Reservoir sweet spot identification method

Abstract

The application provides a reservoir dessert identification method, comprising the following steps: S1, dividing reservoir dessert types according to fracturing results and production results monitored by an optical fiber outside a pipe; S2, generating a rock mechanics depth data set by using drilling data; S3, calibrating a horizontal well perforation position dessert type, logging characteristics and drilling rock breaking characteristic data set, and establishing a research area oil and gas dessert reservoir identification model through cluster analysis; and S4, inputting the drilling data and the logging data into the research area oil and gas dessert reservoir identification model, and outputting a dessert profile combined with the drilling rock breaking data and the logging data. The application solves the problem of low accuracy of the dessert identification method in the prior art.

Description

Technical Field

[0001] This invention relates to the field of oilfield engineering technology, and more specifically, to a method for identifying reservoir sweet spots. Background Technology

[0002] Fracturing design is the core of hydraulic fracturing. Accurately identifying reservoir stimulation sweet spots and achieving fracture-reservoir matching are crucial for improving fracturing production. One proposed solution uses a self-developed computing engine to generate a weighted OmniLog profile (i.e., sweet spot profile) and mass spectrometry input. A simplified well dataset is established using the sweet spot profile and drilling energy data. Neural networks and optimization algorithms are employed to establish a mapping between production capacity and the dataset. Data-driven optimization of horizontal well perforation cluster locations, production capacity prediction, and well completion parameters are then performed. Field tests were conducted in 30 wells across 6 shale basins, resulting in an average production increase of 27% compared to 78 unoptimized wells.

[0003] However, current methods for identifying "sweet spots" rely solely on well logging data to plot geostress profiles, calculate engineering quality, and generate comprehensive evaluation results. Because the data source depends solely on well logging curves, it inevitably suffers from low vertical resolution (0.125m), errors in rock mechanics parameters (due to dynamic calculations), and an over-reliance on electrical parameters for hydrocarbon-bearing properties, resulting in inaccurate and unreliable final results. Summary of the Invention

[0004] The main objective of this invention is to provide a method for identifying reservoir desserts, in order to solve the problem of low accuracy in existing dessert identification methods.

[0005] To achieve the above objectives, this invention provides a method for identifying reservoir sweet spots, comprising: step S1, classifying reservoir sweet spot types based on fracturing and production results monitored by external optical fiber; step S2, generating a rock mechanics depth dataset using drilling data; step S3, calibrating the sweet spot type, logging characteristics, and drilling rock-breaking characteristics datasets for horizontal well perforation locations, and establishing an oil and gas sweet spot reservoir identification model for the study area through cluster analysis; and step S4, inputting drilling data and logging data into the oil and gas sweet spot reservoir identification model for the study area, and outputting a sweet spot profile combining drilling rock-breaking data and logging data.

[0006] Furthermore, the reservoir sweet spot identification method also includes: drilling data including at least one of drill bit type, drilling torque, drilling fluid density, rotational speed, drilling speed, and mechanical drilling pressure.

[0007] Furthermore, methods for identifying reservoir sweet spots also include: rock mechanics, including rock brittleness and compressive strength.

[0008] Further, step S1 includes: Step S11, deploying external dual-mode optical fiber in the horizontal well; Step S12, classifying the reservoir compressibility type F based on the fracturing monitoring results of the external dual-mode optical fiber. The classification criteria for reservoir compressibility type F are: those not opened during fracturing are incompressible; the compressibility of the fracturing section is divided into levels 1-n from good to poor based on the fluid injection rate, and the numerical value is inversely proportional to the compressibility; Step S13, classifying the reservoir potential type Q based on the production monitoring results of the external dual-mode optical fiber. The classification criteria for reservoir potential type Q are: those that do not produce fluid throughout the production process are non-potential; the production reservoir is divided into levels 1-m from high to low based on the oil or gas production rate, and the numerical value is inversely proportional to the potential; Step S14, classifying the reservoir sweet spot type based on reservoir compressibility type F and reservoir potential type Q.

