A method, system, device and storage medium for analyzing main controlling factors of oil well productivity

Abstract

The present invention relates to an analysis method, system, equipment and storage medium for the main controlling factors of oil well productivity. The method comprises the following steps: respectively cleaning the production dynamic factors and static geological factors in the acquired oil well data; using the control variable method to independently analyze the cleaned production dynamic and static geological data, respectively designing their application methods, and using a machine learning algorithm to sort the main controlling factors of oil well productivity, and obtain the importance ranking of each production dynamic factor and static geological factor. The present invention can improve the applicability and accuracy of the machine learning algorithm in the field of research on the main controlling factors of oil well productivity by efficiently cleaning the production dynamic and static geological data and using a data usage method suitable for the characteristics of the production dynamic and static geological data. Therefore, the present invention can be widely used in the field of oil and gas field development.

Description

Technical Field

[0001] The present invention relates to the field of oil and gas field development, and in particular to a data-driven oil well productivity main control factor analysis method, system, equipment and storage medium. Background Art

[0002] The main controlling factors of oil well productivity are the basic research content for oil well productivity evaluation. The existing analysis methods of the main controlling factors of oil well productivity include: theoretical formula, physical experiment, numerical simulation and data analysis method. However, due to the many factors affecting oil well productivity and the strong nonlinear relationship between oil well productivity and various influencing factors, the applicability of the above methods in complex real-world situations is limited.

[0003] In order to more accurately analyze the nonlinear relationship between oil well productivity and various influencing factors, a machine learning method is used to analyze the main controlling factors of oil well productivity using actual data. At present, in this type of work, machine learning algorithms are often used directly to analyze the collected data, and less attention is paid to efficient data cleaning methods for actual data, reasonable data usage methods, and the applicability of machine learning algorithms in the field of oil well productivity research. As a result, a large amount of inaccurate data is used, and the analysis results of the machine learning algorithm have a large deviation from the existing oil field development theory. Summary of the invention

[0004] In view of the above problems, the purpose of the present invention is to provide a data-driven oil well productivity main controlling factor analysis method, system, equipment and storage medium, which can perform efficient data cleaning on actual data and improve the accuracy of oil well productivity main controlling factor analysis.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, a method for analyzing the main controlling factors of oil well productivity is provided, which comprises the following steps:

[0007] Perform data cleaning on the production dynamic factors and static geological factors in the acquired oil well data respectively;

[0008] The dynamic production factors and static geological factors that have undergone data cleaning are analyzed independently, and the machine learning algorithm is used to rank the main controlling factors of oil well capacity, and the importance ranking of each dynamic production factor and static geological factor is obtained.

[0009] Preferably, for production dynamic factors, the steps of data cleaning are:

[0010] Screening the oil wells that have been put into production according to the first preset discrimination criterion;

[0011] Missing data handling;

[0012] Determining the official production date of the oil well according to the second preset judgment standard;

[0013] The data affected by the operation is removed according to the third preset judgment standard.

[0014] Preferably, for static geological factors, the data cleaning steps are:

[0015] Screening geological factor data of perforation layers;

[0016] Calculation of overall geological data for a single well;

[0017] Missing data handling.

[0018] Preferably, the calculation method of the overall geological data of a single well is: for the reservoir thickness and perforation thickness of a single well, the data of each small layer are accumulated, and for the other static geological factors, they are weighted averaged according to the perforation thickness of each layer to calculate the overall geological data of the single well.

[0019] Preferably, the calculation formula for the overall geological data of a single well is:

[0020]

[0021] Where X is the geological data of the whole well; i is the geological data of the i-th layer; H i is the perforation thickness of the i-th layer; i is the small layer number of a single well.

[0022] Preferably, the method for analyzing production dynamic factors comprises:

[0023] ① Select a well from the production dynamics data set, use the daily oil production as the label of the machine learning algorithm, and use the other production dynamic factors as input features. Use the machine learning algorithm as the base model, train the base model with the well data and obtain the importance weight of each feature;

[0024] ② Calculate the importance weight of the features for the production dynamic data of each well in turn according to step ①, and take the average value of the corresponding feature weights of all wells, and determine the importance ranking of each feature according to the average weight;

[0025] ③ Sort by importance, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation;

[0026] ④ Repeat steps ②-③ until the least important features are screened out, and rank the order in which the features are eliminated as the importance of the dynamic factors of oil well productivity.

