Tight gas well productivity intelligent prediction method and system, processing equipment and storage medium

Abstract

The invention relates to a tight gas well productivity intelligent prediction method and system, processing equipment and a storage medium, and the method is characterized in that the method comprises the steps: obtaining key feature parameters of logging data of a to-be-detected gas well reservoir section; inputting the obtained key characteristic parameters into a pre-trained tight gas well productivity prediction model to obtain a meter open flow capacity prediction value of the gas well to be detected; the training process of the tight gas well productivity prediction model comprises the steps that key feature parameters of well logging data of each gas well reservoir section and meter-free flow capacity of each gas well are extracted; carrying out dimensionality reduction on the key characteristic parameters by adopting a principal component analysis method; the XGBoost algorithm is adopted, the dense gas well productivity prediction model is constructed and trained based on the key feature parameters after dimension reduction and the corresponding meter open flow capacity, error layer superposition can be avoided, the prediction result has higher precision, and the method can be widely applied to the field of dense gas well productivity prediction.

Description

Technical Field

[0001] The present invention relates to the field of tight gas well productivity prediction, and in particular to a tight gas well productivity intelligent prediction method, system, processing equipment and storage medium. Background Art

[0002] Tight gas well capacity prediction is of great significance to the exploration and development of tight gas reservoirs. It can provide a scientific basis for the design of gas field development plans, help determine the reasonable well network type, mining layer and mining method, thereby optimizing the development strategy and improving economic benefits.

[0003] At present, the methods for predicting the production capacity of tight gas reservoirs mainly include reservoir engineering, numerical simulation and linear statistics. Among them, reservoir engineering and numerical simulation have strict requirements on data, and require accurate physical parameters and strict dynamic data. Due to the complex seepage and strong heterogeneity of tight sandstone reservoirs, there is a complex nonlinear relationship between physical parameters (porosity, permeability, gas saturation) and production capacity, and the linear statistical method has a large error.

[0004] At the same time, the premise of applying these traditional methods is to clarify the reservoir physical property parameters and then make capacity predictions. The prediction of reservoir physical property parameters is mainly based on the well logging curve through regression analysis and empirical formula calculation. In the actual logging process, the reservoir is complex, and the logging data may be affected by noise and outliers, which will affect the prediction accuracy of the model to a certain extent. This method of predicting physical property parameters based on logging data and then making traditional predictions will lead to layers of errors. Directly using the image recognition method to predict capacity through logging curves requires high quality of logging images, and each image needs to be calibrated, which is a large workload, and the image sample volume is also large. Summary of the invention

[0005] In view of the above problems, the object of the present invention is to provide a method, system, processing equipment and storage medium for intelligent prediction of tight gas well productivity that can reduce errors and avoid image calibration.

[0006] To achieve the above-mentioned object, the present invention adopts the following technical solutions: In a first aspect, a method for intelligently predicting the productivity of a tight gas well is provided, comprising: obtaining key characteristic parameters of logging data of a reservoir section of a gas well to be tested; The key characteristic parameters obtained are input into the pre-trained tight gas well productivity prediction model to obtain the predicted value of the meter-free flow rate of the gas well to be tested; The training process of the tight gas well productivity prediction model is as follows: Extract key characteristic parameters of logging data of reservoir sections of each gas well and meter-free flow rate of each gas well; The principal component analysis method is used to reduce the dimension of key characteristic parameters; The XGBoost algorithm is used to build and train a tight gas well productivity prediction model based on the key feature parameters after dimensionality reduction and the corresponding meter-free flow rate.

[0007] Furthermore, the key characteristic parameters of the logging data of each gas well reservoir section include the mean, median, maximum, minimum, mode, upper quartile, lower quartile, range, variance, standard deviation, deviation, coefficient of variation, average amplitude, average peak, average trough and skewness of the neutron, density, acoustic time difference and resistivity curves of each gas well reservoir section.

[0008] Furthermore, before the principal component analysis method is used to reduce the dimension of the key characteristic parameters, the following steps are also included: KMO test and Bartlett test were performed on the extracted key feature parameters.