[0009] Further, in step S14, the calculation formula for classifying reservoir sweet spot types according to reservoir compressibility type F and reservoir potential type Q is as follows: Let k = max(m, n), and establish a sweet spot matrix coupling reservoir compressibility and reservoir potential of order k:

[0010]

[0011] If m≠n, then blank sequences are added to the dessert matrix;

[0012] The dessert combinations on the diagonal of the coupled dessert matrix are classified into one dessert type, T. 11 As the highest quality type of dessert, T 12 and T 21 As the second best dessert type, T 13 T 22 and T 31 As the 3rd best dessert type… and so on, T mn The lowest quality dessert type is categorized into a total of 2k-1 dessert types.

[0013] Further, step S2 includes: step S21, calculating the bottom hole data drilling pressure, torque, rotational speed, and hydraulic energy based on the drilling data, which includes drill bit type, drilling torque, drilling fluid density, rotational speed, drilling speed, and mechanical drilling pressure; step S22, calculating the rock brittleness using the calculated bottom hole data drilling pressure, torque, rotational speed, and hydraulic energy, combined with the first drill bit data; step S23, calculating the rock compressive strength using the calculated bottom hole data drilling pressure, torque, rotational speed, and hydraulic energy, combined with the second drill bit data; step S24, converting the time-domain second-point data into depth data and performing depth correction; step S25, normalizing all the obtained data and interpolating non-drilling process, same well depth, and null data; step S26, establishing a depth dataset of brittleness index and compressive strength that increases with well depth and is spaced at predetermined values.

[0014] Furthermore, the first drill bit data includes the drill bit diameter, and the second drill bit data includes the drill bit nozzle flow rate and nozzle size, with a predetermined value of 0.01-0.05 meters.

[0015] Furthermore, in step S23, the formula for calculating the compressive strength of the rock is as follows:

[0016]

[0017] Where CCS is the rock compressive strength, in MPa; P e Effective drilling pressure, unit: kN; A b This refers to the cross-sectional area of ​​the drill bit, in mm. 2 N represents rotational speed, in RPM; T e Effective torque, unit: kN·m; V e Effective mechanical drilling speed, in m / h.

[0018] Further, step S3 includes: step S31, calibrating the logging feature dataset for the horizontal well perforation location, the logging feature dataset including at least one of thickness, interpreted permeability, sonic transit time, resistivity, natural gamma, total organic carbon, porosity, and oil or gas saturation; step S32, calibrating the drilling sweet spot feature dataset corresponding to the horizontal well perforation location, the drilling sweet spot feature dataset including brittleness index and compressive strength; step S33, establishing a parameter sample library for horizontal well perforation location and reservoir sweet spot type, logging feature dataset, and drilling sweet spot feature dataset, divided into no more than 2K-1 categories. Clustering is performed, where K is the number of clusters. The similarity between sample points is measured based on Euclidean geometric distance, so that the samples in the clusters reach a predetermined similarity. Step S34: The sum of squared errors is used as the criterion function, and the optimal cluster center is selected by traversing all sample points to determine the sweet spot type to which the sample belongs. Step S35: The classified data are assigned to the category of the cluster center corresponding to the least square value, and a neural network model is established. Step S36: Horizontal well data monitored by external fiber optic cables are collected and the neural network model is trained to obtain the oil and gas sweet spot reservoir identification model in the study area.

[0019] Furthermore, in step S34, the formula for calculating the sum of squared errors is:

[0020]

[0021] Where SSE is the sum of squared errors; n j μ is the number of samples in the j-th cluster. j x is the j-th cluster center point; i denoted as sample points; K represents the number of clusters.

[0022] The technical solution of this invention overcomes the shortcomings of conventional sweet spot identification methods, which rely solely on well logging curves, have low vertical resolution, large errors in rock mechanics parameters, and cannot provide quantitative display. It establishes a sweet spot identification method that combines drilling and rock breaking data, is more realistic, has higher resolution, and provides quantitative data. By incorporating drilling and rock breaking data, the data sources for sweet spot identification are enriched, and the vertical resolution reaches the decimeter level. By combining monitoring data and insights from external optical fibers, the accuracy of sweet spot identification is improved by more than 15%, enabling full utilization of reservoir sweet spots, increasing fracture control reserves, and significantly increasing single-well cumulative production by more than 10% while reducing operational costs. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0024] Figure 1 A flowchart of the reservoir sweet spot identification method of the present invention is shown;

