Method for constructing mature exploration area short-term well site deployment planning model

Abstract

The application provides a mature exploration area medium and short term well site deployment planning model construction method, comprising the following steps: step 1, according to the reservoir oil and gas condition information in the drilling process of the exploration well and the constraint of the adjacent exploration well, the exploration well effect and influence are analyzed; step 2, according to the exploration well effect and influence distribution, the oil and gas reservoir area is analyzed; step 3, according to the oil and gas reservoir area analysis result, the single well reserve growth contribution and investment situation of the region are analyzed; step 4, taking the reserve growth expectation and the planned investment as the constraint condition, the target planning method is used to analyze the exploration well site deployment planning, and the exploration well site deployment planning scheme is formulated. The mature exploration area medium and short term well site deployment planning model construction method carries out the regional reserve growth prediction, establishes an intelligent, simple and practical exploration well site deployment planning model construction method, and provides the exploration well site deployment planning reference scheme.

Description

Technical Field

[0001] This invention relates to the field of oilfield information technology, and in particular to a method for constructing a short-term well location deployment planning model in a mature exploration area. Background Technology

[0002] Predicting reserve growth in mature exploration areas primarily relies on exploratory wells. Exploratory well placement planning is a crucial aspect of oil and gas exploration, directly impacting the achievement of reserve growth targets and the return on exploration investment. Exploratory well placement generally employs target planning, using reserves as the decision objective and economic benefits and workload as constraints to propose different well placement plans. However, current methods often fail to comprehensively consider influencing factors, leading to a decrease in the scientific rigor of the decision-making process.

[0003] Well placement planning relies on the understanding and prediction of geological patterns, depends on the enrichment and occurrence state of underground oil and gas resources, and follows the laws of exploration geology and exploration development. Currently, numerous scholars both domestically and internationally have conducted research on exploration patterns and achieved fruitful results, proposing a series of reserve growth models to guide well placement planning. Typical models include: the Ong's life cycle model, the Logistic production change model, grey system theory, resource-reserve-production prediction models, reservoir distribution probability models, fractal mathematical methods, the Gompattz model, the reservoir scale sequence method, the "broom" model, the reserve growth coefficient method, and geological analysis methods. However, most of these models and methods predict reserve growth expectations from a macroscopic perspective, have complex formulas, and are difficult to use. They are poorly adaptable to exploration areas with complex geological conditions and are not conducive to guiding well placement planning in mature exploration areas under complex geological conditions.

[0004] Therefore, the most important aspect of exploration well location planning research is to organically combine geological laws, geological understanding, exploration well distribution, reservoir status, and reserve growth to establish a simple and easy-to-use exploration well location deployment planning model, and then use this model to find the optimal exploration well location deployment planning scheme.

[0005] Chinese patent application CN201911323760.5 discloses a method for optimizing the deployment of injection-production well networks. This method includes a virtual oil well network optimization deployment method based on production well spacing constraints and maximum controllable movable reserves; a well network optimization deployment method for the comprehensive utilization of old oil wells with maximum controllable movable reserves based on production well spacing constraints and seepage resistance constraints; a virtual water injection well location optimization method based on injection-production well spacing range and minimum seepage resistance level difference; and a water injection well location optimization method for the comprehensive utilization of old wells based on injection-production well spacing range and minimum seepage resistance level difference. This invention solves the problem of vectorized design of injection-production well spacing in the design of well networks for new reservoir construction and the comprehensive adjustment of well networks in old reservoirs, enabling quantitative and scientific decision-making for well network and spacing design and adjustment.

[0006] Chinese patent application CN201911367075.2 discloses a method and apparatus for determining the location of infill wells in complex fault-block oil reservoir development areas. The method includes: acquiring a three-dimensional geological model of the complex fault-block oil reservoir development area; for each well-free grid in the three-dimensional geological model, determining the number of oil layers encountered by a single well and the actual controlled area of ​​a single well in each oil layer based on geological conditions; determining the remaining movable oil reserves of a single well in each oil layer based on the number of grids contained within the actual controlled area of ​​a single well in each oil layer, the remaining oil saturation, residual oil saturation, crude oil density, and pore volume of each grid; superimposing the remaining movable oil reserves of all oil layers to obtain the total remaining movable oil reserves of a single well; and determining the well-free grid corresponding to the maximum total remaining movable oil reserves as the deployment location of the infill well. This application can determine the location of infill wells through a reasonable method, improving the utilization rate of reserves in complex fault-block oil reservoir development areas.

[0007] Chinese patent application CN202111051454.8 discloses an integrated geological and engineering development method for ultra-shallow shale gas. Based on historical geological and survey data, the method involves: deploying drilling well locations; conducting full-hole coring to establish a lithological columnar section; testing gas content to identify high-quality gas-bearing zones; using stress relief methods for downhole in-situ stress measurement, geophysical logging, and borehole stereo imaging to establish a geological and mechanical model of the gas-bearing zone; conducting mineral composition and rock mechanical parameter tests to formulate a fracturing plan; performing fracturing operations; monitoring the hydraulic fracturing network information; testing gas production to determine daily gas production; and assessing whether the economic benefits are met. If not, drilling well locations are redeployed; if so, a rolling deployment of development well networks is implemented. This invention features a simple process, low cost, and effective development of ultra-shallow shale gas resources.

[0008] The existing technologies described above are significantly different from the present invention and have failed to solve the technical problem we want to address. Therefore, we have invented a new method for constructing a short-term well location deployment planning model in mature exploration areas. Summary of the Invention

[0009] The purpose of this invention is to provide a method for constructing a short-term well location planning model in a mature exploration area that can accurately and quickly plan various well location deployment schemes for decision-making reference.

[0010] The objective of this invention can be achieved through the following technical measures: a method for constructing a short-term well location deployment planning model in mature exploration areas, comprising:

[0011] Step 1: Analyze the effectiveness and impact of exploration wells based on reservoir oil and gas conditions and constraints from adjacent exploration wells during the drilling process.

[0012] Step 2: Analyze the oil and gas reservoir areas based on the results and impact distribution of exploration wells;

[0013] Step 3: Based on the results of the oil and gas reservoir area analysis, analyze the contribution of single wells to the growth of reserves and the investment situation in the region;

[0014] Step 4: Using the expected growth of reserves and planned investment as constraints, conduct well site deployment planning analysis using the target planning method, and formulate a well site deployment planning scheme.

[0015] The objective of this invention can also be achieved through the following technical measures:

[0016] Step 1 includes:

[0017] Step 1a: Classify the effectiveness of the exploration well based on the reservoir oil and gas condition information, including whether the reservoir has oil and gas indications or electrical logging indications during the drilling process and the results and conclusions after the casing test.