[0009] Furthermore, the principal component analysis method is used to reduce the dimension of key characteristic parameters, including: Standardize the key characteristic parameters of the logging data of each gas well reservoir section that pass the KMO test and Bartlett test; Calculate the covariance matrix of key feature parameters after standardization; Compute the eigenvalues ​​and eigenvectors of the covariance matrix; Determine an eigenvector matrix based on the calculated eigenvalues ​​and eigenvectors of the covariance matrix; Based on the eigenvector matrix, the number of principal components whose cumulative contribution rate exceeds the preset value is selected as the key characteristic parameter after dimensionality reduction.

[0010] Furthermore, the XGBoost algorithm is used to construct and train a tight gas well productivity prediction model based on the key characteristic parameters after dimensionality reduction and the corresponding meter-free flow rate, including: Randomly select 80% of the data in the key feature parameters after dimension reduction and the corresponding meter-free flow as the training set, and the remaining 20% ​​of the data as the test set; The XGBoost algorithm is used to construct an XGBoost model, and the constructed XGBoost model is trained based on the training set to obtain a trained tight gas well capacity prediction model; The test set is used to evaluate the trained tight gas well productivity prediction model.

[0011] Furthermore, the objective function of the tight gas well productivity prediction model is:

[0012] in, For training The objective function of the wheel; The first samples; is the number of regression trees; For training After round The prediction error of samples is is the actual unimpeded flow rate, For training Model behind the wheel; is the complexity of the regression tree.

[0013] Furthermore, the test set is used to evaluate the trained tight gas well productivity prediction model, including: The test set is input into the trained tight gas well productivity prediction model, the determination coefficient of the test set is calculated, and the model effect is evaluated.

[0014] In a second aspect, a tight gas well productivity intelligent prediction system is provided, comprising: A data acquisition module, used to obtain key characteristic parameters of the well logging data of the reservoir section of the gas well to be tested; Model training module, used to train the tight gas well productivity prediction model; A prediction module is used to input the acquired key characteristic parameters into a pre-trained tight gas well productivity prediction model to obtain a predicted value of the meter-wide unobstructed flow rate of the gas well to be tested; The training process of the model training module is as follows: Extract key characteristic parameters of logging data of reservoir sections of each gas well and meter-free flow rate of each gas well; The principal component analysis method is used to reduce the dimension of key characteristic parameters; The XGBoost algorithm is used to build and train a tight gas well productivity prediction model based on the key feature parameters after dimensionality reduction and the corresponding meter-free flow rate.

[0015] In a third aspect, a processing device is provided, comprising computer program instructions, wherein the computer program instructions, when executed by the processing device, are used to implement the steps corresponding to the above-mentioned method for intelligent prediction of tight gas well productivity.

[0016] In a fourth aspect, a computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions, when executed by a processor, are used to implement the steps corresponding to the above-mentioned method for intelligent prediction of tight gas well productivity.

[0017] The present invention adopts the above technical solution, which has the following advantages: 1. The present invention directly performs an intelligent production capacity prediction method based on well logging data. Compared with the traditional method of interpreting physical parameters from logging curves and then establishing a production capacity prediction model based on the physical parameters, the present invention can directly bypass well logging interpretation and logging images, and use statistical parameters in combination with statistical methods to extract key characteristic parameters of logging data, which can avoid error layer superposition and have higher prediction results.

[0018] 2. Compared with the logging curve image recognition in the prior art, the present invention extracts information through feature numbers. The present invention does not require high-quality logging images, image calibration, and a large number of learning samples. At the same time, the present invention uses XGBoost for prediction, which can improve the prediction accuracy.