[0025] Figure 2 A flowchart of step S1 is shown;

[0026] Figure 3 A flowchart of step S2 is shown;

[0027] Figure 4 A flowchart of step S3 is shown. Detailed Implementation

[0028] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0029] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0030] In this invention, unless otherwise stated, directional terms such as "upper," "lower," "top," and "bottom" are generally used in relation to the direction shown in the accompanying drawings, or in relation to the vertical, perpendicular, or gravitational direction of the component itself; similarly, for ease of understanding and description, "inner" and "outer" refer to the inner and outer contours of each component itself, but the above directional terms are not intended to limit this invention.

[0031] To address the issue of low accuracy in existing dessert identification methods, this invention provides a method for identifying desserts in reservoirs.

[0032] like Figures 1 to 4The method for identifying reservoir sweet spots includes: Step S1, classifying reservoir sweet spot types based on fracturing and production results monitored by external fiber optic cables; Step S2, generating a rock mechanics depth dataset using drilling data; Step S3, calibrating the sweet spot type, logging characteristics, and drilling rock-breaking characteristics datasets for horizontal well perforation locations, and establishing an oil and gas sweet spot reservoir identification model for the study area through cluster analysis; Step S4, inputting the drilling data and logging data into the oil and gas sweet spot reservoir identification model for the study area, and outputting a sweet spot profile combining drilling rock-breaking data and logging data.

[0033] This embodiment addresses the shortcomings of conventional sweet spot identification methods, which rely solely on well logging curves, have low vertical resolution, significant errors in rock mechanics parameters, and cannot provide quantitative data. It establishes a practical, high-resolution, and quantitative sweet spot identification method that incorporates drilling and rock breaking data. By combining drilling and rock breaking data, the data sources for sweet spot identification are enriched, achieving decimeter-level vertical resolution. Furthermore, by integrating monitoring data and insights from external fiber optic cables, the accuracy of sweet spot identification is improved by over 15%, enabling full utilization of reservoir sweet spots, increasing fracture control reserves, and significantly increasing single-well cumulative production by over 10% while reducing operational costs.

[0034] The reservoir sweet spot identification method in this embodiment is mainly suitable for sweet spot identification and perforation location selection in horizontal wells, directional wells and highly deviated wells of oil and gas.

[0035] In this embodiment, the reservoir sweet spot identification method further includes: the drilling data includes at least one of drill bit type, drilling torque, drilling fluid density, rotational speed, drilling rate, and mechanical pressure on the drill bit. Similarly, the reservoir sweet spot identification method further includes: the rock mechanics includes rock brittleness and compressive strength. Of course, other relevant data can be added as needed.

[0036] Based on existing methods for identifying sweet spots, this embodiment expands the data sources for sweet spot identification. Considering that adding drilling data with high vertical resolution and reflecting rock brittleness characteristics, such as drilling specific energy, would significantly improve the accuracy and resolution of sweet spot identification, this embodiment still faces four challenges: First, drilling and fracturing belong to different fields; how to convert drilling data into usable data for fracturing sweet spots? Second, drilling data is discrete; how to convert it into continuous, high-resolution depth profiles requires complex data processing. Third, how to ensure that the reservoir sweet spots are consistent with reality; conventional algorithms, such as selecting reservoirs with high brittleness index, often produce screening results that differ significantly from actual fracturing monitoring and production results. Fourth, how to integrate drilling and logging data to establish an evaluation model. To address these issues, this embodiment specifically designs steps S1 to S3 to better solve these problems.

[0037] like Figure 2 As shown, in this embodiment, step S1, which involves classifying reservoir sweet spots based on the fracturing and production results monitored by the external optical fiber, specifically includes:

[0038] Step S11: Deploy dual-mode optical fiber outside the casing in the horizontal well;

[0039] Step S12: Based on the fracturing monitoring results of the external dual-mode optical fiber, the reservoir compressibility type F is classified. The classification criteria for reservoir compressibility type F are as follows: those that are not opened during the fracturing process are incompressible. The compressibility of the fracturing segment is divided into 1-n levels from good to poor according to the fluid injection volume, and the numerical value is inversely proportional to the compressibility, that is, the lower the value, the better the compressibility.