[0018] Step 1b: Conduct an impact analysis on the exploration well based on its geological conditions and the constraints of adjacent exploration wells;

[0019] Step 1c: Divide the study area into several grids, and radiate the well effectiveness with the grid where the well is located as the center to obtain the distribution of the well effectiveness influence.

[0020] In step 1a, the exploration wells are classified into effective categories based on the oil and gas conditions of the reservoir they pass through. Exploration wells with oil layers, heavy oil layers, or water-bearing oil layers are classified into the first category; exploration wells with oil-water co-layers or oil-water layers are classified into the second category; exploration wells with oil and gas show reservoirs or electrical logging showing that there may be oil and gas reservoirs are classified into the third category; and the others are classified into the fourth category.

[0021] In step 1b, for areas where geological conditions change very rapidly, the reservoir can still be considered homogeneous within a limited area around the exploration well, so the effectiveness of a single exploration well can cover a certain range; in mature exploration areas, due to the large number of exploration wells, the coverage range of a single well is constrained by adjacent exploration wells, and the size of the coverage radius has little impact on most areas.

[0022] Step 1c includes:

[0023] Step 1c1: Connect to the database, obtain initial data, including the x and y coordinates and classification of the exploration wells, and normalize the x and y coordinates of the data;

[0024] Step 1c2: Multiply the coordinate-normalized data by the grid size factor and round it down. Merge the data according to the coordinates, retaining the optimal value for each category.

[0025] Step 1c3: Create a grid image. Based on the merged data, fill the image with the classification results according to the coordinate index. At this point, the image contains the classification information of all exploration wells.

[0026] Step 1c4: Assign validity values ​​to the grid image. Based on the different classifications of the exploration wells, assign different validity values ​​A1, A2, A3, and A4 to the corresponding locations, with the validity decreasing in that order.

[0027] Step 1c5: Perform radial iteration based on the total number of iterations S1. Each iteration traverses the entire image: If the validity value a of a point (x, y) on the image is in {A1, A2, A3} and the iteration number s is less than the next iteration number S2, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value ai to the point that does not contain a validity value; If the validity value a of a point (x, y) on the image is A4, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value a to the point that does not contain a validity value.

[0028] Step 1c6: After the iteration is complete, output the influence distribution map.

[0029] Step 2 includes:

[0030] Step 2a: Perform effectiveness clustering on the exploration wells based on the effectiveness classification results;

[0031] Step 2b: Based on the clustering results of valid exploration wells, classify invalid exploration wells.

[0032] Step 2a includes:

[0033] Step 2a1: Connect to the database, obtain the initial data, and normalize the original data;

[0034] Step 2a2: Assign different weight coefficients to each feature;

[0035] Step 2a3: Randomly select k samples from the processed data as the initial k centroid vectors: {μ1,μ2,...,μk};

[0036] Step 2a4: Cluster the data around the k centroids to divide it into different clusters;

[0037] Step 2a5, output the cluster partition C = {C1, C2, ..., Ck}.

[0038] Step 2a4 includes:

[0039] Step 2a4a, for n = 1, 2, ..., N: Initialize the cluster partition C as follows (t = 1, 2...k);

[0040] Steps 2a4b: For i = 1, 2, ..., m, use the vq algorithm to calculate sample x. i and each centroid vector μ j The distance (j=1,2,...k) will be x i The smallest one is d. ij The corresponding category λ i At this point, update Cλ i =Cλ i ∪{x i};

[0041] Step 2a4c: For j = 1, 2, ..., k, recalculate the new centroids for all sample points in Cj; if the centroid vectors change, repeat step 2a4a; if none of the k centroid vectors change, proceed to step 2a5.

[0042] Step 2b includes:

[0043] Step 2b1: Based on the effective well clustering results, take the centroid U = {μ1,μ2,...μj} (j = 1,2,...k) of each class;

[0044] Step 2b2: For invalid wells s = 1, 2, ..., S, use Euclidean distance to calculate the distance between each invalid well si and each centroid vector μj (j = 1, 2, ... k), and label the si with the smallest value as the category λi corresponding to dij. At this time, update Cλi = Cλi∪{xi}.

[0045] Step 2b3, output the final cluster partition Cs = {Cs1, Cs2, ..., Csk};

[0046] Step 2b4: Obtain each type of data and use the Graham scan algorithm to obtain the set of points that constitute each convex hull;

[0047] Step 2b5: Using the set of convex hull points, draw the convex hull of each cluster.

[0048] Step 2b4 includes:

[0049] Step 2b4a: Start from the lowest point on the y-axis as the starting point p0;

[0050] Step 2b4b: Start with p0 and perform a polar coordinate scan, traversing all points in the graph in a counter-clockwise direction according to the polar coordinate angle.

[0051] In step 2b4c, if the newly traversed point can produce a left rotation, add the point to the convex hull; otherwise, discard it.

[0052] Step 3 includes:

[0053] Step 3a: Based on the effective well clustering results, take the effective well cluster division C = {C1, C2, ... Ck};

[0054] Step 3b: Based on the clustering results of all exploration wells, divide all exploration well clusters into Cs = {Cs1, Cs2, ..., Csk};

[0055] Step 3c: Traverse the effective well cluster division C = {C1, C2, ... Ck} and all exploration well cluster divisions Cs = {Cs1, Cs2, ... Csk}, calculate the number of effective exploration wells and the total number of exploration wells, and obtain their percentages;

[0056] Step 3d: Traverse the effective well cluster division Co = {Co1, Co2, ..., Cok}, store the number of each data in the original data list, find the corresponding reported block reserves according to the number, calculate the number of reported blocks and the cumulative reserves of the region, and obtain the average reserves per well. Traverse all exploration well cluster divisions Cs = {Cs1, Cs2, ..., Csk}, calculate the number of all exploration wells, and obtain the percentage of the number of reported blocks and the number of all exploration wells.

[0057] Step 3e: Traverse all well clusters and divide them into Cos = {Cos1, Cos2, ... Cosk}. Store the number of each data in the original data list. Find the drilling investment of the corresponding well name based on the number, calculate the number of wells and the cumulative investment in the region, and obtain the average investment per well.

[0058] Step 3f: Iterate through the centroids U = {μ1, μ2, ... μj} (j = 1, 2, ... k) of each type, and display the data calculated in step 3e above the centroids of each region.

[0059] Step 4 includes:

[0060] Step 4a: Obtain block q and the number of clusters k from the model. First, calculate the average reserves Ra = {Ra1, Ra2, Ra3, Ra4….Rak} and the average investment Ia = {Ia1, Ia2, Ia3, Ia4,…..Iak} for all sample point partitions C = {C1, C2,...Ck} corresponding to each class; at the same time, remove the data with an average reserve of 0 in the class.

