Reserve exploration area evaluation method and system based on classification algorithm

Abstract

The invention relates to the technical field of intelligent geological exploration, in particular to a reserve exploration area evaluation method and system based on a classification algorithm, and the method comprises the following steps: extracting an included angle between a tectonic line and a lithologic zone trend, and generating a tectonic trend label according to space combination, calling a label to obtain mineralized element concentration, calculating a distance between a distribution center and a mass center to generate an enrichment type, dividing density grades through an area ratio, and generating an alteration gradient classification set in combination with a grid thickness trend and a mineral concentration ratio. According to the method, the spatial combination is identified by extracting the relationship between the structure and the lithology direction, the coupling expression of the geologic structure and the metallogenic mode is improved, the feature consistent region is screened by combining the spatial distance between the mineralized element enrichment center and the regional mass center, the target region focusing precision is improved, and the abnormal cluster is extracted based on the convergence of the abnormal body point location and the quantity feature; abnormal intensity quantification is enhanced, and the multi-factor collaborative recognition path significantly enhances the multi-source data fusion judgment capability and spatial precision.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent geological exploration, and in particular to a method and system for evaluating a reserve exploration area based on a classification algorithm. Background Art

[0002] The field of intelligent geological exploration technology encompasses a diverse system of methods for improving the accuracy and efficiency of geological prospecting based on informatization and data. The core of this technology lies in the spatial prediction and detailed evaluation of underground mineral resources through the comprehensive processing and integrated analysis of multi-source geological data, including remote sensing, geophysics, and geochemistry. Its overall technical system encompasses the automated collection of geological data, the identification of geological anomalies, resource potential modeling, and the delineation of prediction areas, striving to enhance the scientific nature and standardization of geological exploration work through intelligent means. Currently, intelligent geological exploration has been gradually applied to key links such as the site selection of reserve exploration areas, the study of mineralization patterns, and the optimization of prospecting targets, becoming an important support method for improving the ability to secure strategic mineral resources.

[0003] Among them, the reserve exploration area evaluation method based on classification algorithm refers to a method of classifying, judging and grading the reserve exploration areas to be evaluated by constructing an evaluation sample system including geological background characteristics, resource potential indicators, exploration degree information and economic feasibility parameters using a supervised classification algorithm. The technical matters targeted by this patent subject cover the data modeling and quantitative evaluation issues required for the pre-selection stage of reserve exploration areas. Specifically, based on the attribute indicators such as structural type, lithologic combination, ore body morphology, width of surrounding rock alteration zone, and abundance value of mineralized elements extracted from historical exploration results, a support vector machine model or decision tree classifier is constructed through the training set to classify and identify samples of new exploration areas, complete the evaluation and grading, and form a data support system for reserve planning.

[0004] Existing technologies primarily rely on the overlay analysis and statistical modeling of multi-source geological information in data processing. However, at the level of quantitative analysis of spatial structure, they lack the ability to identify the directional coupling relationship between structural and lithologic trends, resulting in a crude representation of structurally controlled mineralization patterns and limiting the extraction of detailed mineralization patterns. Analysis of the spatial distribution of mineralized elements often relies on concentration threshold screening, ignoring the spatial relationship between enrichment centers and regional structure, making it difficult to identify weak anomaly concentrations and, consequently, missing potential mineralization areas. Anomaly identification fails to incorporate the spatial statistical characteristics of point distribution, easily equating discrete anomalies and reducing the reliability of focused anomaly identification. Resource potential assessment primarily relies on qualitative descriptions or static indicators, lacking density quantification and hierarchical expression methods, making it difficult to support differential analysis of spatial distribution. Alteration zone assessment typically relies on changes in the concentration of a single mineral, without incorporating thickness trends, limiting the depth of expression of spatial gradients in alteration intensity. These deficiencies result in blind spots in the identification of preferred areas during initial reserve exploration, resulting in reduced prospecting efficiency and spatial accuracy, and the easy omission of potentially weak anomaly areas, impacting the scientific and systematic nature of reserve planning. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a reserve exploration area evaluation method and system based on a classification algorithm.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for evaluating reserve exploration areas based on a classification algorithm, comprising the following steps:

[0007] S1: Acquire structural line and lithologic zone trend data, extract the directional angle between structural joints and lithologic boundaries, divide structural trend combinations according to spatial position, extract the combination relationships within the matching range of directional differences, establish structural and lithologic trend classification samples, and generate a structural trend classification label set;

[0008] S2: calling the spatial number in the structural trend classification label set, extracting the concentration data of multiple mineralized elements at corresponding positions, calculating the spatial distance between the element distribution center and the centroid of the geological area, extracting areas within the distance threshold and marking them as feature-consistent types, and generating an enrichment type field set;

[0009] S3: Calling the coded information in the enrichment type field set, extracting the coordinate data of the abnormal body in the target area, calculating the average distance and distribution standard deviation of the point to the center position, identifying the distance convergence area, and establishing a spatial classification field based on the number of abnormal points to generate a focused classification feature set;

[0010] S4: Call the target area number corresponding to the focused classification feature set, obtain the occurrence quantity and spatial area data in the area, calculate the unit area ratio, divide the results into multiple levels according to the natural segmentation method and encode and label them to generate an occurrence density level label set.

