logging curve sample set according to the data completeness of the logging curve samples in the logging curve sample set, wherein the set of logging curve samples with complete data is the original data, and the set of logging curve samples with missing data is the verification set; The apparent resistivity curve, well diameter curve, well temperature curve, natural potential curve and three-lateral resistance curve of the logging curve samples in the original data are used as the first input data set, and the natural gamma logging curve is used as the first label data set. The data of the first input data set are paired with the corresponding natural gamma logging curve in the first label data set to obtain a first logging data pair set; the first logging data pair set is divided into a first training set and a first test set in a ratio of 8:

2.

5. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The step S3 is specifically as follows: Define the hyperparameter range of the random forest regression model, set the number of trees n_estimators to (50, 500), the maximum depth to max_depth to (5, 30), the minimum number of sample splits to min_samples_split to (2, 10), the minimum number of leaf node samples to min_samples_leaf to (1, 5), and the maximum number of features to max_features to (0.5, 1.0), and set the random forest regression model to default; Initialize the Bayesian optimization algorithm, select the random forest default model for the evaluator, the hyperparameter range of the random forest regression model for the search range, 200 for the number of search iterations, 5 for cross validation, negative mean square error for scoring, and 66 for the random number; The first training set is searched using the Bayesian optimization algorithm, and after obtaining the optimal hyperparameters, the performance of the random forest regression model is evaluated using the first test set to obtain a verified and qualified random forest regression model; The validation set is input into the verified random forest regression model to obtain the predicted filling GR curve.

6. The intelligent identification method of rock properties of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The step S4 specifically includes: deleting the abnormal part of the natural gamma logging curve in the logging curve sample of the original data, and then adding the part of the predicted filled GR curve corresponding to the abnormal part of the natural gamma logging curve to the natural gamma logging curve with the abnormal part deleted to obtain the filled data.

7. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: Step S5 is specifically as follows: The apparent resistivity curve, natural gamma curve, well diameter curve, well temperature curve, natural potential curve, and three-lateral resistance curve of the original data and the filled data are used as the second input data set and the third input data set respectively, and the lithology data is used as the label data set; Pairing the data in the second input data set with the corresponding lithology data in the label data set to obtain a second well logging data pair set; Pairing the data in the third input data set with the corresponding lithology data in the label data set to obtain a third well logging data pair set; Dividing the second well logging data pair set and the third well logging data pair set into a second training set, a second test set and a third training set, a third test set in a ratio of 7:3 respectively; The lithology data include mudstone, siltstone, fine sandstone, medium sandstone, coarse sandstone, and conglomerate.

8. The intelligent identification method of rock lithology of sandstone-type uranium ore-bearing strata according to claim 1 is characterized in that: The step S6 is specifically as follows: The second training set, the second test set, the third training set, and the third test set are preprocessed, and the XGBoost model is set by default to not use the target variable for label encoding, and the evaluation indicator of the XGBoost model is the logistic loss; Define the hyperparameter range of the XGBoost model, set the number of trees n_estimators to (50, 500), the maximum depth max_depth to (3, 10), the learning rate learning_rate to (0.01, 0.3, 'log-uniform'), the subsampling parameter subsample to (0.5, 1.0), the proportion of the features of the logging curve samples used to train each tree to the total features colsample_bytree to (0.5, 1.0), the minimum weight of a leaf node and min_child_weight to (1, 10), and the minimum loss reduction gamma required to split a node to (0, 5); Initialize the XGBoost Bayesian algorithm, select the XGBooost model as the evaluator, the hyperparameter range of the XGBoost model as the search range, 50 as the number of search iterations, 3 as the cross validation, accuracy as the score, and 45 as the random number; The XGBoost Bayesian algorithm was used to search the second training set and the third training set respectively. After obtaining the optimal hyperparameters, performance evaluation was performed on the second test set and the third test set respectively to obtain a qualified second XGBoost model and a qualified third XGBoost model.

9. The intelligent identification method of rock properties of sandstone-type uranium ore-bearing strata according to claim 8 is characterized in that: The preprocessing is to use standardization to make the well logging curve samples uniformly distributed.

10. The intelligent identification method of rock properties of sandstone-type uranium ore-bearing strata according to claim 5 is characterized in that: When evaluating the performance of the random forest regression model, the R-square value of the prediction model fitting accuracy is used. As the first evaluation metric for random forest regression models.

Images