[0061] Step 4b: Let the investment be I and the expected reserve be R. By traversing the average reserve set Ra and the average investment set Ia of the clusters, obtain the most greedy investment-reserve expectation RImax, that is, through RImax, the minimum investment Imin can be calculated to obtain the maximum reserve Rmax.

[0062] Step 4c: Judge whether RImax satisfies the investment I and the expected reserve R at this time. If it does not satisfy, loosen the investment condition and calculate the minimum investment Imin’ that can meet the expected reserve R. At this time, Imin’ is greater than I.

[0063] Step 4d: If RImax satisfies the investment I and the expected reserve, that is, there exists a certain set of solutions E = {e1, e2, e3….} such that the investment Ie = {Ie1, Ie2, Ie3…..} corresponding to each solution satisfies Ie <= -I, and the reserves Re = {Re1, Re2, Re3….} that each solution can obtain satisfy Re >= R; by traversing, obtain a partial set of solutions Eg = {Eg1, Eg2…..} such that these solutions meet the following: the most greedy solution, drill wells in the area with the largest I / R ratio; average well placement, when the investment I’ < I, drill wells evenly in the area with reserves.

[0064] This invention presents a method for constructing a short-term well placement planning model in mature exploration areas. Addressing the challenges of representing drilling effectiveness in mature exploration areas with complex geological structures and the poor scientific rigor of well placement planning, this invention proposes a drilling effectiveness characterization method based on the distribution of exploration wells and their contribution, constrained by oil and gas reservoir patterns. This method then constructs a well placement planning model to solve the problem of exploration well placement planning. This invention approximates the geological structure distribution using the distribution of exploration wells, and primarily relies on the effectiveness of exploration wells and their contribution to reserve growth. It employs artificial intelligence methods to delineate oil and gas reservoir distribution areas and construct the exploration well placement planning model. Using this model, multiple exploration well placement schemes can be accurately and quickly planned for decision-making reference, without requiring specific geological structural factors for each region. This method for constructing a short-term well location deployment planning model in mature exploration areas is designed for the reserve growth prediction approach based primarily on exploration wells. It organically combines geological laws, geological understanding, well distribution, reservoir status, and reserve growth to predict regional reserve growth. It establishes an intelligent, simple, and practical method for constructing an exploration well location deployment planning model, and provides a reference scheme for exploration well location deployment planning based on this model. Attached Figure Description

[0065] Figure 1 This is a flowchart of a specific embodiment of the method for constructing a short-term well location deployment planning model in mature exploration areas according to the present invention;

[0066] Figure 2 This is a flowchart of a specific embodiment of the well impact analysis of the present invention;

[0067] Figure 3 This is a flowchart of a specific embodiment of the well influence distribution map construction method of the present invention;

[0068] Figure 4 A flowchart of a specific embodiment of the effective well clustering construction method of the present invention;

[0069] Figure 5 This is a flowchart of a specific embodiment of the invalid well classification construction method of the present invention;

[0070] Figure 6 A flowchart illustrating a specific embodiment of the analysis of regional single-well reserve growth contribution and investment in this invention;

[0071] Figure 7 This is a flowchart illustrating a specific embodiment of the well location deployment planning scheme analysis of the present invention. Detailed Implementation

[0072] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[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 of the present invention. 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, and / or combinations thereof.

[0074] like Figure 1 As shown, Figure 1 This is a flowchart illustrating the method for constructing a short-term well location deployment planning model in mature exploration areas according to the present invention. The method for constructing a short-term well location deployment planning model in mature exploration areas includes:

[0075] Step 1: Analyze the effectiveness and impact of exploration wells based on reservoir oil and gas conditions and constraints from adjacent exploration wells during the drilling process.

[0076] Step 2: Analyze the oil and gas reservoir areas based on the results and impact distribution of exploration wells;

[0077] Step 3: Based on the results of the oil and gas reservoir area analysis, analyze the contribution of single wells to the growth of reserves and the investment situation in the region;

[0078] Step 4: Using the expected growth of reserves and planned investment as constraints, conduct well site deployment planning analysis using the target planning method, and formulate a well site deployment planning scheme.

[0079] The following are several specific embodiments of the application of the present invention.

[0080] Example 1

[0081] In a specific embodiment 1 of the present invention, the method for constructing a short-term well location deployment planning model in mature exploration areas includes the following steps:

[0082] Step 1: Analyze the effectiveness and impact of exploration wells based on reservoir oil and gas conditions and constraints from adjacent exploration wells during the drilling process.

[0083] Step 1a: Classify the effectiveness of the exploration well based on the reservoir oil and gas condition information, including whether the reservoir has oil and gas indications or electrical logging indications during the drilling process and the results and conclusions after the casing test.

[0084] Based on the oil and gas conditions of the reservoir through which the exploration well passes, exploration wells are classified into four categories based on their effectiveness: the first category includes exploration wells with oil layers (including reservoirs that have yielded industrial oil and gas flow), heavy oil layers, and water-bearing oil layers; the second category includes exploration wells with oil-water co-layers and oil-water layers; the third category includes exploration wells with oil and gas show reservoirs or electrical logging indicating that there may be oil and gas reservoirs; and the fourth category includes all others.

[0085] Step 1b: Conduct an impact analysis on the exploration well based on its geological conditions and the constraints of adjacent exploration wells;

[0086] In the Bohai Bay Basin, where geological conditions change very rapidly, the reservoir can still be considered homogeneous within a limited area around the exploration well. Therefore, the effectiveness of a single exploration well can cover a certain range. In mature exploration areas, due to the large number of exploration wells, the coverage range of a single well is constrained by adjacent exploration wells, and the size of the coverage radius has little impact on most areas.

[0087] Step 1c: Divide the study area into several grids, and radiate the well effectiveness with the grid where the well is located as the center to obtain the distribution of the well effectiveness influence.

[0088] Step 1c1: Connect to the database, obtain initial data, including the x and y coordinates and classification of the exploration wells, and normalize the x and y coordinates of the data;

[0089] Step 1c2: Multiply the coordinate-normalized data by the grid size factor and round it down. Merge the data according to the coordinates, retaining the optimal value for each category.

[0090] Step 1c3: Create a grid image. Based on the merged data, fill the image with the classification results according to the coordinate index. At this point, the image contains the classification information of all exploration wells.

[0091] Step 1c4: Assign validity values ​​to the grid image. Based on the different classifications of the exploration wells, assign different validity values ​​A1, A2, A3, and A4 to the corresponding locations, with the validity decreasing in that order.