[0011] As a further solution of the present invention, the structural trend classification label set includes structural strike combination type, direction matching interval, and lithologic trend category; the enrichment type field set includes element distribution center coordinates, centroid distance threshold, and feature consistent area identifier; the focused classification feature set includes average distance value of outliers, distribution standard deviation parameter, spatial convergence area identifier, and outlier quantity classification; the occurrence density level label set includes unit area ratio data, natural segmentation level division, and density level coding.

[0012] As a further solution of the present invention, the specific steps of S1 are:

[0013] S101: Acquire spatial coordinate information of the structural line and the lithologic zone boundary, extract the structural line direction and the lithologic zone strike based on the path direction relationship between the line segments, calculate the angle between the structural and lithologic boundaries by using the node direction difference, and obtain a structural and lithologic direction angle sequence;

[0014] S102: clustering the structural lines according to the spatial distribution position in the structural and lithologic direction angle sequence, and screening them based on the angle value and the set interval to extract structural and lithologic combinations that meet the conditions, and obtain a matching trend combination comparison table;

[0015] S103: extracting the direction difference set and combination identifier of structure and lithology based on the combination information in the matching trend combination comparison table, calculating the difference variation and establishing a classification map to generate a structural lithology trend classification label set.

[0016] As a further solution of the present invention, the specific calculation formula for calculating the difference variability is:

[0017]

[0018] Calculate the difference variation index, establish a classification map, and generate a structural lithology trend classification label set;

[0019] Among them, H represents the difference variability index of the structure and lithology direction difference set, represents the absolute difference of the eth element in the set of structural and lithological direction differences, w e is the normalized weight factor corresponding to the combination identifier, μ Δ is the weighted arithmetic mean of the direction difference set, k is the lithologic trend adjustment coefficient, λ g is the matching strength parameter of the g-th combination identifier, max(Δd ) and min(Δ d ) represent the maximum and minimum values ​​of the direction difference set, is the unbiased variance of the directional difference, z is the total number of elements in the directional difference set, and y is the total number of combined identification categories.

[0020] As a further solution of the present invention, the specific steps of S2 are:

[0021] S201: calling the spatial number in the structural lithologic trend classification label set, extracting the mineralized element concentration data at the corresponding position, associating the concentration value with the number according to the spatial coordinate, establishing a spatial distribution set of the mineralized elements, and generating an element distribution value matrix;

[0022] S202: Based on the element distribution value matrix, extract the distribution center of the mineralized elements, calculate the spatial distance between the distribution center of the mineralized elements and the center of mass of the geological area, extract the offset distance data of all elements, and generate an element offset distance value set;

[0023] S203: Based on the element offset distance value set, the spatial distance threshold in the trend classification label set is called to filter the numbered areas within the range, mark the corresponding numbers as feature consistent types, add a mark field, and generate an enriched type field set.

[0024] As a further solution of the present invention, the specific steps of S3 are:

[0025] S301: calling the coding information and anomaly distribution data in the enrichment type field set, extracting the coordinate values ​​of the anomalies in the target area, calculating the distance from the point to the center position, summarizing the average distance and standard deviation of all points, and generating a point distance distribution parameter set;

[0026] S302: Calculate the difference between the point and the center position based on the average distance and standard deviation in the point distance distribution parameter set, perform screening, select a concentrated point set, calculate the spatial density characteristics, and generate a coefficient set for the outlier dense area;

[0027] S303: Calling the outlier dense area coefficient set and outlier quantity data, summarizing the number of outliers in the spatial area, performing classification judgment based on the outlier proportion and density conditions, and generating a focused classification feature set.

[0028] As a further solution of the present invention, the specific calculation formula for the number of outliers in the summary spatial area is:

[0029]

[0030] Calculate the summary value of spatial outliers and generate a focused classification feature set;

[0031] Among them, Nk Represents the total number of outliers in the k-th spatial region, w i Represents the coefficient of the ith outlier dense area, d i represents the number of abnormal points in the corresponding area, c i represents the spatial radius of the ith dense area, α represents the smoothing constant, β represents the cross-region influence weight factor, and p j represents the outlier impact factor of the jth adjacent area, γ represents the benchmark adjustment amount, ∈ j Represents the distance attenuation coefficient of the jth adjacent area.