[0019] In summary, the present invention can be widely used in the field of tight gas well productivity prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings: Figure 1 It is a schematic diagram of a method flow provided by an embodiment of the present invention; Figure 2 It is a structural diagram of the number of principal components and the cumulative contribution rate provided by an embodiment of the present invention; Figure 3 is a schematic diagram of a test set determination coefficient provided by an embodiment of the present invention; Figure 4 It is a schematic diagram of meter-wide unimpeded flow rate predicted by a conventional logging curve interpretation method provided by an embodiment of the present invention; Figure 5 It is a schematic diagram of the system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0022] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0023] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0024] At present, the methods for predicting the production capacity of tight gas reservoirs mainly include reservoir engineering, numerical simulation and linear statistics. Among them, reservoir engineering and numerical simulation have strict requirements on data, and require accurate physical parameters and strict dynamic data. Due to the complex seepage and strong heterogeneity of tight sandstone reservoirs, there is a complex nonlinear relationship between physical parameters (porosity, permeability, gas saturation) and production capacity, and the linear statistical method has large errors. At the same time, the premise of applying these traditional methods is to clarify the reservoir physical parameters and then predict production capacity. The prediction of reservoir physical parameters is mainly based on well logging curves through regression analysis and empirical formulas. In the actual logging process, the reservoir is complex, and the logging data may be affected by noise and outliers, which affects the prediction accuracy of the model to a certain extent. This method of predicting physical parameters based on logging data and then performing traditional prediction will lead to the superposition of errors. The direct use of image recognition method to predict production capacity through logging curves has high requirements on the quality of logging images, and each image needs to be calibrated, which is a large workload, and the image sample volume is large. The present invention provides a method for intelligent prediction of the production capacity of a tight gas well, comprising: obtaining key characteristic parameters of logging data of a reservoir section of a gas well to be tested; inputting the obtained key characteristic parameters into a pre-trained tight gas well production capacity prediction model to obtain a predicted value of the meter-free flow of the gas well to be tested; the training process of the tight gas well production capacity prediction model is: extracting the key characteristic parameters of the logging data of each gas well reservoir section and the meter-free flow of each gas well; using the principal component analysis method to reduce the dimension of the key characteristic parameters; using the XGBoost algorithm, based on the key characteristic parameters after the dimension reduction and the corresponding meter-free flow, constructing and training the tight gas well production capacity prediction model. The present invention uses statistical indicators to extract the key characteristic parameters of the logging data of the entire reservoir section and then characterize the shape of the logging curve, uses the principal component analysis method to reduce the dimension, and finally uses the XGBoost algorithm to couple the key characteristic parameters after the dimension reduction with the corresponding meter-free flow to perform production capacity prediction. The method directly predicts production capacity based on logging data to avoid the superposition of errors, and at the same time avoids image calibration by extracting information through feature numbers, and uses the XGBoost algorithm for nonlinear fitting to improve the prediction accuracy.

[0025] Example 1 like Figure 1 As shown, this embodiment provides a method for intelligently predicting the productivity of a tight gas well, comprising the following steps: 1) Extract the key characteristic parameters of the logging data of each gas well reservoir section and the meter-free flow rate of each gas well.

[0026] Specifically, the key characteristic parameters of the logging data of each gas well reservoir section include the mean, median, maximum, minimum, mode, upper quartile, lower quartile, range, variance, standard deviation, deviation, coefficient of variation, average amplitude, average peak, average trough and skewness of the neutron, density, acoustic time difference and resistivity curves of each gas well reservoir section. A total of 64 key characteristic parameters are used as independent variables, and the meter open flow rate (open flow rate divided by gas layer thickness) of each gas well is used as the dependent variable.

[0027] 2) Perform KMO test (sampling suitability test) and Bartlett test (Bartlett sphericity test) on the extracted key feature parameters.

[0028] Specifically, when the KMO value of the KMO test is greater than 0.5 and the p value of the Bartlett test is less than 0.05, it indicates that the extracted key characteristic parameters are suitable for principal component analysis, and proceed to step 3); otherwise, increase the number of key characteristic parameters of the logging data of the gas well reservoir section until the requirements are met.

[0029] 3) The principal component analysis (PCA) method was used to reduce the dimension of the key characteristic parameters that passed the KMO test and the Bartlett test. Specifically: 3.1) Standardize the key characteristic parameters of the logging data of each gas well reservoir section that pass the KMO test and Bartlett test to ensure that the mean value of each key characteristic parameter is 0 and the standard deviation is 1.

[0030] 3.2) Calculate the covariance matrix of the key feature parameters after standardization.

[0031] 3.3) Calculate the eigenvalues ​​and eigenvectors of the covariance matrix.