[0092] Step 1c5: Perform radial iteration based on the total number of iterations S1. Each iteration traverses the entire image: If the validity value a of a point (x, y) on the image is in {A1, A2, A3} and the iteration number s is less than the next iteration number S2, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value ai to the point that does not contain a validity value; If the validity value a of a point (x, y) on the image is A4, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value a to the point that does not contain a validity value.

[0093] Step 1c6: After the iteration is complete, output the influence distribution map.

[0094] Step 2: Analyze the oil and gas reservoir area based on the exploration well results and impact distribution; including:

[0095] Step 2a involves clustering exploration wells based on their effectiveness classification results; this includes the following steps:

[0096] Step 2a1: Connect to the database, obtain the initial data, and normalize the original data;

[0097] Step 2a2: Assign different weight coefficients to each feature;

[0098] Step 2a3: Randomly select k samples from the processed data as the initial k centroid vectors: {μ1,μ2,...,μk};

[0099] Step 2a4a, for n = 1, 2, ..., N: Initialize the cluster partition C as follows t=1,2...k, t = 1, 2...k;

[0100] Steps 2a4b: For i = 1, 2, ..., m, use the vq algorithm to calculate sample x. i and each centroid vector μ j The distance (j=1,2,...k) will be x i The smallest one is d. ij The corresponding category λ i At this point, update Cλ i =Cλ i ∪{x i};

[0101] Step 2a4c: For j = 1, 2, ..., k, recalculate the new centroids for all sample points in Cj. If the centroid vectors change, repeat step 2a4a; if all k centroid vectors remain unchanged, proceed to step 2a5.

[0102] Step 2a5, output the cluster partition C = {C1, C2, ..., Ck}.

[0103] Step 2b: Based on the clustering results of valid exploration wells, classify invalid exploration wells.

[0104] Includes the following steps:

[0105] Step 2b1: Based on the effective well clustering results, take the centroid U = {μ1,μ2,...μj} (j = 1,2,...k) of each class;

[0106] Step 2b2, for invalid wells s = 1, 2, ..., S. Use Euclidean distance to calculate the distance between each invalid well si and each centroid vector μj (j = 1, 2, ..., k), and label the si with the smallest distance as the category λi corresponding to dij. At this time, update Cλi = Cλi∪{xi}.

[0107] Step 2b3, output the final cluster partition Cs = {Cs1, Cs2, ..., Csk};

[0108] Step 2b4: Obtain each type of data and use the Graham scan algorithm to obtain the set of points that constitute each convex hull;

[0109] Step 2b4a: Start from the lowest point on the y-axis as the starting point p0;

[0110] Step 2b4b: Start with p0 and perform a polar coordinate scan, traversing all points in the graph in a counter-clockwise direction according to the polar coordinate angle.

[0111] Step 2b4c: If the newly traversed point can produce a left rotation, add the point to the convex hull; otherwise, discard it.

[0112] Step 2b5: Using the set of convex hull points, draw the convex hull of each cluster.

[0113] Step 3, based on the analysis results of the oil and gas reservoir area, analyze the contribution of single wells to the growth of reserves and the investment situation in the region; including the following steps:

[0114] Step 3a: Based on the effective well clustering results, take the effective well cluster division C = {C1, C2, ... Ck};

[0115] Step 3b: Based on the clustering results of all exploration wells, divide all exploration well clusters into Cs = {Cs1, Cs2, ..., Csk};

[0116] Step 3c: Traverse the effective well cluster division C = {C1, C2, ... Ck} and all exploration well cluster divisions Cs = {Cs1, Cs2, ... Csk}, calculate the number of effective exploration wells and the total number of exploration wells, and obtain their percentages;

[0117] Step 3d: Traverse the effective well cluster division Co = {Co1, Co2, ..., Cok} (store the number of each data in the original data list), find the reserves of the corresponding reporting block according to the number, calculate the number of reporting blocks and the cumulative reserves of the region, and obtain the average reserves of a single well. Traverse all exploration well cluster divisions Cs = {Cs1, Cs2, ..., Csk}, calculate the number of all exploration wells, and obtain the percentage of the number of reporting blocks and the number of all exploration wells.

[0118] Step 3e: Traverse all exploration well clusters Cos = {Cos1, Cos2,... Cosk} (the numbers storing each data in the original data list), find the drilling investment corresponding to the exploration well name according to the numbers, calculate the number of exploration wells and the cumulative investment in the region, and obtain the average investment per well.

[0119] Step 3f: Traverse the centroids of each type U = {μ1, μ2,... μj} (j = 1, 2,... k), and display the calculated data above the centroid of each region.

[0120] Step 4: Taking the reserve growth expectation and planned investment as constraint conditions, conduct exploration well location deployment planning analysis by the goal programming method, and formulate an exploration well location deployment planning scheme. It includes the following steps:

[0121] Step 4a: Obtain the block q and the number of clusters k from the model. First, calculate the average reserves Ra = {Ra1, Ra2, Ra3, Ra4…. Rak} and the average investment Ia = {Ia1, Ia2, Ia3, Ia4,….. Iak} of all sample point partitions C = {C1, C2,... Ck} corresponding to each class. At the same time, remove the data with an average reserve of 0 in the class.

[0122] Step 4b: Let the investment be I and the expected reserve be R. By traversing the average reserve set Ra and the average investment set Ia of the clusters, obtain the most greedy investment-reserve expectation RImax, that is, through RImax, the minimum investment Imin can be calculated to obtain the maximum reserve Rmax.

[0123] Step 4c: Judge whether RImax meets the investment I and the expected reserve R at this time. If it does not meet, we relax the investment condition and calculate the minimum investment Imin’ that can meet the expected reserve R. At this time, Imin’ is greater than I.

[0124] Step 4d: If RImax meets the investment I and the expected reserve, that is, there exists a certain plan E = {e1, e2, e3….}, such that the investment Ie = {Ie1, Ie2, Ie3…..} corresponding to each plan satisfies Ie <= -I, and the reserves Re = {Re1, Re2, Re3….} that each plan can obtain satisfy Re >= R. We obtain some plans Eg = {Eg1, Eg2…..} by traversing, such that these plans meet: the most greedy plan, drilling wells in the region with the largest I / R ratio; average well drilling, when the investment I’ < I, drilling wells evenly in the regions with reserves.

[0125] Embodiment 2

[0126] In a specific Embodiment 2 of applying the present invention, the method for constructing the medium- and short-term well location deployment planning model in the mature exploration area includes the following steps:

[0127] In step S1, the effectiveness and impact of the exploration well are analyzed.

[0128] In step S1a, exploration wells are classified according to their effectiveness based on the oil and gas conditions of the reservoirs they traverse. Wells with oil layers (including reservoirs yielding industrial oil and gas flow), heavy oil layers, or water-bearing oil layers are classified as Category I; wells with oil-water co-layers or oil-water-bearing layers are classified as Category II; wells with oil and gas show reservoirs or those showing potential oil and gas reservoirs through logging are classified as Category III; and all others are classified as Category IV. This classification is relatively rough and the number and criteria for classification can be adjusted based on the actual conditions of the study area.