[0027] Preferably, the method for analyzing static geological factors comprises:

[0028] ① Using the static geological factor data of all wells as the training set, daily oil production as the label of the machine learning algorithm, and other static geological factors as input features, the machine learning algorithm is used as the base model to train the model and obtain the importance weight of each feature;

[0029] ② Determine the importance of each feature according to the weight, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation;

[0030] ③ Repeat step ② until the least important feature is screened out, and the order in which the features are eliminated is used as the importance ranking of the static geological factors of oil well productivity.

[0031] The second aspect is to provide an oil well productivity main control factor analysis system, which includes:

[0032] A data cleaning module is used to clean the production dynamic factors and static geological factors in the acquired oil well data respectively;

[0033] The main controlling factor analysis module is used to independently analyze the production dynamic factors and static geological factors after data cleaning, and use machine learning algorithms to rank the main controlling factors of oil well capacity to obtain the importance ranking of each production dynamic factor and static geological factor.

[0034] The third aspect is to provide a processing device, which includes at least a processor and a memory, wherein the memory stores a computer program, and when the processor runs the computer program, it executes the steps of the method for analyzing the main controlling factors of oil well productivity.

[0035] A fourth aspect is to provide a computer storage medium having computer-readable instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the steps of the method for analyzing the main controlling factors of oil well productivity.

[0036] The present invention adopts the above technical solution, which has the following advantages:

[0037] (1) The present invention combines the prior knowledge of oil field development theory to efficiently clean production dynamic and static geological data;

[0038] (2) The present invention adopts a data usage method suitable for the characteristics of dynamic and static geological data of production, and cooperates with a main control factor analysis method designed according to the characteristics of factors affecting oil well productivity, so as to eliminate inaccurate data in the factors affecting oil well productivity, and at the same time effectively improve the accuracy of the analysis of the main control factors of oil well productivity.

[0039] Therefore, the present invention can be widely applied to the field of oil and gas field development. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flow chart of a method for analyzing main controlling factors of oil well productivity provided by an embodiment of the present invention;

[0041] Figure 2(a) to Figure 2(h) 2(a) is a comparison diagram of production dynamic data before and after data cleaning in an embodiment of the present invention, wherein FIG2(b) is a comparison diagram of daily oil production; FIG2(c) is a comparison diagram of production pressure difference; FIG2(d) is a comparison diagram of gasoline ratio; FIG2(e) is a comparison diagram of water content; FIG2(f) is a comparison diagram of casing pressure; FIG2(g) is a comparison diagram of nozzle size; and FIG2(h) is a comparison diagram of pump frequency. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0043] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0044] The present invention provides a method for analyzing the main controlling factors of oil well productivity, which cleans the production dynamic factors and static geological factors in the acquired oil well data respectively; adopts the control variable method to independently analyze the cleaned production dynamic and static geological data, respectively designs their application methods, and uses a machine learning algorithm to rank the main controlling factors of oil well productivity, and obtains the importance ranking of each production dynamic factor and static geological factor. The present invention can improve the applicability and accuracy of the machine learning algorithm in the field of research on the main controlling factors of oil well productivity by efficiently cleaning the production dynamic and static geological data and using a data usage method suitable for the characteristics of the production dynamic and static geological data.

[0045] Correspondingly, some embodiments of the present invention also provide an oil well productivity main control factor analysis system, equipment and storage medium.

[0046] Example 1

[0047] like Figure 1 As shown, the present embodiment provides a method for analyzing the main controlling factors of oil well productivity, comprising the following steps:

[0048] 1) Perform data cleaning on the production dynamic factors and static geological factors in the acquired oil well data respectively;

[0049] 2) The dynamic production factors and static geological factors that have been cleaned are analyzed independently, and the machine learning algorithm is used to rank the main controlling factors of oil well capacity, and the importance ranking of each dynamic production factor and static geological factor is obtained.

[0050] Preferably, in the above step 1), the oil well data obtained can be divided into two categories: production dynamic factors and static geological factors.

[0051] Specifically, production dynamic factors include: date, daily production time, daily oil production, water cut, gas-oil ratio, production pressure difference, wellhead pressure, casing pressure, pump frequency, nozzle size, etc.;

[0052] Static geological factors include: permeability, porosity, oil saturation, mud content, reservoir thickness, perforation thickness, crude oil density, crude oil viscosity, etc.

[0053] Preferably, in the above step 1), when the acquired oil well data is cleaned, different data cleaning methods are used according to the data type.

[0054] Specifically, for production dynamic factors, the steps of data cleaning are as follows:

[0055] ① Screening and obtaining the production data of the oil wells that have been put into production according to the first preset discrimination standard;

[0056] The method is: reading the daily production time data of the oil well in date order, and when the daily production time data meets the first preset judgment standard, entering the next step, otherwise the production data of the oil well is discarded.