[0032] 3.4) Determine the eigenvector matrix based on the calculated eigenvalues ​​and eigenvectors of the covariance matrix.

[0033] 3.5) Based on the eigenvector matrix, the number of principal components with cumulative contribution rates exceeding 80% is selected as the key feature parameters after dimensionality reduction.

[0034] Specifically, starting from the first principal component in the eigenvector matrix, the contribution rate of each principal component is calculated cumulatively. Specifically, the contribution rate of the first principal component is the ratio of its own variance to the total variance of the original data, and the contribution rate of the second principal component is the ratio of its own variance to the remaining variance after deducting the first principal component, and so on, until the last principal component, the cumulative contribution rate is obtained. The number of principal components with a cumulative contribution rate exceeding 80% is selected as the key feature parameter after dimensionality reduction.

[0035] 4) Using the XGBoost (distributed gradient boosting library) algorithm, based on the key feature parameters after dimensionality reduction and the corresponding meter-free flow rate, a tight gas well productivity prediction model is constructed and trained. Specifically: 4.1) Randomly select 80% of the key feature parameters after dimension reduction and the corresponding meter-free flow as the training set, and the remaining 20% ​​of the data as the test set.

[0036] 4.2) Using the XGBoost algorithm, an XGBoost model is constructed, and the constructed XGBoost model is trained based on the training set to obtain a trained tight gas well capacity prediction model, namely the PCA-XGBoost model.

[0037] Specifically, the XGBoost model is a powerful prediction model composed of a series of regression trees. The core idea is to continuously add regression trees and fit the residual of the previous tree by generating new trees. When the training is completed and N regression trees are obtained, the scores corresponding to each tree are added up to get the predicted value of the sample. The model is as follows: (1) in, The first Key characteristic parameters of samples (the well logging data of the gas well reservoir section consists of several learning samples, each of which has corresponding key characteristic parameters); is the predicted unobstructed flow value in meters; is the number of regression trees; For the Weight value; is the key feature parameter after dimensionality reduction. In each iteration, a regression tree is added to fit the residual between the prediction result of the previous regression tree and the true value. The iterative process is:

[0038]

[0039]

[0040]

[0041] (2) in, For training Model behind the wheel; Functions added in the previous round for retention; For the newly added function; For the The objective function of the tight gas well productivity prediction model is: (3) in, For training The objective function of the wheel; For training After round The prediction error of samples is is the actual unobstructed flow rate in meters; is the complexity of the regression tree. When the objective function no longer decreases with the number of iterations, the model is trained.

[0042] 4.3) The trained tight gas well productivity prediction model is evaluated using the test set, and the error index between the predicted result and the actual value is calculated to evaluate the performance and accuracy of the model.

[0043] Specifically, the present invention uses the coefficient of determination as an error indicator for evaluation, inputs the test set into the trained tight gas well productivity prediction model, calculates the coefficient of determination of the test set, and evaluates the model effect. The coefficient of determination is defined as: (4) in, is the coefficient of determination; is the predicted unimpeded flow in meters; is the average unimpeded flow rate in meters; To extract the actual meter unobstructed flow.

[0044] 5) Obtain the key characteristic parameters of the reservoir logging data of the gas well to be tested, and input them into the trained tight gas well productivity prediction model to obtain the predicted value of the meter-free flow rate of the gas well to be tested.

[0045] The following describes in detail the intelligent prediction method for the production capacity of a tight gas well according to the present invention through specific embodiments: Neutron, density, acoustic time difference, resistivity curves and corresponding meter-free flow rate of 200 gas wells in a tight sandstone gas reservoir were selected. The mean, median, maximum, minimum, mode, upper quartile, lower quartile, range, variance, standard deviation, deviation, coefficient of variation, average amplitude, average peak, average trough and skewness of the four characteristic curves of each well were counted, and a total of 64 key characteristic parameters were used as independent variables, and the meter-free flow rate of each well was used as the dependent variable.

