Solid asphalt content prediction method and system based on gradient lifting regression method

Abstract

The invention provides a solid asphalt content prediction method and system based on a gradient lifting regression method, and the method comprises the steps: collecting a casting body sheet of a single well target layer, carrying out the image analysis of the casting body sheet, and obtaining the asphalt content data of different depths; collecting well logging data of a single well, and constructing the asphalt content data and the well logging data into a deep matching data set; training a gradient lifting regression model through the data set to obtain a prediction model; and performing solid asphalt content prediction on the uncored reservoir section of the to-be-detected single well through the prediction model to obtain a prediction result. According to the method, the complex nonlinear relation in the logging data can be learned and fitted, and the content of the solid asphalt can be predicted more accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of petroleum geological exploration, and in particular to a method and system for predicting solid asphalt content based on a gradient boosting regression method. Background Art

[0002] Ancient oil reservoirs refer to oil reservoirs that were once formed in the geological history, and later underwent transformation to form new oil and gas reservoirs or suffered destruction due to tectonic movement, temperature rise, etc. For deep-ultra-deep layers, ancient oil reservoirs undergo in-situ cracking due to the increase in burial temperature to form ancient gas reservoirs, and gradually form modern gas reservoirs. Ancient gas reservoirs formed by the cracking of ancient oil reservoirs provide sufficient gas sources for deep-ultra-deep gas reservoirs. Among the deep-ultra-deep gas reservoirs discovered today, most have been confirmed to be formed by the cracking of ancient oil reservoirs. The original scale and distribution of ancient oil reservoirs often directly determine the resource potential and distribution of modern gas reservoirs. Therefore, accurate identification and quantification of ancient oil reservoirs are crucial to the accumulation and evolution process of deep-ultra-deep oil and gas reservoirs and natural gas exploration.

[0003] Solid asphalt in the reservoir is the direct evidence of the existence of ancient oil reservoirs and is an important medium for studying ancient oil reservoirs. Previous studies mainly used core data to directly evaluate the content of solid asphalt in the reservoir. Due to the scarcity of core data, scholars have proposed a method to identify and predict the content of solid asphalt through logging data. Among them, advanced logging technologies such as nuclear magnetic resonance and array acoustic waves can well identify and predict the content of solid asphalt, but since these logging projects are rarely implemented in actual exploration wells, their application scope is significantly restricted. Although the method of combining multi-mineral inversion model with conventional logging has good practicality, the model involves many parameters, and the selection of these parameters is often affected by regional and human factors, resulting in complexity and uncertainty in the application of the model. Summary of the invention

[0004] The present invention provides a solid asphalt content prediction method and system based on the gradient boosting regression method to address the defects of the prior art. The invention aims to develop a method based on conventional logging data and utilizing high-precision machine learning to significantly reduce the influence of human factors, thereby achieving objectivity and accuracy in predicting the solid asphalt content of uncored well sections, and providing a basic basis for the quantitative identification of ancient oil reservoirs.

[0005] The present invention provides a method for predicting solid asphalt content based on a gradient boosting regression method, comprising: S1: Collecting cast thin sections of target layers in a single well, performing image analysis on the cast thin sections, and obtaining asphalt content data at different depths; S2: collecting well logging data of a single well, and constructing the bitumen content data and the well logging data into a depth matching data set; S3: training a gradient boosting regression model using the data set to obtain a prediction model; S4: predicting the solid asphalt content of the uncored reservoir section of the single well to be tested by using the prediction model to obtain a prediction result.

[0006] According to a solid asphalt content prediction method based on gradient boosting regression method provided by the present invention, step S1 also includes: S11: Correcting the asphalt content data.

[0007] According to a solid asphalt content prediction method based on gradient boosting regression provided by the present invention, step S11 further comprises: S111: performing image analysis on the casting thin slice to obtain residual porosity; S112: performing physical property analysis on the casting thin slice at the same depth to obtain gas-measured porosity; S113: establishing a linear regression relationship according to the residual porosity and the gas porosity; S114: Correcting the asphalt content data through the linear regression relationship.