[0032] As a further solution of the present invention, the specific steps of S4 are:

[0033] S401: Call the target area number corresponding to the focused classification feature set, extract the area boundary and attribute field information under the number, extract the target objects in the area based on the boundary, count the occurrence quantity and spatial area in each area, calculate the area ratio between the two data, and generate the occurrence value per unit area;

[0034] S402: Based on the occurrence value per unit area, the ratio data of all regions are summarized, the grade interval boundaries are established by sorting, and the data are divided into multiple grade segments according to the value range to generate the occurrence density grade interval value;

[0035] S403: According to the occurrence density level interval value, the unit area occurrence value of each area is called to determine the level interval to which it belongs, and a corresponding level coding identifier is assigned. A corresponding set of numbers and regional spatial positions is constructed, and the coding layer is superimposed to generate an occurrence density level label set.

[0036] As a further embodiment of the present invention, the method further comprises:

[0037] S5: calling the high-level area number in the occurrence density level label set, extracting the grid thickness sequence and calculating the adjacent change trend, obtaining the concentration data of key minerals and calculating the ratio with the background concentration, extracting the spatial position that meets the change trend and concentration difference characteristics, and generating the alteration gradient classification field set;

[0038] The alteration gradient classification field set includes grid thickness variation sequence, concentration difference ratio, and gradient trend matching area;

[0039] The specific steps of S5 are:

[0040] S501: calling the high-level area number in the density level label set, matching the spatial grid number and extracting the thickness sequence of the corresponding area, calculating the adjacent thickness difference according to the adjacent grid number sequence, and judging the thickness change direction according to the difference sign, and generating the adjacent thickness change trend value;

[0041] S502: extracting the concentration value of key minerals from the corresponding spatial positions according to the adjacent thickness change trend values, calculating the concentration reference value of the area, and comparing the ratio of the point concentration value to the reference value. Based on the trend change direction, select the positions where the concentration ratio is greater than the reference value to generate a concentration trend compliance position set;

[0042] S503: Extract the occurrence density grade, thickness change trend value and mineral concentration ratio according to the concentration trend compliance position set, construct a feature combination and divide the combination interval, fill the classification number into the spatial grid field record, and generate the alteration gradient classification field set.

[0043] A reserve exploration area evaluation system based on a classification algorithm includes:

[0044] The structural classification module obtains spatial trend data of structural lines and lithologic zones, calculates the directional angle between structural joints and lithologic boundaries, uses spatial partitioning information to divide structural combinations by number, extracts the directional difference between each structural and lithologic combination, determines whether the fluctuation range of the directional difference sequence meets the classification criteria, incorporates the combination relationships that meet the criteria into the sample set, extracts the trend mean of the combination and binds it to a number to generate a structural trend classification label set;

[0045] The enrichment extraction module calls the number information in the structural trend classification label set, extracts the concentrations of multiple mineralized elements in the corresponding area, calculates the spatial distance between the element distribution center and the regional centroid, selects the areas that meet the distance threshold as consistency types, and generates an enrichment type field set;

[0046] The focus identification module calls the coded information in the enrichment type field set, extracts the coordinate data of the abnormal body in the corresponding area, calculates the average distance and standard deviation of the abnormal point set to the center position, selects the numbered areas with strong convergence and records the number of abnormal points, and generates a focus classification feature set;

[0047] The density assessment module calls the target area number in the focused classification feature set, extracts the regional occurrence quantity and spatial area, calculates the unit area ratio, and divides the levels in a segmented manner, assigns level identifiers, and generates an occurrence density level label set;

[0048] The alteration classification module calls the high-level number in the occurrence density level label set, extracts the grid thickness sequence of the corresponding area, calculates the adjacent change trend and determines the directional stability, extracts the concentration of key minerals and performs ratio analysis with the background value, screens the spatial locations that meet the trend and concentration difference characteristics, and generates the alteration gradient classification field set.