[0040] Step S13: Based on the production monitoring results of the external dual-mode optical fiber, the reservoir potential type Q is classified. The classification criteria for reservoir potential type Q are as follows: reservoirs that have not produced liquid throughout the entire production process are considered to have no potential. The reservoirs in production are classified into 1-m levels from high to low based on the oil production or gas production, and the numerical value is inversely proportional to the potential, that is, the lower the value, the higher the potential.

[0041] Step S14: Classify the reservoir sweet spot type according to the reservoir compressibility type F and the reservoir potential type Q.

[0042] Furthermore, in step S14, the calculation formula for classifying reservoir sweet spot types based on the reservoir compressibility type F and the reservoir potential type Q is as follows: Let k = max(m, n), and establish a sweet spot matrix coupling k-order reservoir compressibility and reservoir potential:

[0043]

[0044] If m≠n, then blank sequences are added to the dessert matrix. Specifically, blanks can be added using symbols or other means to make the k-order matrix complete.

[0045] The dessert combinations on the diagonal of the coupled dessert matrix are classified into one dessert type, T 11 As the highest quality type of dessert, T 12 and T 21 As the second best dessert type, T 13 T 22 and T 31 As the 3rd best dessert type… and so on, T mn The lowest quality dessert type is categorized into a total of 2k-1 dessert types.

[0046] like Figure 3 As shown, in this embodiment, step S2, namely generating a depth dataset for rock mechanics using drilling data, specifically includes:

[0047] Step S21: Calculate the bottom hole data, drilling pressure, torque, rotational speed, and hydraulic energy based on the drilling data. The drilling data includes the drill bit type, drilling torque, drilling fluid density, rotational speed, drilling speed, and mechanical drilling pressure.

[0048] Step S22: Using the calculated bottom hole data, drilling pressure, torque, rotational speed, and hydraulic energy, combined with the first drill bit data, the rock brittleness is calculated. The first drill bit data includes the drill bit diameter.

[0049] Step S23: Using the calculated bottom hole data (drill pressure, torque, rotational speed, and hydraulic energy), and combining it with the second drill bit data, the rock compressive strength is calculated. The second drill bit data includes the drill bit nozzle flow velocity and nozzle size. The formula for calculating the rock compressive strength is as follows:

[0050]

[0051] In the formula, CCS represents the rock compressive strength, in MPa; P e Effective drilling pressure, unit: kN; A b This refers to the cross-sectional area of ​​the drill bit, in mm. 2 N represents rotational speed, in RPM; T e Effective torque, unit: kN·m; V e Effective mechanical drilling speed, in m / h.

[0052] Step S24: Convert the second data in the time domain into depth data and perform depth correction;

[0053] Step S25: Normalize all the obtained data and interpolate the non-drilling process, the same well depth, and the empty data.

[0054] Step S26: Establish a depth dataset of brittleness index and compressive strength that increases with well depth and is spaced at predetermined values. In this embodiment, the predetermined values ​​are preferably 0.01-0.05 meters.

[0055] like Figure 4 As shown, in this embodiment, step S3, namely, calibrating the dataset of sweet spot type, logging characteristics, and drilling rock breaking characteristics of horizontal well perforation locations, and establishing an oil and gas sweet spot reservoir identification model for the study area through cluster analysis, specifically includes:

[0056] Step S31: Calibrate the logging feature dataset of the horizontal well perforation location. The logging feature dataset includes at least one of the following: thickness, interpreted permeability, sonic transit time, resistivity, natural gamma, total organic carbon, porosity, and oil or gas saturation. This embodiment includes all of the above items.

[0057] Step S32: Calibrate the drilling sweet spot feature dataset corresponding to the perforation location of the horizontal well, wherein the drilling sweet spot feature dataset includes brittleness index and compressive strength;

[0058] Step S33: Establish a parameter sample library for the horizontal well perforation location, the reservoir sweet spot type, the logging feature dataset, and the drilling sweet spot feature dataset, and divide it into no more than 2K-1 clusters, where K is the number of clusters. The similarity between sample points is measured based on Euclidean geometric distance, so that the samples in the clusters reach a predetermined similarity.