[0008] According to a solid asphalt content prediction method based on gradient boosting regression provided by the present invention, the expression of the asphalt content data in step S1 is:

[0009] in, To calculate the asphalt content data at the current depth, is the solid asphalt pixel value extracted at the current depth, is the original image pixel value corresponding to the casting slice.

[0010] According to a solid asphalt content prediction method based on gradient boosting regression method provided by the present invention, the logging data in step S2 includes: gamma logging data, original formation resistivity logging data, flushing zone resistivity logging data, compensated density logging data, compensated acoustic logging data, and compensated neutron logging data.

[0011] According to a solid asphalt content prediction method based on gradient boosting regression method provided by the present invention, step S2 also includes: S21: performing preprocessing including denoising, clearing and normalization on the logging data.

[0012] According to a solid asphalt content prediction method based on gradient boosting regression provided by the present invention, step S3 further comprises: S31: dividing the data set into a training set and a validation set; S32: training a gradient boosting regression model using the training set to obtain a fitting model; S33: Validate the fitting model through the validation set to obtain a prediction model.

[0013] According to a solid asphalt content prediction method based on gradient boosting regression method provided by the present invention, step S32 further includes: S321: Establish a gradient boosting regression model; S322: Setting training parameters of the gradient boosting regression model, wherein the training parameters include at least the number of decision trees and random seeds; S322: Inputting the training set into the gradient boosting regression model for iterative training to reduce the prediction value residual of the gradient boosting regression model and obtain a fitting model.

[0014] According to a solid asphalt content prediction method based on gradient boosting regression provided by the present invention, step S33 further includes: S331: inputting the validation set into the fitting model; S332: Evaluate multiple performance indicators of the fitting model through mean square error and determination coefficient to obtain evaluation results; S333: According to the evaluation results, the number of decision trees and random seeds are selected through cross-validation, and the corresponding fitting model is used as the prediction model.

[0015] The present invention also provides a solid asphalt content prediction system based on the gradient boosting regression method, which is used for the solid asphalt content prediction method based on the gradient boosting regression method as described in any one of the above items, comprising: Analysis module: used to collect cast thin sections of target layers in a single well, perform image analysis on the cast thin sections, and obtain asphalt content data at different depths; Acquisition module: used to collect the logging data of a single well, and construct the asphalt content data and the logging data into a depth-matched data set; Training module: used to train the gradient boosting regression model through the data set to obtain a prediction model; Prediction module: used to predict the solid asphalt content of the uncored reservoir section of the single well to be tested through the prediction model obtained by the training module to obtain the prediction result.

[0016] The present invention provides a solid asphalt content prediction method and system based on the gradient boosting regression method. By using multiple decision trees as base learners and an integrated learning method, the method and system can effectively handle nonlinear relationships and are particularly suitable for prediction tasks of complex geological data sets, thereby improving the accuracy of prediction. In addition, the GRB algorithm integrates multiple weak learners and combines strategies such as learning rate adjustment and early stopping to effectively control the risk of overfitting in the model training process, significantly improve the generalization ability of the model, and make the prediction results more robust and reliable. Secondly, the prediction model of the present invention has a good tolerance for outlier data, can reduce the impact of abnormal data on prediction results, improve the stability and reliability of prediction results, and can make full use of the characteristic information of all relevant elements in the study area, and further improve the prediction accuracy and comprehensiveness by deeply mining the potential relationships between data.