[0049] Compared with the prior art, the advantages and positive effects of the present invention are:

[0050] In the present invention, the spatial combination is identified by extracting the relationship between structure and lithologic direction, the coupled expression of geological structure and mineralization mode is improved, the spatial distance between the mineralization element enrichment center and the regional centroid is combined to screen the characteristic consistent area, the focus accuracy of the target area is improved, the anomaly cluster is extracted based on the convergence and quantitative characteristics of the anomaly point, the quantification of anomaly intensity is enhanced, the occurrence density level is divided by unit area ratio, the expression of resource distribution difference is strengthened, the thickness trend and concentration ratio are integrated to extract the alteration gradient area, and the alteration intensity positioning capability is improved. The multi-factor collaborative identification path significantly enhances the multi-source data fusion judgment capability and spatial accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the steps of the present invention

[0052] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0054] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0055] The method of the present invention is mainly applicable to the following three types of minerals and regional types:

[0056] (1) Applicable Mineral Scope: This includes 12 scarce metal minerals, including iron, manganese, chromium, copper, nickel, cobalt, gold, lithium, beryllium, niobium, tantalum, and platinum group metals; 2 scarce non-metallic minerals, including potassium salt and boron; and 14 advantageous minerals, including coal, tungsten, tin, molybdenum, antimony, vanadium, titanium, rare earth, indium, germanium, gallium, crystalline graphite, fluorite, and phosphorus. In addition, it is also applicable to regional key minerals such as silver, lead and zinc, talc, and vein quartz.

[0057] (2) Applicable target area types: This method can be used for the following four types of reserve exploration units with different resource bases and prospecting potentials:

[0058] Category A target areas: known mining areas where the sum of inferred resources (TD) and potential resources (QZ1) has reached the scale of medium-sized or above deposits;

[0059] Category B target area: exploration work has clearly delineated the ore body. Although TD has not been estimated, the potential resource QZ1 has reached a medium scale;

[0060] Category C target area: The ore-bearing geological body is clear, the mineralization conditions are favorable, and it is expected to become a medium-sized deposit through further exploration;

[0061] Category D target area: Exploration work has not yet been carried out, but the potential evaluation predicts that the QZ2 resources may reach medium or above scale.

[0062] The above-mentioned mineral species and regional types provide clear boundary conditions for the structural trend extraction, enrichment type discrimination and abnormal focus identification of the present invention, and can serve as an important priori basis for target area number screening and data input.

[0063] See also Figure 1 , a reserve exploration area evaluation method based on a classification algorithm, comprising the following steps:

[0064] S1: Acquire structural line and lithologic zone trend data, extract the directional angle between structural joints and lithologic boundaries, divide structural trend combinations according to spatial position, extract combination relationships within the matching range of directional differences, establish structural and lithologic trend classification samples, and generate structural trend classification label sets;

[0065] (Before calling the spatial number, priority can be given to screening the numbered areas that meet the A to D reserve target area type conditions to ensure the preliminary geological rationality of enrichment feature extraction).

[0066] S2: Call the spatial number in the structural trend classification label set, extract the concentration data of multiple mineralized elements at the corresponding positions, calculate the spatial distance between the element distribution center and the centroid of the geological area, extract the areas within the distance threshold and mark them as feature consistent types, and generate the enrichment type field set;

[0067] (Priority is given to numbering and screening the areas that have been classified as type A to D target areas in the enrichment type field concentration to enhance the geological rationality and mineralization prospect basis of the focused identification of anomalies).

[0068] S3: Call the coded information in the enrichment type field set, extract the coordinate data of the abnormal body in the target area, calculate the average distance and distribution standard deviation of the point to the center position, identify the distance convergence area, and establish the spatial classification field based on the number of abnormal points to generate the focused classification feature set;

[0069] S4: Call the target area number corresponding to the focused classification feature set, obtain the occurrence quantity and spatial area data in the area, calculate the unit area ratio, divide the results into multiple levels according to the natural segmentation method and perform coding and labeling to generate the occurrence density level label set;

[0070] S5: Call the high-level area number in the occurrence density level label set, extract the grid thickness sequence and calculate the adjacent change trend, obtain the concentration data of key minerals and calculate the ratio with the background concentration, extract the spatial position that meets the change trend and concentration difference characteristics, and generate the alteration gradient classification field set.

[0071] The structural trend classification label set includes structural strike combination type, direction matching interval, and lithologic trend category; the enrichment type field set includes element distribution center coordinates, centroid distance threshold, and feature consistent area identification; the focus classification feature set includes the average distance value of anomaly points, distribution standard deviation parameter, spatial convergence area identification, and anomaly number classification; the occurrence density level label set includes unit area ratio data, natural segmentation level division, and density level coding; the alteration gradient classification field set includes grid thickness change sequence, concentration difference ratio, and gradient trend matching area.