[0046] The results of KMO and Bartlett tests are shown in Table 1. The KMO test value of this example is 0.817>0.5. At the same time, the results of the Bartlett test show that the chi-square test value is 3228.361, and the P value is 0.000<0.01. The level is significant, and the original hypothesis is rejected. There is a correlation between the variables, so the data is suitable for principal component analysis: Table 1: KMO and Bartlett test results

[0047] The number of principal components with cumulative contribution rates exceeding 80% is selected. Figure 2 It can be seen that the number of principal components is 6. 80% of the key characteristic parameters after dimensionality reduction and the corresponding meter-free flow are randomly selected as the training set for algorithm training. When the objective function no longer decreases with the number of iterations, the model is trained. The determination coefficient is the evaluation criterion, such as Figure 3 As shown, the coefficient of determination of the model test set is 0.98.

[0048] like Figure 4 The figure shows the schematic diagram of the meter-free flow rate predicted by the traditional logging curve interpretation method. The physical parameters interpreted by the logging curve interpretation method are: permeability (K), gas layer thickness (h), porosity (Φ), gas saturation (s g ), the accuracy determination coefficient of fitting the characteristic parameters with the meter open flow rate is 0.29. It can be seen that the accuracy of the meter open flow rate predicted by the traditional method is much lower than that of the meter open flow rate predicted by the method of the present invention.

[0049] Example 2 like Figure 5 As shown, this embodiment provides a tight gas well productivity intelligent prediction system, including: A data acquisition module, used to obtain key characteristic parameters of the well logging data of the reservoir section of the gas well to be tested; Model training module, used to train the tight gas well productivity prediction model; A prediction module is used to input the acquired key characteristic parameters into a pre-trained tight gas well productivity prediction model to obtain a predicted value of the unobstructed flow rate per meter to be tested; The training process of the model training module is: Extract key characteristic parameters of logging data of each gas well reservoir section; KMO test and Bartlett test were performed on the extracted key feature parameters; The principal component analysis method was used to reduce the dimension of the key characteristic parameters that passed the KMO test and Bartlett test; The XGBoost algorithm is used to build and train a tight gas well productivity prediction model based on the key feature parameters after dimensionality reduction.

[0050] The system provided in this embodiment is used to execute the above-mentioned method embodiments. Please refer to the above-mentioned embodiments for specific processes and detailed contents, which will not be repeated here.

[0051] Example 3 This embodiment provides a processing device corresponding to the method for intelligent prediction of tight gas well productivity provided in Example 1. The processing device can be suitable for client processing devices, such as mobile phones, laptops, tablet computers, desktop computers, etc., to execute the method of Example 1.

[0052] The processing device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. The memory stores a computer program that can be run on the processing device, and the processing device executes the intelligent prediction method for the production capacity of tight gas wells provided in this embodiment 1 when running the computer program.

[0053] In some implementations, the memory may be a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage.

[0054] In some other implementations, the processor may be a central processing unit (CPU), a digital signal processor (DSP), or other general-purpose processors of various types, which are not limited herein.

[0055] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0056] Those skilled in the art will understand that the structure of the above-mentioned computing device is only a partial structure related to the solution of the present invention, and does not constitute a limitation on the computing device to which the solution of the present invention is applied. The specific computing device may include more or fewer components, or combine certain components, or have a different arrangement of components.

[0057] Example 4 This embodiment provides a computer program product corresponding to the method for intelligent prediction of tight gas well productivity provided in Example 1. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the method for intelligent prediction of tight gas well productivity described in Example 1 are loaded.

[0058] Computer readable storage media can be tangible devices that hold and store instructions used by instruction execution devices. Computer readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any combination thereof.

[0059] The above embodiment provides a computer-readable storage medium, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0060] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks 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 produce 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.

[0061] 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.

[0062] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0063] The above embodiments are only used to illustrate the present invention, wherein the structure, connection mode and manufacturing process of each component may be changed. Any equivalent transformations and improvements based on the technical solution of the present invention should not be excluded from the protection scope of the present invention.