[0129] In step S1b, the well influence refers to the effective coverage area of ​​a single well. In the Bohai Bay Basin, where geological conditions change rapidly, the reservoir can still be considered homogeneous within a limited area around the well; therefore, the effectiveness of a single well can cover a certain range. Similarly, due to the rapid changes in geological conditions, the coverage area of ​​a single well is limited. The specific coverage radius needs to be given based on the actual geological conditions within the study area. In mature exploration areas, due to the large number of wells, the coverage area of ​​a single well is constrained by adjacent wells, and the size of the coverage radius has little impact in most areas.

[0130] In step S1c, the study area is divided into several grids, and each well's grid is assigned a well validity. Radiation is then performed outwards from the well, meaning the validity of adjacent grids is assigned to the well's grid validity. Radiation continues outwards from the newly assigned grid until a coverage area is reached or a collision with another grid occurs. Since the attributes do not decay during radiation, and the expansion distance is the same (one grid) each time, the coverage radius can be equivalently converted into the number of radiation iterations.

[0131] The specific process is as follows: Figure 3As shown, step S1c1 involves connecting to the database to obtain initial data, including the x and y coordinates and classification of the wells, and normalizing the x and y coordinates. Step S1c2 involves multiplying the normalized data by a grid size coefficient and rounding it down, then merging the data based on the coordinates, retaining the optimal value for each classification. Step S1c3 involves creating a grid image, filling the image with the classification results according to the coordinate indices based on the merged data; at this point, the image contains the classification information for all wells. Step S1c4 involves assigning validity values ​​to the grid image, assigning different validity values ​​A1, A2, A3, and A4 to the corresponding locations based on the well classification, with the validity decreasing sequentially. Step S1c5 involves performing radial iteration based on the total number of iterations S1, traversing the entire image in each iteration. If the validity value 'a' of a point (x, y) on the image lies within {A1, A2, A3} and the iteration count 's' is less than the iteration count 'S2', then step S1c5a is executed. Radial judgment is performed on the four points above, below, left, and right of this point, and the points that do not contain a validity value are assigned the validity value 'a'. If the validity value 'a' of a point (x, y) on the image is A4, then step S1c5b is executed. Radial judgment is performed on the four points above, below, left, and right of this point, and the points that do not contain a validity value are assigned the validity value 'a'. After the iteration in step S1c5 is completed, step S1c6 is executed to output the influence distribution map.

[0132] like Figure 1 , 2 As shown in Figure 4, in step S2, an oil and gas reservoir analysis is performed.

[0133] In step S2a, effective exploration wells are clustered. First, cluster analysis is performed on the effective exploration wells; wells in categories I, II, and III are all considered effective. Clustering factors include: well location (i.e., x, y coordinates), well effectiveness category, and well contribution to reserve growth. All factors are normalized, and then a clustering algorithm is designed for clustering.

[0134] The specific process is as follows: Figure 4 As shown. In step S2a1, the database is connected to obtain initial data, and the original data is normalized. In step S2a2, different weight coefficients are assigned to each feature. In step S2a3, k samples are randomly selected from the processed data as the initial k centroid vectors: {μ1,μ2,...,μk}. In step S2a4, for n=1,2,...,N: In step S2a4a, the cluster partition C is initialized as... t=1,2...k, t = 1, 2...k; In step S2a4b, for i = 1, 2...m, the vq algorithm is used to calculate sample x. i and each centroid vector μj The distance (j=1,2,...k) will be x i The smallest one is d. ij The corresponding category λ i At this point, update Cλ. i =Cλ i ∪{x i In step S2a4c, for j = 1, 2, ..., k, recalculate the new centroids for all sample points in Cj. If the centroid vectors change, repeat step S2a4. If none of the k centroid vectors change, proceed to step S2a5 and output the cluster partition C = {C1, C2, ..., Ck}.

[0135] In step S2b, invalid exploration wells are classified. Invalid exploration wells refer to the fourth category of exploration wells.

[0136] The specific implementation process is as follows: Figure 5 As shown. In step S2b1, based on the clustering results of valid wells, the centroids of each class are taken as U = {μ1, μ2, ... μj} (j = 1, 2, ... k). In step S2b2, for invalid wells s = 1, 2, ..., S, the distance between each invalid well si and each centroid vector μj (j = 1, 2, ... k) is calculated using Euclidean distance. The class λi corresponding to the smallest si is marked as dij. At this time, Cλi = Cλi ∪ {xi} is updated. In step S2b3, the final cluster partition Cs = {Cs1, Cs2, ... Csk} is output. In step S2b4, data for each class is obtained, and the Graham scan algorithm is used to obtain the set of points constituting each convex hull. In step S2b4a, the lowest point on the y-axis is taken as the starting point p0. In step S2b4b, polar coordinate scanning starts from p0, and all points in the graph are traversed in turn, according to the polar coordinate angle, in a counterclockwise direction. In step S2b4c, if a newly traversed point can generate a left rotation, then add that point to the convex hull; otherwise, discard it. Finally, execute step S2b5 to draw the convex hull for each cluster using the convex hull point set. During the classification process, the centroids of each category remain unchanged; that is, invalid wells do not affect the clustering results of valid wells, only changing the category coverage.

[0137] In step S3, a regional single-well reserve growth contribution and investment analysis is performed. Based on the results of the oil and gas reservoir regional analysis, all reported reserves are divided according to the well cluster coverage area. The total reserve growth for each region is calculated, and then the equivalent reserve growth contribution of each well (such as the average reserve growth contribution) is calculated. Similarly, the total regional investment is calculated, and the average investment for each well is calculated.

[0138] The specific implementation process is as follows: Figure 6As shown. In step S3a, based on the effective well clustering results, an effective well cluster division C = {C1, C2, ... Ck} is taken. In step S3b, based on the clustering results of all wells, an all well cluster division Cs = {Cs1, Cs2, ... Csk} is taken. In step S3c, the effective well cluster division C = {C1, C2, ... Ck} and the all well cluster division Cs = {Cs1, Cs2, ... Csk} are traversed, the number of effective wells and the number of all wells are calculated, and their percentages are obtained. In step S3d, the effective well cluster division Co = {Co1, Co2, ... Cok} (which stores the number of each data point in the original data list) is traversed, the reserves of the corresponding reporting block are found according to the number, the number of reporting blocks and the cumulative reserves of the region are calculated, and the average reserves per well are obtained. In step S3e, all well clusters are iterated through, Cs = {Cs1, Cs2, ..., Csk}, and the total number of wells is calculated to obtain the percentage of reported blocks and all wells. In step S3f, all well clusters are iterated through, Cos = {Cos1, Cos2, ..., Cosk} (containing the number of each data point in the original data list), and the drilling investment corresponding to the well name is found based on the number. The number of wells and the cumulative investment for the region are calculated to obtain the average investment per well. In step S3f, the centroids of each type are iterated through, U = {μ1, μ2, ..., μj} (j = 1, 2, ..., k), and the calculated data is displayed above the centroids of each region.