[0017] The single-well solid asphalt content prediction method of the present invention aims to improve the accuracy of single-well solid asphalt prediction by applying a gradient boosting regression algorithm. Generally speaking, the present invention optimizes the prediction process, which not only improves the prediction speed and efficiency, but also reduces the need for experimental analysis, thereby simplifying the operation steps. Therefore, the present invention can provide a more accurate geological basis for oil and gas exploration, and effectively address the problems of complex prediction methods and low computational efficiency in the process of predicting the solid asphalt content of uncored well sections. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A schematic flow chart of a method for predicting solid asphalt content based on a gradient boosting regression method provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a solid asphalt content prediction system based on a gradient boosting regression method provided in an embodiment of the present invention; Figure 3 A linear regression relationship diagram between the residual porosity obtained by the image analysis method and the gas porosity obtained by the physical property analysis provided in an embodiment of the present invention; Figure 4 A result diagram of the accuracy evaluation of the single-well solid asphalt content prediction method based on the gradient boosting regression algorithm provided in an embodiment of the present invention on a validation set; Figure 5A graph showing the prediction results of the solid asphalt content in a reservoir in an uncored well section by a single well solid asphalt content prediction method based on a gradient boosting regression method provided in an embodiment of the present invention.

[0020] Reference numerals: 100, analysis module; 200, acquisition module; 300, training module; 400, prediction module. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. 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. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0022] Combine the following Figures 1 to 5 Embodiments of the present invention are described.

[0023] like Figure 1 As shown, the present invention provides a solid asphalt content prediction method based on gradient boosting regression method, comprising: S1: Collect cast thin sections of target layers in a single well, perform image analysis on the cast thin sections, and obtain asphalt content data at different depths.

[0024] In step S1, firstly, a cast thin section of the reservoir of the core section of the target layer of a single well is collected or prepared, and the asphalt content data of the reservoir of the core section of the target layer at different depths can be obtained by using a cast thin section image analysis and processing method with the aid of MATLAB software.

[0025] Wherein, step S1 also includes: S11: Correcting the asphalt content data.

[0026] Wherein, step S11 further comprises: S111: performing image analysis on the casting thin slice to obtain residual porosity; S112: performing physical property analysis on the casting thin slice at the same depth to obtain gas-measured porosity; S113: establishing a linear regression relationship according to the residual porosity and the gas porosity; S114: Correcting the asphalt content data through the linear regression relationship.

[0027] In step S11, the residual porosity obtained by the image analysis processing method at the same depth point (the residual porosity is equal to the ratio of the extracted residual pore space pixel value to the original image pixel value) and the gas measurement porosity obtained by physical property analysis are first subjected to linear regression analysis to establish a relationship between the two. Subsequently, this correction relationship is applied to the correction of solid asphalt content.

[0028] The expression of the asphalt content data in step S1 is:

[0029] in, To calculate the asphalt content data at the current depth, is the solid asphalt pixel value extracted at the current depth, is the original image pixel value corresponding to the casting slice.

[0030] The above formula, i.e., the solid asphalt content obtained in step S1, is equal to the solid asphalt pixel value extracted by analyzing the cast thin film image divided by the original image pixel value.

[0031] S2: Collect the logging data of a single well, and construct the asphalt content data and the logging data into a depth-matched data set.

[0032] First, conventional logging data of a single well are collected. Then, depth matching is performed on the corrected solid asphalt content data and the processed conventional logging data to construct a data set containing solid asphalt content and multiple logging response characteristic values ​​at the same depth.

[0033] The logging data in step S2 include: gamma logging data, undisturbed formation resistivity logging data, flushing zone resistivity logging data, compensated density logging data, compensated sonic logging data, and compensated neutron logging data.

[0034] Wherein, step S2 also includes: S21: performing preprocessing including denoising, clearing and normalization on the logging data.

[0035] After collecting single-well logging data, it is necessary to perform environmental correction on the logging curve, remove the "noise" of the curve, and clean and normalize the logging data in order to eliminate the impact of the data's own dimension and ensure data quality.

[0036] S3: Train a gradient boosting regression model using the data set to obtain a prediction model.

[0037] Wherein, step S3 further comprises: S31: dividing the data set into a training set and a validation set; Step S31 aims to randomly divide the data set into a training set and a validation set in a certain ratio for subsequent model training.