[0057] Optionally, the first preset criterion is: the number of days with daily production time greater than 0 hours is greater than 0 days.

[0058] ② Process missing data of the screened production wells;

[0059] Read the daily production time, nozzle size and production pressure difference data of the oil well in date order. If any of the data is missing, remove the production data of that date.

[0060] ③ Determine the official production date of the oil well based on the second preset judgment standard;

[0061] Read the daily production time and water content data of the oil well in date order. If the daily production time and water content data of the day meet the second preset judgment standard, then the day is used as the official production date of the oil well, and the production data before the day is removed;

[0062] Optionally, the second preset criterion is: the daily production time is greater than 0 hours and the moisture content is less than 5% for 7 consecutive days;

[0063] ④ Remove the data affected by the operation according to the third preset judgment standard;

[0064] Read the daily production time, nozzle size and production pressure difference data of the oil well in date order, and remove the production data affected by the operation according to the third preset judgment standard;

[0065] Optionally, the third preset judgment criterion is: the production dynamic data corresponding to the date when the daily production time is less than 24 hours, excluding the production data corresponding to the date when the nozzle size changes by more than 3 mm and the production pressure difference changes by more than 0.2 MPa.

[0066] ⑤The oil wells and their corresponding production data retained after processing in the above steps can be used as data sets for training machine learning algorithms.

[0067] For static geological factors, the steps of data cleaning are as follows:

[0068] ① Screening geological factor data of perforation layers;

[0069] Read the perforated layers of the oil well in order of reservoir depth, and select the corresponding geological factor data;

[0070] ②Calculation of overall geological data of a single well;

[0071] For the reservoir thickness and perforation thickness of a single well, the accumulation of data of each small layer is adopted, and the weighted average of the other geological factor data is performed according to the perforation thickness of each layer according to formula (1) to calculate the overall geological data of the single well;

[0072]

[0073] In the formula, X is the geological data of the whole well; x i is the geological data of the i-th layer; H i is the perforation thickness of the i-th layer; i is the small layer number of a single well.

[0074] ③ Missing data processing;

[0075] If the fluid property data of a certain oil well is missing, the corresponding data of the adjacent wells will be used to supplement the well. If the reservoir property data of a certain oil well is missing, the data of this layer will not be used in the weighted average of step ②.

[0076] ④ The static geological factor data of the oil well calculated after processing in the above steps can be used as a data set for training the machine learning algorithm.

[0077] Preferably, in the above step 2), specifically, for the production dynamic factors, the importance ranking of each production dynamic factor is determined by first analyzing a single well and then taking an average of each well, including the following steps:

[0078] ① Select a well from the data set, use the daily oil production as the label of the machine learning algorithm, use the other production dynamic factors as input features, use machine learning algorithms with importance evaluation capabilities such as Random Forest or XGBoost as the base model, train the base model with the well data and obtain the importance weights of each input feature;

[0079] ② Calculate the importance weight of each feature according to step ① for each well data, and take the average of the corresponding feature weights of all wells, and determine the importance ranking of each feature based on the average weight;

[0080] ③ Sort by importance, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation;

[0081] ④ Repeat steps ②-③ until the least important features are screened out. The order in which the features are eliminated is the order of their importance from strong to weak;

[0082] ⑤The analysis of the main dynamic control factors of oil well production capacity can be completed based on the importance ranking.

[0083] Preferably, in the above step 2), the static geological factors are analyzed by multi-well data joint analysis to determine the importance ranking of each static geological factor, and the main steps are:

[0084] ①Use the static geological factor data of all wells in the data set as the training set, the daily oil production as the label of the machine learning algorithm, various static geological factors as input features, and machine learning algorithms with importance evaluation capabilities such as Random Forest or XGBoost as the base model to train the model and obtain the importance weight of each feature;

[0085] ② Determine the importance of each feature according to the weight, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation;

[0086] ③ Repeat step ② until the least important feature is selected. The order in which the features are eliminated is the order of their importance from strong to weak;

[0087] ④ The analysis of static geological main controlling factors of oil well productivity can be completed based on the importance ranking.

[0088] This embodiment uses a data-driven approach to study the main controlling factors of oil well productivity using machine learning algorithms, taking into account the prior knowledge of oil field development theory, and provides data cleaning methods and main controlling factor analysis methods for production dynamic factors and static geological factors, respectively, to form a set of data-driven oil well productivity main controlling factor analysis methods and systems. The present invention can achieve efficient data cleaning, provide a reasonable data usage method for the application of machine learning algorithms in the field of oil well productivity main controlling factor research, and improve its accuracy.