[0139] In step S4, an exploration well location deployment planning scheme analysis is performed. The analysis uses goal programming, with constraints including expected reserve growth and planned investment, where expected reserve growth is a mandatory condition. The expected reserve growth and investment for each exploration well in all regions are examined, and regional exploration well allocation schemes are calculated using both the minimum and maximum regional involvement strategies, followed by the required investment. Investments exceeding the required amount are considered risky exploration well investments, and risky exploration well locations are not recommended.

[0140] The specific implementation process is as follows: Figure 7As shown in the figure. In step S4a, obtain block q and the number of clusters k from the model. First, calculate the average reserves Ra = {Ra1, Ra2, Ra3, Ra4….Rak} and the average investment Ia = {Ia1, Ia2, Ia3, Ia4,…..Iak} for all sample point partitions C = {C1, C2,...Ck} corresponding to each class. At the same time, remove the data with an average reserve of 0 in the class. In step S4b, let the investment be I and the expected reserve be R. By traversing the average reserve set Ra and the average investment set Ia of the clusters, obtain the most greedy investment-reserve expectation RImax. That is, through RImax, the minimum investment Imin can be calculated to obtain the maximum reserve Rmax. In step S4c, judge whether Rimax satisfies the investment I and the expected reserve R at this time. If it does not satisfy, we relax the investment condition and calculate the minimum investment Imin’ that can meet the expected reserve R. At this time, Imin’ is greater than I. In step S4d, if RImax satisfies the investment I and the expected reserve, that is, there exists a certain scheme E = {e1, e2, e3….}, such that the investment Ie = {Ie1, Ie2, Ie3…..} corresponding to each scheme satisfies Ie <= -I, and the reserve Re = {Re1, Re2, Re3….} that each scheme can obtain satisfies Re >= R. We obtain some schemes Eg = {Eg1, Eg2…..} by traversing, such that these schemes satisfy: the most greedy scheme, drilling wells in the area with the largest I / R ratio; average well drilling, when the investment I’ < I, drilling wells evenly in the area with reserves.

[0141] Example 3:

[0142] In a specific Example 3 of applying the present invention, the method for constructing a short-term well location deployment planning model in a mature exploration area includes the following steps:

[0143] In step P1, analyze the effectiveness and influence of exploration wells.

[0144] In step P1a, classify the exploration wells according to whether there is oil and gas show, electric logging show in the reservoir of the exploration well, and the result after running casing and testing for oil. Classify the exploration wells with the conclusion of oil layer (including reservoirs with industrial oil and gas flow), heavy oil layer, and water-bearing oil layer into the first category, classify the exploration wells with the conclusion of oil-water co-layer and water-bearing oil layer into the second category, classify the exploration wells with oil and gas show in the reservoir or electric logging show that may have oil and gas reservoirs into the third category, and the others into the fourth category.

[0145] In step P1b, the influence of the exploration well is analyzed based on the geological conditions of the exploration well and the constraints of adjacent exploration wells. In the Bohai Bay Basin, where geological conditions change very rapidly, the reservoir can still be considered homogeneous within a limited area around the exploration well. Therefore, the effectiveness of a single exploration well can cover a certain range. In mature exploration areas, due to the large number of exploration wells, the coverage range of a single well is constrained by adjacent exploration wells, and the size of the coverage radius has little impact on most areas.

[0146] In step P1c, the study area is divided into several grids, and the effectiveness of the well is radiated with the grid where the well is located as the center to obtain the distribution of the influence of the well's effectiveness.

[0147] Step P1c1: Connect to the database and obtain initial data, including the x and y coordinates and classification number of the exploration well. First, normalize the x and y coordinates of the data.

[0148] Step P1c2: Multiply the coordinate-normalized data by the grid size factor and round it down. Then merge the data according to the coordinates, retaining the optimal value for each category.

[0149] Step P1c3: Create a grid image. Based on the merged data, fill the image with the classification results according to the coordinate index. At this time, the image contains the classification information of all exploration wells.

[0150] Step P1c4: Assign validity values ​​to the grid image. Based on the different classifications of the exploration wells, assign different validity values ​​B1, B2, B3, and B4 to the corresponding locations, with the validity decreasing in that order.

[0151] Step P1c5: Perform radial iteration based on the total number of iterations D1. Each iteration traverses the entire image: If the validity value b of a point (x, y) on the image is in {B1, B2, B3} and the number of the main iteration d is less than the number of the secondary iteration D2, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value bi to the point that does not contain a validity value; If the validity value b of a point (x, y) on the image is B4, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value b to the point that does not contain a validity value.

[0152] Step P1c6: After the iteration is complete, output the influence distribution map.

[0153] In step P2, an oil and gas reservoir analysis is performed.

[0154] Step P2a involves clustering exploration wells based on their effectiveness classification results; this includes the following steps:

[0155] Step P2a1: Connect to the database, obtain the initial data, and normalize the original data.

[0156] Step P2a2: Assign different weight coefficients to each feature;

[0157] Step P2a3: Randomly select k samples from the processed data as the initial k centroid vectors: {μ1,μ2,...,μk};

[0158] Step P2a4: Cluster the effective exploration wells around the k cluster centers;

[0159] Step P2a4a, for n = 1, 2, ..., N: Initialize the cluster partition C as follows t=1,2...k, t = 1, 2...k;

[0160] Step P2a4b, for i = 1, 2... m, use the vq algorithm to calculate sample x. i and each centroid vector μ j The distance (j=1,2,...k) will be x i The smallest one is d. ij The corresponding category λ i At this point, update Cλ i =Cλ i ∪{x i};

[0161] Step 2a4c: For j = 1, 2, ..., k, recalculate the new centroids for all sample points in Cj. If the centroid vectors change, repeat step 2a4a; if all k centroid vectors remain unchanged, proceed to step 2a5.

[0162] Step P2a5 outputs the cluster partition C = {C1, C2, ..., Ck}.