[0059] Step S34: Using the sum of squared errors as the criterion function, and selecting the optimal cluster center by traversing all sample points, the dessert type to which the sample belongs is determined; wherein, the formula for calculating the sum of squared errors is:

[0060]

[0061] In the formula, SSE is the sum of squared errors; n j μ is the number of samples in the j-th cluster. j x is the j-th cluster center point; i denoted as sample points; K represents the number of clusters.

[0062] Step S35: Assign the categorized data to the category of the cluster center corresponding to the least square value and establish a neural network model;

[0063] Step S36: Collect horizontal well data for external fiber optic monitoring and train the neural network model to obtain the oil and gas sweet spot reservoir identification model for the study area.

[0064] In practical applications, further methods can be employed, such as coring typical wells in the target block and conducting triaxial mechanical experiments on the cored rock to obtain measured data on the static Young's modulus, Poisson's ratio, compressive strength, fracture toughness, and reservoir stress of the cored rock. A calculation model can then be established to convert the measured mechanical parameters of the target block rock into the calculated results of the drilling data.

[0065] It should be noted that the calculation, fitting and other data processing methods in the above embodiments are conventional data processing methods in the field, and therefore will not be described in detail in this embodiment.

[0066] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects:

[0067] 1. This solves the problem of low accuracy in existing dessert identification methods;

[0068] 2. Based on drilling rock breaking data, it is accurate and has high resolution;

[0069] 3. It has enriched the data sources for dessert identification, and the vertical resolution has reached the decimeter level;

[0070] 4. Improves the accuracy of sweet spot identification by more than 15%, enabling full utilization of sweet spots in reservoirs and increasing fracture control reserves.

[0071] While reducing operating costs, the cumulative production per well has increased significantly by more than 10%.

[0072] Obviously, the embodiments described above are merely some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0073] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0074] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying reservoir sweet spots, characterized in that, include: Step S1: Based on the fracturing results and production results monitored by the external optical fiber, classify the reservoir sweet spot type; Step S2: Generate a depth dataset for rock mechanics using drilling data; Step S3: Calibrate the sweet spot type of horizontal well perforation location, the logging feature dataset obtained from logging data, and the drilling rock breaking feature dataset, and establish an oil and gas sweet spot reservoir identification model in the study area through cluster analysis; Step S4: Input the drilling data and the logging data into the oil and gas sweet spot reservoir identification model of the study area, and output a sweet spot profile combining drilling rock breaking data and logging data.

2. The method for identifying reservoir sweet spots according to claim 1, characterized in that, The reservoir sweet spot identification method also includes: The drilling data includes at least one of the following: drill bit type, drilling torque, drilling fluid density, rotational speed, drilling rate, and mechanical drilling pressure.

3. The method for identifying reservoir sweet spots according to claim 1, characterized in that, The reservoir sweet spot identification method also includes: Rock mechanics includes rock brittleness and compressive strength.

4. The method for identifying reservoir sweet spots according to claim 1, characterized in that, Step S1 includes: Step S11: Deploy dual-mode optical fiber outside the casing in the horizontal well; Step S12: Based on the fracturing monitoring results of the external dual-mode optical fiber, the reservoir compressibility type F is classified. The classification criteria for reservoir compressibility type F are as follows: those that are not opened during the fracturing process are incompressible. The compressibility of the fracturing segment is divided into 1-n levels from good to poor according to the fluid injection volume, and the numerical value is inversely proportional to the compressibility. Step S13: Based on the production monitoring results of the external dual-mode optical fiber, the reservoir potential type Q is classified. The classification criteria for reservoir potential type Q are as follows: reservoirs that have not produced liquid throughout the entire production process are considered to have no potential. The reservoirs in production are classified into 1-m levels from high to low based on oil production or gas production, and the numerical value is inversely proportional to the potential. Step S14: Classify the reservoir sweet spot type according to the reservoir compressibility type F and the reservoir potential type Q.