[0089] Example 2

[0090] This example takes 87 oil wells in the P oil field as an example, and applies the method of Example 1 to analyze the main controlling factors of the oil well productivity. The details are as follows:

[0091] (1) The dynamic and static geological data of oil well production were processed by using a data cleaning method combined with prior knowledge of oil field development theory, and Well W1 was selected as a demonstration example.

[0092] According to the data cleaning method described in Example 1, the production dynamic data such as daily oil production, water content, gas-oil ratio, production pressure difference, wellhead pressure, casing pressure, pump frequency, and nozzle size are processed, and the processing results are as follows: Figure 2(a) to Figure 2(h) shown.

[0093] The processing results of static geological data such as permeability, porosity, oil saturation, mud content, reservoir thickness, perforation thickness, crude oil density, and crude oil viscosity according to the data cleaning method are shown in Table 1.

[0094] Table 1 Comparison of static geological data before and after data cleaning

[0095]

[0096] Using the Random Forest algorithm as the base model, multiple rounds of training were performed based on a data set of 87 wells. After each round of training, the factors with the highest weight were removed until the least important factors were screened out. The results of each training round are shown in Table 2. The order of elimination of each factor can be obtained from the table, that is, the influence of each factor on the oil well productivity is: production pressure difference> gas-oil ratio> water content> wellhead pressure> pump frequency> nozzle size> casing pressure. According to this importance ranking, the top 1 / 2 factors can be selected as the main controlling factors of oil well productivity, namely production pressure difference, gas-oil ratio, and water content.

[0097] Table 2 Weights of production dynamic factors in each training round

[0098]

[0099]

[0100] Using the Random Forest algorithm as the base model, multiple rounds of training were performed based on a data set of 87 wells. After each round of training, the factors with the highest weight were removed until the least important factors were screened out. The results of each training round are shown in Table 3. The order of elimination of each factor can be obtained from the table, that is, the influence of each factor on oil well productivity is: permeability> perforation thickness> reservoir thickness> oil saturation> porosity> shale content> crude oil density> crude oil viscosity. According to this importance ranking, combined with actual application needs, the top 1 / 2 factors can be selected as the main controlling factors of oil well productivity, namely permeability, perforation thickness, reservoir thickness, and oil saturation.

[0101] Table 3 Weights of static geological factors in each training round

[0102]

[0103] Example 3

[0104] The above-mentioned embodiment 1 provides a method for analyzing the main controlling factors of oil well productivity. Correspondingly, this embodiment provides a system for analyzing the main controlling factors of oil well productivity. The identification system provided in this embodiment can implement a method for analyzing the main controlling factors of oil well productivity in embodiment 1. The identification system can be implemented by software, hardware, or a combination of software and hardware. For example, the system may include integrated or separate functional modules or functional units to execute the corresponding steps in each method of embodiment 1. Since the identification system of this embodiment is basically similar to the method embodiment, the description process of this embodiment is relatively simple. For relevant parts, please refer to the partial description of embodiment 1. The embodiment of the system of this embodiment is only illustrative.

[0105] This embodiment provides an oil well productivity main control factor analysis system, including:

[0106] A data cleaning module is used to clean the production dynamic factors and static geological factors in the acquired oil well data respectively;

[0107] The main controlling factor analysis module is used to independently analyze the production dynamic factors and static geological factors after data cleaning, and use machine learning algorithms to rank the main controlling factors of oil well capacity to obtain the importance ranking of each production dynamic factor and static geological factor.

[0108] Example 4

[0109] This embodiment provides a processing device corresponding to the oil well productivity main control factor analysis method provided in this embodiment 1. The processing device can be a processing device for a client, such as a mobile phone, a laptop computer, a tablet computer, a desktop computer, etc., to execute the method of embodiment 1.

[0110] 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 processor, and the processor executes a method for analyzing the main control factors of oil well productivity provided in Example 1 when running the computer program.

[0111] 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 memory.

[0112] 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.

[0113] Example 5

[0114] The method for analyzing the main controlling factors of oil well productivity of this embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the method for analyzing the main controlling factors of oil well productivity described in this embodiment 1 are loaded.

[0115] 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.

[0116] It should be noted that the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present application. Each box in the flowchart or block diagram may represent a module, a program segment or a part of the code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions.