Claims

1. A method for intelligent prediction of tight gas well productivity, characterized in that: include: Obtain key characteristic parameters of the logging data of the reservoir section of the gas well to be tested; The key characteristic parameters obtained are input into the pre-trained tight gas well productivity prediction model to obtain the predicted value of the meter-free flow rate of the gas well to be tested; The training process of the tight gas well productivity prediction model is as follows: Extract key characteristic parameters of logging data of reservoir sections of each gas well and meter-free flow rate of each gas well; The principal component analysis method is used to reduce the dimension of key characteristic parameters; The XGBoost algorithm is used to build and train a tight gas well productivity prediction model based on the key feature parameters after dimensionality reduction and the corresponding meter-free flow rate.

2. The method for intelligent prediction of tight gas well productivity according to claim 1, characterized in that: The key characteristic parameters of the logging data of each gas well reservoir section include the mean, median, maximum, minimum, mode, upper quartile, lower quartile, range, variance, standard deviation, deviation, coefficient of variation, average amplitude, average peak, average trough and skewness of the neutron, density, acoustic time difference and resistivity curves of each gas well reservoir section.

3. The method for intelligent prediction of tight gas well productivity according to claim 1, characterized in that: Before using principal component analysis to reduce the dimension of key feature parameters, it also includes: KMO test and Bartlett test were performed on the extracted key feature parameters.

4. A method for intelligent prediction of tight gas well productivity according to claim 3, characterized in that: The principal component analysis method is used to reduce the dimension of key characteristic parameters, including: Standardize the key characteristic parameters of the logging data of each gas well reservoir section that pass the KMO test and Bartlett test; Calculate the covariance matrix of key feature parameters after standardization; Compute the eigenvalues ​​and eigenvectors of the covariance matrix; Determine an eigenvector matrix based on the calculated eigenvalues ​​and eigenvectors of the covariance matrix; Based on the eigenvector matrix, the number of principal components whose cumulative contribution rate exceeds the preset value is selected as the key characteristic parameter after dimensionality reduction.

5. The method for intelligent prediction of tight gas well productivity according to claim 1, characterized in that: The XGBoost algorithm is used to construct and train a tight gas well productivity prediction model based on the key characteristic parameters after dimensionality reduction and the corresponding meter-free flow rate, including: Randomly select 80% of the data in the key feature parameters after dimension reduction and the corresponding meter-free flow as the training set, and the remaining 20% ​​of the data as the test set; The XGBoost algorithm is used to construct an XGBoost model, and the constructed XGBoost model is trained based on the training set to obtain a trained tight gas well capacity prediction model; The test set is used to evaluate the trained tight gas well productivity prediction model.

6. A method for intelligent prediction of tight gas well productivity according to claim 5, characterized in that: The objective function of the tight gas well productivity prediction model is: in, For training The objective function of the wheel; The first samples; is the number of regression trees; For training After round The prediction error of samples is is the actual unimpeded flow rate, For training Model behind the wheel; is the complexity of the regression tree.

7. The method for intelligent prediction of tight gas well productivity according to claim 5, characterized in that: The test set is used to evaluate the trained tight gas well productivity prediction model, including: The test set is input into the trained tight gas well productivity prediction model, the determination coefficient of the test set is calculated, and the model effect is evaluated.

8. An intelligent prediction system for tight gas well productivity, characterized in that: include: A data acquisition module, used to obtain key characteristic parameters of the well logging data of the reservoir section of the gas well to be tested; Model training module, used to train the tight gas well productivity prediction model; A prediction module is used to input the acquired key characteristic parameters into a pre-trained tight gas well productivity prediction model to obtain a predicted value of the meter-free flow rate of the gas well to be tested; The training process of the model training module is as follows: Extract key characteristic parameters of logging data of reservoir sections of each gas well and meter-free flow rate of each gas well; The principal component analysis method is used to reduce the dimension of key feature parameters; The XGBoost algorithm is used to build and train a tight gas well productivity prediction model based on the key feature parameters after dimensionality reduction and the corresponding meter-free flow rate.

9. A processing device, characterized in that: It includes computer program instructions, wherein the computer program instructions, when executed by a processing device, are used to implement the steps corresponding to the method for intelligent prediction of tight gas well productivity according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, wherein the computer program instructions, when executed by a processor, are used to implement the steps corresponding to the method for intelligent prediction of tight gas well productivity according to any one of claims 1 to 7.

Images