[0038] S32: training a gradient boosting regression model using the training set to obtain a fitting model; Wherein, step S32 further comprises: S321: Establish a gradient boosting regression model; S322: Setting training parameters of the gradient boosting regression model, wherein the training parameters include at least the number of decision trees and random seeds; S322: Inputting the training set into the gradient boosting regression model for iterative training to reduce the prediction value residual of the gradient boosting regression model and obtain a fitting model.

[0039] In step S32, a gradient boosting regression model is first established, and parameters such as the number of decision trees (n_estimators) and random seeds (random_state) are set. Then, the data of the training set is input into the model to learn the intrinsic relationship between the asphalt content and each conventional logging curve data. Through continuous iterative training, multiple decision tree models are integrated to continuously optimize the prediction performance, gradually reduce the prediction value residual, and improve the model's fitting ability.

[0040] S33: Validate the fitting model through the validation set to obtain a prediction model.

[0041] Wherein, step S33 further comprises: S331: inputting the validation set into the fitting model; S332: Evaluate multiple performance indicators of the fitting model through mean square error and determination coefficient to obtain evaluation results; S333: According to the evaluation results, the number of decision trees and random seeds are selected through cross-validation, and the corresponding fitting model is used as the prediction model.

[0042] In step S33, the validation set data is input into the trained model, and the mean square error (MSE) and determination coefficient (R²) are used to evaluate the performance of the model on the validation set. Cross-validation is used to evaluate the performance of the model under different n_estimators values. By comparing the performance indicators of different models on the validation set (mean square error MSE and root mean square error RMSE), the n_estimators value and random_state corresponding to the model with the best performance are selected, and finally a model that can accurately predict the solid asphalt content is obtained.

[0043] S4: predicting the solid asphalt content of the uncored reservoir section of the single well to be tested by using the prediction model to obtain a prediction result.

[0044] like Figure 2 As shown, the present invention also provides a solid asphalt content prediction system based on the gradient boosting regression method, which is used for a solid asphalt content prediction method based on the gradient boosting regression method as described in any one of the above items, comprising: Analysis module 100: used to collect cast thin sections of target layers in a single well, perform image analysis on the cast thin sections, and obtain asphalt content data at different depths; Acquisition module 200: used to collect well logging data of a single well, and construct the asphalt content data and the well logging data into a depth matching data set; Training module 300: used to train a gradient boosting regression model using the data set to obtain a prediction model; Prediction module 400: used to predict the solid asphalt content of the uncored reservoir section of the single well to be tested by using the prediction model obtained by the training module to obtain a prediction result.

[0045] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0046] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0047] The following is an explanation of a solid asphalt content prediction method and system based on the gradient boosting regression method provided by the present invention through a specific example.

[0048] First, 100 cast thin sections of the core section reservoir of the target layer in Well A were collected or prepared. The solid asphalt content data at 100 depth points of the core section reservoir of the target layer were obtained by cast thin section image analysis and processing method (implemented with MATLAB software) (solid asphalt content = extracted solid asphalt pixel value / original image pixel value), and the solid asphalt content obtained by the image method was corrected. Specifically, the following two steps were included: The specific correction steps are as follows: first, the residual porosity obtained by image analysis and processing method at the same depth point (residual porosity = extracted residual pore space pixel value / original image pixel value) and the gas measurement porosity obtained by physical property analysis are subjected to linear regression analysis to establish the relationship between the two. The established relationship is as follows: Figure 3 As shown; this correction relationship is then applied to the correction of solid asphalt content: Corrected solid asphalt content = 0.8171 × solid asphalt content obtained by image analysis processing method + 0.6054, where the results corresponding to 10 exemplary depth data are shown in Table 1.