[0117] 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

[0118] 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 analyzing the main controlling factors of oil well productivity, characterized in that: The following steps are involved: Perform data cleaning on the production dynamic factors and static geological factors in the acquired oil well data respectively; Among them, for production dynamic factors, the steps of data cleaning are: Screening the oil wells that have been put into production according to the first preset discrimination criterion; Missing data handling; Determining the official production date of the oil well according to the second preset judgment standard; removing data affected by the operation according to a third preset judgment standard; For static geological factors, the steps of data cleaning are as follows: Screening geological factor data of perforation layers; Calculation of overall geological data for a single well; Missing data handling; We analyze the production dynamic factors and static geological factors after data cleaning, and use machine learning algorithms to rank the main control factors of oil well productivity, and get the importance ranking of each production dynamic factor and static geological factor; Methods for analyzing production dynamics factors include: ① Select a well from the production dynamics data set, use the daily oil production as the label of the machine learning algorithm, and use the other production dynamic factors as input features. Use the machine learning algorithm as the base model, train the base model with the well data and obtain the importance weight of each feature; ② Calculate the importance weight of the features for the production dynamic data of each well in turn according to step ①, and take the average value of the corresponding feature weights of all wells, and determine the importance ranking of each feature according to the average weight; ③ Sort by importance, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation; ④ Repeat steps ②-③ until the least important features are screened out, and the order in which the features are eliminated is used as the importance ranking of the dynamic factors of oil well productivity; Methods for analyzing static geological factors include: ① Using the static geological factor data of all wells as the training set, daily oil production as the label of the machine learning algorithm, and other static geological factors as input features, the machine learning algorithm is used as the base model to train the model and obtain the importance weight of each feature; ② Determine the importance of each feature according to the weight, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation; ③ Repeat step ② until the least important feature is screened out, and the order in which the features are eliminated is used as the importance ranking of the static geological factors of oil well productivity.

2. The method for analyzing the main controlling factors of oil well productivity according to claim 1, characterized in that: The calculation method of the overall geological data of a single well is as follows: for the reservoir thickness and perforation thickness of a single well, the data of each small layer are accumulated, and for the other static geological factors, the weighted average is performed according to the perforation thickness of each layer to calculate the overall geological data of the single well.

3. The method for analyzing the main controlling factors of oil well productivity according to claim 2, characterized in that: The calculation formula for the overall geological data of a single well is: In the formula, It is the geological data of a single well as a whole; For the geological data of the layers; For the The perforation thickness of the layer; Number the sublayers of a single well.

4. An oil well productivity main control factor analysis system, characterized in that: include: A data cleaning module is used to clean the production dynamic factors and static geological factors in the acquired oil well data respectively; Among them, for production dynamic factors, the steps of data cleaning are: Screening the oil wells that have been put into production according to the first preset discrimination criterion; Missing data handling; Determining the official production date of the oil well according to the second preset judgment standard; removing data affected by the operation according to a third preset judgment standard; For static geological factors, the steps of data cleaning are as follows: Screening geological factor data of perforation layers; Calculation of overall geological data for a single well; Missing data handling; The main control factor analysis module is used to independently analyze the production dynamic factors and static geological factors after data cleaning, and use machine learning algorithms to rank the main control factors of oil well capacity to obtain the importance ranking of each production dynamic factor and static geological factor; The analysis of production dynamic factors includes: ① Select a well from the production dynamics data set, use the daily oil production as the label of the machine learning algorithm, and use the other production dynamic factors as input features. Use the machine learning algorithm as the base model, train the base model with the well data and obtain the importance weight of each feature; ② Calculate the importance weight of the features for the production dynamic data of each well in turn according to step ①, and take the average value of the corresponding feature weights of all wells, and determine the importance ranking of each feature according to the average weight; ③ Sort by importance, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation; ④ Repeat steps ②-③ until the least important features are screened out, and the order in which the features are eliminated is used as the importance ranking of the dynamic factors of oil well productivity; The analysis of static geological factors includes: ① Using the static geological factor data of all wells as the training set, daily oil production as the label of the machine learning algorithm, and other static geological factors as input features, the machine learning algorithm is used as the base model to train the model and obtain the importance weight of each feature; ② Determine the importance of each feature according to the weight, remove the most important one or several features, use the remaining features to form a new data set, and use the base model for the next round of training and feature importance evaluation; ③ Repeat step ② until the least important feature is screened out, and the order in which the features are eliminated is used as the importance ranking of the static geological factors of oil well productivity.

5. A processing device, comprising at least a processor and a memory, wherein a computer program is stored in the memory, wherein: When the processor runs the computer program, the processor executes the steps of the method for analyzing the main controlling factors of oil well productivity according to any one of claims 1 to 3.

6. A computer storage medium, characterized in that: Computer-readable instructions are stored thereon, and the computer-readable instructions can be executed by a processor to implement the steps of the method for analyzing the main controlling factors of oil well productivity according to any one of claims 1 to 3.

Images