[0072] See also Figure 1 , the specific steps of S1 are:

[0073] S101: Acquire spatial coordinate information of the structural line and the lithologic zone boundary, extract the structural line direction and the lithologic zone strike based on the path direction relationship between the line segments, calculate the angle between the structural and lithologic boundaries by using the node direction difference, and obtain a structural and lithologic direction angle sequence;

[0074] The spatial coordinates of the boundaries of tectonic lines and lithologic zones can be obtained by parsing vector data from remote sensing images, geological maps, or 3D modeling results. The data format usually uses vector layers such as shp or GeoJSON. After loading through GIS tools, they are uniformly projected into a consistent coordinate system to ensure consistent spatial accuracy. The tectonic lines and lithologic zone segments are segmented according to nodes. The direction of each segment is determined by the coordinates of its two ends. The direction can be obtained by the difference between the coordinates of the two points. The direction information of all segments is converted into angle values, and the direction angles are normalized to the range of 0° to 180° to avoid negative values ​​and angles exceeding 180° interfering with the calculation. Then, the path direction of the tectonic line is determined according to the node order. Assuming that a certain tectonic line Connect the five spatial nodes in sequence to form four line segments. Calculate the direction angle of each segment, and then compare the direction angle with the line segments of the corresponding strike in the adjacent lithologic belt. Calculate the absolute difference between the direction angles of each group of structural segments and the adjacent lithologic belt segments. If the direction of a structural line segment is 60° and the lithologic belt is 45°, the angle is 15°. If the angle exceeds 90°, subtract the angle from 180° to correct it to an acute angle. Repeat the operation to construct a complete angle sequence. For example, in the five line segments, the calculated angles are 12.3°, 24.7°, 7.9°, 14.5°, and 20.1°, forming an angle series corresponding to the structural and lithologic directions for subsequent spatial distribution analysis.

[0075] S102: Clustering the structural lines according to the spatial distribution of the angle sequence between the structural and lithologic directions, and screening them based on the angle value and the set interval to extract structural and lithologic combinations that meet the conditions and obtain a matching trend combination comparison table;

[0076] The angle series and its corresponding spatial position constitute the input features. Clustering algorithms such as density-based methods are used to classify and divide structural lines. Before clustering, the angle values ​​and their corresponding coordinate values ​​need to be normalized to avoid the influence of scale differences on the clustering effect. For example, the angle range is 0° to 90°, and they are uniformly mapped to the interval of 0 to 1. At the same time, the coordinate values ​​are also standardized to form a composite feature vector for clustering operations. The density clustering algorithm parameters need to set the minimum number of neighboring points and the distance threshold. Common settings include the minimum number of points 3 and the distance threshold 0.05. Based on the clustering results, multiple spatial categories are obtained. The line segments in each category are screened according to their angle value distribution. For example, line segments with angles between 10° and 20° are screened. After screening, the spatial trend of the line segments is matched with the set structural trend standard. For example, the structural line trend is set to N45E with a tolerance of 15°, and the lithologic zone trend is set to N30E with a tolerance of 15°. If the direction of a structural line segment is 43

[0077] °, the lithologic belt is 31°, and the included angle between the two is 12°, which is within the allowable error range. It is considered that the structural and lithologic strikes are successfully matched. Such matching combinations are included to form a strike combination corresponding table. The table records include the structural line number, lithologic belt number, matching angle value and corresponding matching level code, which is convenient for subsequent identification and call.

[0078] S103: extracting the direction difference set and combination identifier of structure and lithology based on the combination information in the matching trend combination comparison table, calculating the difference variation and establishing a classification map to generate a structural lithology trend classification label set;

[0079] The specific calculation formula for calculating the difference variation is:

[0080]

[0081] Calculate the difference variation index, establish a classification map, and generate a structural lithology trend classification label set;

[0082] Among them, H represents the difference variability index of the structure and lithology direction difference set, represents the absolute difference of the eth element in the set of structural and lithological direction differences, w e is the normalized weight factor corresponding to the combination identifier, μ Δ is the weighted arithmetic mean of the direction difference set, k is the lithologic trend adjustment coefficient, λ g is the matching strength parameter of the g-th combination identifier, max(Δ d ) and min(Δ d ) represent the maximum and minimum values ​​of the direction difference set, is the unbiased variance of the directional difference, z is the total number of elements in the directional difference set, and y is the total number of combined identification categories;

[0083] Directional difference set: 5 sets of data are obtained through geological survey, specifically Δ d =[1.2°,0.8°,2.1°,1.5°,0.9°];

[0084] The total number of elements z = 5;

[0085] Combination identification category: According to the "Lithologic Combination Classification Standard", there are three types of combination identification, with a total number of categories y = 3;

[0086] Normalized weight factor: based on combined identifier matching data [0.85, 0.92, 0.78, 0.88, 0.81];

[0087] Calculate the weight factor:

[0088]

[0089] Weighted arithmetic mean:

[0090]