[0163] Step P2b, based on the clustering results of valid wells, classifies invalid wells; includes the following steps:

[0164] Step P2b1: Based on the effective well clustering results, take the centroid U = {μ1, μ2, ... μj} (j = 1, 2, ... k) of each class;

[0165] Step P2b2, for invalid wells s = 1, 2, ..., S. Use Euclidean distance to calculate the distance between each invalid well si and each centroid vector μj (j = 1, 2, ... k), and label the si with the smallest distance as the category λi corresponding to dij. At this time, update Cλi = Cλi∪{xi}.

[0166] Step P2b3, output the final cluster partition Cs = {Cs1, Cs2, ..., Csk};

[0167] Step P2b4: Obtain each type of data and use the Graham scan algorithm to obtain the set of points that constitute each convex hull;

[0168] Step P2b4a: Start from the lowest point on the y-axis as the starting point p0;

[0169] Step P2b4b: Start the polar coordinate scan from p0, and traverse all points in the graph in turn, according to the polar coordinate angle, in a counterclockwise direction.

[0170] In step P2b4c, if the newly traversed point can produce a left rotation, add the point to the convex hull; otherwise, discard it.

[0171] In step P2b5, use the set of convex hull points to draw the convex hull of each cluster.

[0172] In step P3, an analysis of the contribution of regional single-well reserves growth and investment is conducted; the specific steps are as follows:

[0173] Step P3a: Based on the effective well clustering results, take the effective well cluster division C = {C1, C2, ... Ck};

[0174] Step P3b: Traverse the effective well clusters C = {C1, C2, ... Ck} to obtain the number of effective wells in each cluster;

[0175] Step P3c: Traverse the effective well cluster division Co = {Co1, Co2, ..., Cok} (storing the number of each data point in the original data list), find the reserves of the corresponding reporting block based on the number, and calculate the number of reporting blocks and the cumulative reserves of the region.

[0176] Step P3d: Calculate the average reserves per well based on the number of exploration wells and the cumulative reserves in the region.

[0177] Step P3e: Based on the clustering results of all exploration wells, divide all exploration well clusters into Cs = {Cs1, Cs2, ..., Csk};

[0178] Step P3f: Traverse all well clusters and divide them into Cs = {Cs1, Cs2, ..., Csk}, and calculate the total number of wells.

[0179] Step P3g: Based on the number of reported blocks and the total number of exploration wells, obtain the percentage of the number of reported blocks and the total number of exploration wells;

[0180] Step P3h: Traverse all well clusters and divide them into Cos = {Cos1, Cos2, ... Cosk} (store the number of each data in the original data list). Find the drilling investment of the corresponding well name based on the number, calculate the number of wells and the cumulative investment in the region, and obtain the average investment per well.

[0181] Step P3i, traverse the centroids U = {μ1, μ2,... μj} (j = 1, 2,... k) of each type, and display the calculated data above the centroid of each region.

[0182] In step P4, with the expected reserve growth and planned investment as the constraint conditions, use the goal programming method to conduct exploration well location deployment planning analysis, and formulate an exploration well location deployment planning scheme, which specifically includes the following steps:

[0183] Step P4a, obtain the block q and the number of clusters k from the model. First, calculate the average reserve Ra = {Ra1, Ra2, Ra3, Ra4…. Rak} and the average investment Ia = {Ia1, Ia2, Ia3, Ia4,….. Iak} of all sample point partitions C = {C1, C2,... Ck} corresponding to each class. At the same time, remove the data with an average reserve of 0 in the class;

[0184] Step P4b, set the investment as M and the expected reserve as R. By traversing the average reserve set Ra and the average investment set Ma of the clusters, obtain the most greedy investment-reserve expectation RMmax, that is, through RMmax, the minimum investment Mmin can be calculated to obtain the maximum reserve Rmax;

[0185] Step P4c, judge whether Rimax meets the investment M and the expected reserve R at this time. If it does not meet, we relax the investment condition and calculate the minimum investment Mmin' that can meet the expected reserve R. At this time, Mmin' is greater than M;

[0186] Step P4d, if RMmax meets the investment I and the expected reserve, that is, there is a certain plan E = {e1, e2, e3….}, such that the investment Ie = {Ie1, Ie2, Ie3…..} corresponding to each plan satisfies Ie <= -I, and the reserve Re = {Re1, Re2, Re3….} that each plan can obtain satisfies Re >= R. We obtain some plans Eg = {Eg1, Eg2…..} by traversing, such that these plans meet: the most greedy plan, drill wells in the region with the largest M / R ratio; average well placement, when the investment M' < M, drill wells evenly in the regions with reserves.

[0187] The present invention approximately represents the geological structure distribution by the exploration well distribution, takes the exploration well effectiveness and the reserve growth contribution as the main basis, uses the artificial intelligence method to divide the oil and gas reservoir distribution area, and constructs an exploration well location deployment planning model. Using this model, multiple exploration well location deployment plans can be accurately and quickly planned for decision-making reference, without the need for the geological structure factors of each specific region.

[0188] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0189] Except for the technical features described in the specification, all other technologies are known to those skilled in the art.