[0049] Table 1 Data on solid asphalt content in the core section of the reservoir at the target layer of Well A

[0050] Secondly, conventional logging data of Well A were collected, including gamma logging (GR) data, undisturbed formation resistivity logging (RT) data, flushing zone resistivity logging (RXO) data, compensated density logging (DEN) data, compensated acoustic logging (AC) data, and compensated neutron logging (CNL) data. The results are shown in Table 2.

[0051] Table 2 Conventional logging data of Well A

[0052] Subsequently, the logging curves were environmentally corrected to remove the "noise" of the curves, and the logging data were cleaned and normalized to eliminate the impact of the data dimension and ensure data quality. Then, the corrected solid asphalt content data and the processed conventional logging data were depth matched to construct a data set containing the solid asphalt content of 100 depth points and multiple logging response characteristic values ​​at the same depth. The results are shown in Table 3.

[0053] Table 3 Dataset after depth matching of the corrected solid bitumen content data and the processed conventional logging data

[0054] After obtaining the above dataset, the dataset containing 100 depth points is randomly divided into a training set and a validation set in a ratio of 7:3.

[0055] Then, a gradient boosting regression model was established, and parameters such as the number of decision trees (n_estimators) and random seeds (random_state) were set. Then, the data of the training set containing 70 depth points was input into the model to learn the intrinsic relationship between the asphalt content and the data of each conventional logging curve. Through continuous iterative training, multiple decision tree models were integrated to continuously optimize the prediction performance, gradually reduce the residual of the prediction value, and improve the fitting ability of the model.

[0056] Input the validation set data into the trained model, use mean square error (MSE) and determination coefficient (R²) to evaluate the performance of the model on the validation set, and use cross-validation to evaluate the performance of the model under different n_estimators values. By comparing the performance indicators of different models on the validation set (mean square error MSE and root mean square error RMSE), select the n_estimators value and random_state corresponding to the best performing model. In this example, set n_estimators=1000, random_state=33, MSE=0.063400, R²=0.824631, such as Figure 4 The evaluation results shown are Figure 4 A is a distribution diagram of the true value and the predicted value in the test set. Figure 4 B is a schematic diagram of the distribution relationship between the residual and predicted value of the test set. Figure 4 C is the relationship between the true value and the predicted value in the test set, where the horizontal axis is the sample index, Figure 4 D is the mean square error and determination coefficient of the calculated training set and test set, Figure 4 The results show that the model has good fitting effect and prediction accuracy, proving that the model can better predict the content of solid asphalt, and finally obtain a model that can accurately predict the content of solid asphalt.

[0057] Finally, this model was used to quantitatively predict the solid asphalt content of the uncored reservoir section of Well A. The results are shown in Table 4. Figure 5 The predicted solid asphalt content of the target layer in the vertical direction of well A is shown in a. Then the relevant data of the reservoir section of well B in the work area with known asphalt content is input into the single well solid asphalt content prediction method of the present invention, and the result is as follows: Figure 5 The prediction result of the solid asphalt content of the target layer in the vertical direction of well B shown in b shows that the solid asphalt content prediction result of the method of the present invention is highly consistent with the measured solid asphalt content, which proves the accuracy of the method in predicting the solid asphalt content. Similarly, by applying this method to other wells in the work area, the vertical distribution law of solid asphalt in the reservoir section of all single wells in the study area can be studied.

[0058] Table 4 Quantitative prediction of solid asphalt content in the reservoir of uncored section of Well A