[0091] Matching strength parameters: According to the "Standard for Structural Lithology Strength Classification", the corresponding strength parameters of the three types of combination identifications are: λ g =[0.35,0.50,0.15];

[0092] Lithologic trend adjustment coefficient: Based on the ratio of the thickness of the Quaternary sedimentary layer to the exposed area of ​​bedrock in the "Guidelines for Regional Geological Stability Assessment" (the measured ratio is 1:0.83), k = 1.2 is obtained from the table;

[0093] Unbiased variance calculation:

[0094]

[0095] Range calculation: max(Δ d )-min(Δ d )=2.1°-0.8°=1.3°;

[0096] First molecular calculation:

[0097]

[0098] Calculation of the first denominator:

[0099]

[0100] First result:

[0101] Secondary calculation:

[0102] Final result: H = 0.348 + 4.377 = 4.725;

[0103] Result interpretation: The difference variation index H = 4.725 exceeds the Class III anomaly threshold of 4.0 in the "Structural and Lithologic Trend Classification Procedure", triggering the multi-parameter joint inversion process.

[0104] See also Figure 1 , the specific steps of S2 are:

[0105] S201: calling the spatial number in the structural lithologic trend classification label set, extracting the mineralized element concentration data at the corresponding position, associating the concentration value with the number according to the spatial coordinate, establishing the spatial distribution set of the mineralized elements, and generating an element distribution value matrix;

[0106] To access the spatial numbers in the tectonic lithologic trend classification label set, all numbers within the region and their corresponding trend type identifiers must first be extracted from the lithologic trend model file. For example, numbers ID001 to ID1000 correspond to trend types A, B, and C. Unambiguous attribution of trend classification is achieved by reading the model data line by line and extracting the TrendType field information. Next, the spatial coordinates and mineralized element concentrations corresponding to each number are extracted from the actual sampling dataset. The sampling data includes X and Y coordinates and element concentration values, such as 0.15% copper and 2.3 g / t gold. The sampling point coordinates are spatially matched to the number location, and the spatial join tool is used to establish a correspondence between element concentrations and numbers. For example, number ID150 located at coordinates 200 and 300 corresponds to a copper concentration of 0.12%. This matching result forms a joint dataset containing numbers, spatial locations, and element concentrations. The locations of all numbered points are then plotted within a two-dimensional area and annotated with their element concentrations, forming a preliminary distribution map of the mineralized elements. In the process of constructing the element distribution set, a two-dimensional structure matrix is ​​formed by structuring the numbers and corresponding element concentrations one by one. The horizontal axis is the number or coordinate point, and the vertical axis is different mineralized elements such as copper, lead, zinc, etc. Each cell in the matrix is ​​the element concentration value at the corresponding number. For example, the copper at ID150 is 0.12%, and at ID151 is 0.10%. Batch processing and output are performed through data spreadsheet tools or scripting languages ​​to obtain a complete element distribution value matrix.

[0107] S202: Based on the element distribution value matrix, extract the distribution center of the mineralized elements, calculate the spatial distance between it and the center of mass of the geological area, extract the offset distance data of all elements, and generate an element offset distance value set;

[0108] Based on the element distribution value matrix generated above, a concentration analysis is performed on the spatial distribution position of each element. This is done by performing a weighted position calculation based on the coordinate position of each numbered point combined with its element concentration value to determine the distribution center of gravity of the element across the entire area. For example, if numbered A is at positions 100 and 200, the copper concentration is 0.2%; if numbered B is at positions 120 and 220, the copper concentration is 0.3%; and if numbered C is at positions 140 and 240, the copper concentration is 0.5%. The coordinates of each numbered point are multiplied by the corresponding concentration value, the sum is then divided by the total concentration to determine the horizontal and vertical values ​​of the element at the center of the area. After obtaining the distribution center of gravity, the spatial distance between this location and the overall geometric center of the geological area is calculated. If the area is a square with a side length of 400 units, then its center of mass is 200, 200. By measuring the square of the horizontal and vertical distance offset between the center of gravity and the center of mass, and then taking the square root, the offset distance value of the element is obtained. For example, if the distribution center of gravity is 126, 226, the horizontal difference is 74 and the vertical difference is 26, and the total offset distance is calculated to be approximately 78.4 units. The same steps are used to calculate the distance between the distribution center of gravity of all mineralized elements and the center of mass of the area, and organize it into a set of numerical values ​​with element name and offset distance as units, completing the quantitative analysis of the offset of all elements in the area.