5. The method for identifying reservoir sweet spots according to claim 4, characterized in that, In step S14, the calculation formula for classifying reservoir sweet spot types based on the reservoir compressibility type F and the reservoir potential type Q is as follows: Let k = max(m, n), and establish a sweet spot matrix for coupling reservoir compressibility and reservoir potential of order k: T ij =(F n )×(Q m )= ; If m≠n, then blank sequences are added to the dessert matrix; The dessert combinations on the diagonal of the coupled dessert matrix are classified into one dessert type, T 11 As the highest quality type of dessert, T 12 and T 21 As the second best dessert type, T 13 T 22 and T 31 As the 3rd best dessert type… and so on, T mn The lowest quality dessert type is categorized into a total of 2k-1 dessert types.

6. The method for identifying reservoir sweet spots according to claim 1, characterized in that, Step S2 includes: Step S21: Calculate the bottom hole data, including drilling pressure, torque, rotational speed, and hydraulic energy, based on the drilling data. The drilling data includes drill bit type, drilling torque, drilling fluid density, rotational speed, drilling speed, and mechanical drilling pressure. Step S22: Using the calculated bottom hole data, drilling pressure, torque, rotational speed, and hydraulic energy, combined with the first drill bit data, the rock brittleness is calculated. Step S23: Using the calculated bottom hole data, drilling pressure, torque, rotational speed, and hydraulic energy, combined with the second drill bit data, the rock compressive strength is calculated. Step S24: Convert the second data in the time domain into depth data and perform depth correction; Step S25: Normalize all the obtained data and interpolate the non-drilling process, the same well depth, and the empty data. Step S26: Establish a depth dataset of brittleness index and compressive strength that increases with well depth and is spaced at predetermined values.

7. The method for identifying reservoir sweet spots according to claim 6, characterized in that, The first drill bit data includes the drill bit diameter, and the second drill bit data includes the drill bit nozzle flow rate and nozzle size, wherein the predetermined value is 0.01-0.05 meters.

8. The method for identifying reservoir sweet spots according to claim 6, characterized in that, In step S23, the formula for calculating the compressive strength of the rock is as follows: ； Where CCS is the rock compressive strength, in MPa; P e Effective drilling pressure, unit: kN; A b This refers to the cross-sectional area of ​​the drill bit, in mm. 2 N represents rotational speed, in RPM; T e Effective torque, unit: kN·m; V e Effective mechanical drilling speed, in m / h.

9. The method for identifying reservoir sweet spots according to claim 1, characterized in that, Step S3 includes: Step S31: Calibrate the logging feature dataset of the horizontal well perforation location. The logging feature dataset includes at least one of the following: thickness, interpreted permeability, sonic transit time, resistivity, natural gamma, total organic carbon, porosity, and oil or gas saturation. Step S32: Calibrate the drilling sweet spot feature dataset corresponding to the perforation location of the horizontal well, which is obtained from the drilling rock breaking feature dataset. The drilling sweet spot feature dataset includes brittleness index and compressive strength. Step S33: Establish a parameter sample library for the horizontal well perforation location, the reservoir sweet spot type, the logging feature dataset, and the drilling sweet spot feature dataset, and divide it into no more than 2K-1 clusters, where K is the number of clusters. The similarity between sample points is measured based on Euclidean geometric distance, so that the samples in the clusters reach a predetermined similarity. Step S34: Use the sum of squared errors as the criterion function, and select the optimal cluster center by traversing all sample points to determine the dessert type to which the sample belongs; Step S35: Assign the categorized data to the category of the cluster center corresponding to the least square value and establish a neural network model; Step S36: Collect horizontal well data for external fiber optic monitoring and train the neural network model to obtain the oil and gas sweet spot reservoir identification model for the study area.

10. The method for identifying reservoir sweet spots according to claim 9, characterized in that, In step S34, the formula for calculating the sum of squared errors is: ； Where SSE is the sum of squared errors; n j μ is the number of samples in the j-th cluster. j x is the j-th cluster center point; i denoted as sample points; K represents the number of clusters.

Images