[0059] The present invention provides a method and system for predicting solid asphalt content based on the gradient boosting regression method. The solid asphalt content is predicted by the gradient boosting regression model, and the complex nonlinear relationship in the data can be automatically learned and fitted. Compared with the traditional method, the solid asphalt content can be predicted more accurately. Secondly, the present invention combines the image analysis data and logging data of the cast thin section to construct a depth matching data set, and improves the accuracy and reliability of the data through the correction step, providing a high-quality data foundation for model training. The entire prediction process of the present invention is based on the machine learning method, which realizes automation and intelligence, reduces human intervention, improves work efficiency, and reduces errors caused by human factors. The integrated prediction model is not only applicable to the cored reservoir section, but also can predict the solid asphalt content of the uncored reservoir section, which broadens the scope of application and provides a powerful tool for oil and gas exploration and development.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting solid asphalt content based on gradient boosting regression method, characterized in that: include: S1: Collecting cast thin sections of target layers in a single well, performing image analysis on the cast thin sections, and obtaining asphalt content data at different depths; S2: collecting well logging data of a single well, and constructing the bitumen content data and the well logging data into a depth matching data set; S3: training a gradient boosting regression model using the data set to obtain a prediction model; S4: predicting the solid asphalt content of the uncored reservoir section of the single well to be tested by using the prediction model to obtain a prediction result.

2. The solid asphalt content prediction method based on the gradient boosting regression method according to claim 1 is characterized in that: Step S1 also includes: S11: Correcting the asphalt content data.

3. The solid asphalt content prediction method based on the gradient boosting regression method according to claim 2 is characterized in that: Step S11 further comprises: S111: performing image analysis on the casting thin slice to obtain residual porosity; S112: performing physical property analysis on the casting thin slice at the same depth to obtain gas-measured porosity; S113: establishing a linear regression relationship according to the residual porosity and the gas porosity; S114: Correcting the asphalt content data through the linear regression relationship.

4. The solid asphalt content prediction method based on the gradient boosting regression method according to claim 1 is characterized in that: The expression of the asphalt content data in step S1 is: in, To calculate the asphalt content data at the current depth, is the solid asphalt pixel value extracted at the current depth, is the original image pixel value corresponding to the casting slice.

5. The method for predicting solid asphalt content based on gradient boosting regression method according to claim 1 is characterized in that: The logging data in step S2 include: gamma logging data, undisturbed formation resistivity logging data, flushing zone resistivity logging data, compensated density logging data, compensated sonic logging data, and compensated neutron logging data.

6. The solid asphalt content prediction method based on the gradient boosting regression method according to claim 1 is characterized in that: Step S2 also includes: S21: performing preprocessing including denoising, clearing and normalization on the logging data.

7. The method for predicting solid asphalt content based on gradient boosting regression method according to claim 1, characterized in that: Step S3 further comprises: S31: dividing the data set into a training set and a validation set; S32: training a gradient boosting regression model using the training set to obtain a fitting model; S33: Validate the fitting model through the validation set to obtain a prediction model.

8. The method for predicting solid asphalt content based on gradient boosting regression method according to claim 7, characterized in that: Step S32 further includes: S321: Establish a gradient boosting regression model; S322: Setting training parameters of the gradient boosting regression model, wherein the training parameters include at least the number of decision trees and random seeds; S322: Inputting the training set into the gradient boosting regression model for iterative training to reduce the prediction value residual of the gradient boosting regression model and obtain a fitting model.

9. The method for predicting solid asphalt content based on gradient boosting regression method according to claim 8, characterized in that: Step S33 further comprises: S331: inputting the validation set into the fitting model; S332: Evaluate multiple performance indicators of the fitting model through mean square error and determination coefficient to obtain evaluation results; S333: According to the evaluation results, the number of decision trees and random seeds are selected through cross-validation, and the corresponding fitting model is used as the prediction model.

10. A solid asphalt content prediction system based on gradient boosting regression method, used to execute a solid asphalt content prediction method based on gradient boosting regression method as claimed in any one of claims 1 to 9, characterized in that: include: Analysis module: used to collect cast thin sections of target layers in a single well, perform image analysis on the cast thin sections, and obtain asphalt content data at different depths; Acquisition module: used to collect the logging data of a single well, and construct the asphalt content data and the logging data into a depth-matched data set; Training module: used to train the gradient boosting regression model through the data set to obtain a prediction model; Prediction module: used to predict the solid asphalt content of the uncored reservoir section of the single well to be tested through the prediction model obtained by the training module to obtain the prediction result.

Images