[0109] S203: Based on the element offset distance value set, the spatial distance threshold in the trend classification label set is called to filter the numbered areas within the range, mark the corresponding numbers as feature consistent types, add the mark field, and generate an enriched type field set;

[0110] Based on the generated set of element offset distance values, the spatial distance threshold corresponding to each type in the lithologic trend classification label is further applied. For example, the threshold for Class A is 50 units, for Class B it is 75 units, and for Class C it is 100 units. The offset distance is then compared against the threshold specified for each trend type. For example, if a number is labeled Class A and its offset distance is 45 units, it is considered a consistent type because it falls below the threshold of 50 units. This comparison can be performed in batches, with the offset values ​​compared against the threshold row by row. Those that meet the criteria are recorded as consistent, while those that do not are recorded as inconsistent. A consistency determination field is added as an identifier. For example, a new field, "Enrichment Consistency," indicates whether the distance criteria for the trend type are met. The field value can be "Consistent" or "Inconsistent." Finally, this field is merged with the original number information set to form a complete field set containing the number, trend type, offset distance, and consistency label. This enrichment type field set is constructed, achieving a consistent description of the enrichment characteristics of each number.

[0111] See also Figure 1 , the specific steps of S3 are:

[0112] S301: Calling the coding information and anomaly distribution data in the enrichment type field set, extracting the coordinate values ​​of the anomaly in the target area, calculating the distance from the point to the center position, summarizing the average distance and standard deviation of all points, and generating a point distance distribution parameter set;

[0113] When accessing the coded information and anomaly distribution data from the enrichment type field set, it's necessary to first extract coded fields such as geological type, structural unit, and tectonic attributes. For example, lithology code 001 represents granite, and 002 represents diorite. The three-dimensional spatial location of the anomaly is then extracted, typically consisting of x, y, and z coordinates. The coordinates of all anomaly points can be retrieved using geological modeling software or a spatial data interface. Next, the center of the target area is determined. For example, in the coordinate system, the center is set to x = 100, y = 200, and z = 300. The distance to each anomaly coordinate is calculated sequentially, using the root sum of squares of the coordinate differences to determine the spatial distance from each point to the center. For example, if the coordinates of a point are 103, 198, and 298, the distance from the center is approximately 4.1. The distance data for all anomaly points is recorded as a distance set, which is then statistically analyzed for mean and deviation. For example, if the distances are 3.5, 4.2, 4.0, 3.9, and 4.1, the average distance is 3.94 with a standard deviation of 0.25. All calculation results constitute a point distance distribution parameter set, including two core elements: mean distance and standard deviation, which are used for subsequent point screening and density analysis.

[0114] S302: Calculate the difference between the point and the center position based on the average distance and standard deviation in the point distance distribution parameter set, filter and select a concentrated point set, calculate the spatial density characteristics, and generate a coefficient set for the outlier dense area;

[0115] Based on the average distance and standard deviation generated in the previous stage, a screening interval is first set. For example, if the average distance is 3.94 and the standard deviation is 0.25, the range is 3.69 to 4.19. Outliers outside this range are filtered out, and only concentrated points are retained. Next, the target area is spatially divided using cubic grid units, for example, into cube grids with a side length of 5. The number of concentrated points contained in each grid is counted. For example, if a grid contains 10 points, the density is 10 points per cubic unit. After counting the number of points in all grids one by one, the area with the highest point density is found. If the maximum value is 20 points per cubic unit, the density values ​​of other areas are divided by 20 to obtain the dense area coefficient. For example, if the density of an area is 10, the coefficient is 0.5. This coefficient expresses the relative density of each area. Finally, the coefficients of all significantly dense areas are aggregated to form a coefficient set of dense outlier areas, which provides a basis for the next step of regional focus determination.

[0116] S303: Calling the coefficient set of the outlier dense area and the outlier number data, summarizing the number of outliers in the spatial area, performing classification judgment based on the outlier proportion and density conditions, and generating a focused classification feature set;

[0117] The specific calculation formula for the number of outliers in the summary spatial area is:

[0118]

[0119] Calculate the summary value of spatial outliers and generate a focused classification feature set;

[0120] Among them, N k Represents the total number of outliers in the k-th spatial region, w i Represents the coefficient of the ith outlier dense area, d i represents the number of abnormal points in the corresponding area, c i represents the spatial radius of the ith dense area, α represents the smoothing constant, β represents the cross-region influence weight factor, and p j represents the outlier impact factor of the jth adjacent area, γ represents the benchmark adjustment amount, j Represents the distance attenuation coefficient of the jth adjacent area;

[0121] Parameter assignment and data source:

[0122] w i (Outlier density area coefficient): The density of each area is calculated using a spatial clustering algorithm (such as DBSCAN), and the density value is normalized and mapped to the range of 0.1 and 0.9;

[0123] For example, if the density of region 1 is 0.8 and the density of region 2 is 0.6, the normalized values ​​are w1 = 0.57 and w2 = 0.43.