Claims

1. A method for constructing a short-term well location deployment planning model in mature exploration areas, characterized in that, The methods for constructing short- and medium-term well location deployment planning models in this mature exploration area include: Step 1: Analyze the effectiveness and impact of exploration wells based on reservoir oil and gas conditions and constraints from adjacent exploration wells during the drilling process. Step 2: Analyze the oil and gas reservoir areas based on the results and impact distribution of exploration wells; Step 3: Based on the results of the oil and gas reservoir area analysis, analyze the contribution of single wells to the growth of reserves and the investment situation in the region; Step 4: Using the expected growth of reserves and planned investment as constraints, conduct well site deployment planning analysis using the target planning method, and formulate a well site deployment planning scheme; Step 1 includes: Step 1a: Classify the effectiveness of the exploration well based on the reservoir oil and gas condition information, including whether the reservoir has oil and gas indications or electrical logging indications during the drilling process and the results and conclusions after the casing test. Step 1b: Conduct an impact analysis on the exploration well based on its geological conditions and the constraints of adjacent exploration wells; Step 1c: Divide the study area into several grids, and radiate the well effectiveness with the grid where the well is located as the center to obtain the distribution of well effectiveness and influence. Step 2 includes: Step 2a: Perform effectiveness clustering on the exploration wells based on the effectiveness classification results; Step 2b: Based on the clustering results of valid exploration wells, classify invalid exploration wells. Step 3 includes: Step 3a: Based on the effective well clustering results, take the effective well cluster division C={C1,C2,...Ck}; Step 3b: Based on the clustering results of all exploration wells, divide all exploration well clusters into Cs={Cs1,Cs2,...Csk}; Step 3c: Traverse the effective well cluster division C={C1,C2,...Ck} and all exploration well cluster divisions Cs={Cs1,Cs2,...Csk}, calculate the number of effective exploration wells and the total number of exploration wells, and obtain their percentages; Step 3d: Establish an effective well cluster division Co={Co1,Co2,...Cok}, store the number of each data in the original data list, find the reserves of the corresponding reporting block according to the number, calculate the number of reporting blocks and the cumulative reserves of the region, obtain the average reserves of a single well, traverse all exploration well cluster divisions Cs={Cs1,Cs2,...Csk}, calculate the number of all exploration wells, and obtain the percentage of the number of reporting blocks and the total number of exploration wells; Step 3e: Establish all well clusters Cos={Cos1,Cos2,...Cosk}, store the number of each data in the original data list, find the drilling investment of the corresponding well name according to the number, calculate the number of wells and the cumulative investment in the region, and obtain the average investment per well. Step 3f, traverse the centroids U={μ1,μ2,...μ k }(j=1,2,...k), display the data calculated in step 3e above the centroid of each region; Step 4 includes: Step 4a: Obtain block q and number of clusters k from the model. First, calculate the average reserves Ra = {Ra1, Ra2, Ra3, Ra4, ..., Rak} and average investment Ia = {Ia1, Ia2, Ia3, Ia4, ..., Iak} for all sample point clusters C = {C1, C2, ..., Ck} corresponding to each class; at the same time, remove data with an average reserves of 0 from the class. Step 4b: Let the investment be I and the expected reserves be R. By traversing the average reserve set Ra and the average investment set Ia of the clusters, the most greedy investment-reserve expectation RImax is obtained. That is, through RImax, the minimum investment Imin can be calculated to obtain the maximum reserves Rmax. Step 4c: Determine whether RImax satisfies the investment I and the expected reserves R at this time. If it does not satisfy, loosen the investment condition, and through calculation, obtain the minimum investment Imin’ that can meet the expected reserves R. At this time, Imin’ is greater than I. Step 4d: If RImax satisfies the investment I and the expected reserves, that is, there exists a certain set of solutions E = {e1, e2, e3….}, such that the investments Ie = {Ie1, Ie2, Ie3…..} corresponding to each solution satisfy Ie <= -I, and the reserves Re = {Re1, Re2, Re3….} that each solution can obtain satisfy Re >= R; through traversal, obtain a partial set of solutions Eg = {Eg1, Eg2…..}, such that these solutions satisfy: the most greedy solution, drilling wells in the area with the largest I / R ratio; average well placement, when the investment I’ < I, drilling wells evenly in the area with reserves.

2. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 1, characterized in that, In step 1a, according to the oil and gas conditions of the exploration wells passing through the reservoir, the exploration wells are classified for effectiveness. The exploration wells with oil layers, heavy oil layers, and water-bearing oil layers are the first category, the exploration wells with oil-water layers and water-bearing oil layers are the second category, the exploration wells with oil and gas shows in the reservoir or possible oil and gas reservoirs shown by electric logging are the third category, and others are the fourth category.

3. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 1, characterized in that, In step 1b, for areas with very fast-changing geological conditions, within a limited area around the exploration well, the reservoir can still be considered homogeneous. Therefore, the effectiveness of a single exploration well can cover a certain range; in mature exploration areas, due to the large number of exploration wells, the coverage range of a single well is restricted by adjacent exploration wells, and the size of the coverage radius has little impact in most areas.

4. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 1, characterized in that, Step 1c includes: Step 1c1: Connect to the database to obtain initial data, including the x and y coordinates and classification of the exploration wells, and normalize the x and y coordinates of the data. Step 1c2: Multiply the normalized coordinate data by the grid size coefficient and round it, and merge the data according to the coordinates, where the classification retains the optimal value. Step 1c3: Create a grid image, and according to the merged data, fill the classification results into the image according to the coordinate index. At this time, the image contains the classification information of all exploration wells. Step 1c4: Assign effectiveness values to the grid image. According to the different classifications of the exploration wells, assign different effectiveness values A1, A2, A3, A4 to the corresponding positions, and the effectiveness decreases in turn. Step 1c5: Perform radial iteration based on the total number of iterations S1. Each iteration traverses the entire image: If the validity value a of a point (x, y) in the image is in {A1, A2, A3} and the iteration number s is less than the next iteration number S2, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value ai to the point that does not contain a validity value; If the validity value a of a point (x, y) in the image is A4, then perform radial judgment on the four points above, below, left, and right of that point, and assign the validity value a to the point that does not contain a validity value. Step 1c6: After the iteration is complete, output the influence distribution map.

5. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 4, characterized in that, Step 2a includes: Step 2a1: Connect to the database, obtain the initial data, and normalize the original data; Step 2a2: Assign different weight coefficients to each feature; Step 2a3: Randomly select k samples from the processed data as the initial k centroid vectors: {μ1, μ2, ..., μ k }; Step 2a4: Cluster the data around the k centroids to divide it into different clusters; Step 2a5, output the cluster partition C={C1,C2,...Ck}.

6. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 5, characterized in that, Step 2a4 includes: Step 2a4a, for n=1,2,...,N: initialize the cluster partition C as Ct=∅(t=1,2...k); Steps 2a4b: For i=1,2...m, use the vq algorithm to calculate sample x. i and each centroid vector μ j The distance (j=1,2,...k) will be x i The smallest one is d. ij The corresponding category λ i At this point, update Cλ i =Cλ i ∪{x i }; Step 2a4c: For j=1,2,...,k, recalculate the new centroids for all sample points in cluster C; if the centroid vectors change, repeat step 2a4a; if all k centroid vectors remain unchanged, proceed to step 2a5.

7. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 1, characterized in that, Step 2b includes: Step 2b1: Based on the effective well clustering results, take the centroid x = {μ1μ2...μ} of each class. k }(j=1,2,...k); Step 2b2, for invalid exploration wells s=1,2,...,S, calculate the s of each invalid exploration well using Euclidean distance. i and each centroid vector μ j The distance (j=1,2,...k) will be s i The smallest one is d. ij The corresponding category λ i At this point, update Cλ i =Cλ i ∪{s i }; Step 2b3, output the final cluster partition Cs={Cs1,Cs2,...Csk}; Step 2b4: Obtain each type of data and use the Graham scan algorithm to obtain the set of points that constitute each convex hull; Step 2b5: Using the set of convex hull points, draw the convex hull of each cluster.

8. The method for constructing a short-term well location deployment planning model in mature exploration areas according to claim 7, characterized in that, Step 2b4 includes: Step 2b4a: Start from the lowest point on the y-axis as the starting point p0; Step 2b4b: Start with p0 and perform a polar coordinate scan, traversing all points in the graph in a counter-clockwise direction according to the polar coordinate angle. In step 2b4c, if the newly traversed point can produce a left rotation, add the point to the convex hull; otherwise, discard it.

Images