[0124] d i (Number of abnormal points): directly derived from real-time statistics of the monitoring system, the detection value of area 1 is 15, and the detection value of area 2 is 10;

[0125] c i (Spatial radius): Calculates the area boundaries based on GPS coordinate data using the minimum circumscribed circle algorithm. The radius for area 1 is 120 meters, and for area 2 is 90 meters.

[0126] α=1 (smoothing constant): set according to the principle that the denominator cannot be zero, and the empirical value range is 0.5, 2;

[0127] β = 0.2 (cross-area impact weight factor): This factor is set based on traffic flow monitoring data. When the average daily traffic volume in adjacent areas is 500 vehicles, the value is set to 0.2, and increases by 0.05 for every 100 vehicles that the traffic volume increases.

[0128] pj (Adjacent area impact factor): Quantified by the proportion of outliers. Area A has 20 outliers (total outliers are 50), and we get p A =20 / 50=0.4;

[0129] γ=1 (benchmark adjustment): maintains the same dimension as α to prevent the denominator from being too small;

[0130] j =0.5 (distance attenuation coefficient): calculated based on the distance between adjacent areas being 2 kilometers, the inverse of 1 / 2 = 0.5 is taken, and the coefficient decreases by 0.2 for every kilometer increase in distance;

[0131] Calculation derivation process:

[0132] First calculation:

[0133]

[0134] The sum is: 0.0712 + 0.0478 = 0.1190;

[0135] Second calculation:

[0136]

[0137] Final result:

[0138] N k =0.1190+0.0533=0.1723×100≈17.23;

[0139] The results show that the summary value of spatial area abnormal points is 17.23. After being magnified by the percentage system, the value matches the magnitude of the original monitoring data and is directly used as the input parameter of the focused classification feature set. The higher the value, the more significant the abnormal characteristics of the area.

[0140] See also Figure 1 , the specific steps of S4 are:

[0141] S401: Call the target area number corresponding to the focused classification feature set, extract the area boundary and attribute field information under the number, extract the target objects in the area based on the boundary, count the occurrence quantity and spatial area in each area, calculate the area ratio between the two data, and generate the occurrence value per unit area;

[0142] The target area number corresponding to the focused classification feature set usually comes from the screening result after the classification of the target object. For example, a set of spatial numbers, such as T001, T002, etc., are generated after a focused analysis of a certain type of mineral resources. The vector boundary information corresponding to the number can be retrieved through the spatial data platform, and the attribute fields of the boundary, such as land type name, administrative district, soil type, etc., are extracted at the same time. Then, in the spatial database, a spatial clipping operation is performed based on the above boundary to extract all target objects covered by the boundary, including point-distributed mineral points, surface-represented ore bodies, polygonal contaminated blocks, etc., and statistics are performed. The number of these objects within the boundary range is calculated and the sum of their geometric areas is calculated. The total area of ​​the region is directly calculated through the geometric properties of the boundary polygon. The total area of ​​the target object is obtained by accumulating the geometric area of ​​each object. For example, a number T001 has a range area of ​​100ha and contains 20 ore body targets with a total area of ​​5ha. The corresponding unit area occurrence is calculated as the total target area divided by the regional area, that is, 0.05. After processing all target area numbers one by one, the unit area occurrence values ​​corresponding to each number are summarized to form a set of correspondence tables between numbers and occurrence data, which is convenient for unified processing in subsequent operations.

[0143] S402: Based on the per-unit-area inventory values, the ratio data of all regions are summarized, and the grade interval boundaries are established by sorting. The data are divided into multiple grade segments according to the value range to generate the inventory density grade interval values;

[0144] Based on the unit area inventory values ​​extracted from all regions, a value set is formed and the values ​​in the set are sorted from small to large. Natural breakpoint segmentation, equidistant segmentation, or segmentation based on standard deviation can be selected to define the grade interval. If the natural breakpoint method is used, the built-in tools of the spatial analysis software can be used to automatically identify data aggregation trends and form boundaries. For example, in the case of a set of values ​​of 0.02, 0.05, 0.07, 0.08, 0.1, 0.12, 0.15, and 0.2, the natural breakpoints may be divided into five segments, namely less than or equal to 0.04, 0.04 to 0.08, 0.08 to 0.12, 0.12 to 0.16, and greater than 0.16. Each segment is assigned a grade number, such as L1 to L5. All grade interval boundaries and numbers should be recorded in a separate form for subsequent grade determination operations of numbered areas. This step ensures the comparability of the inventory density of each area and provides a unified coding standard.

[0145] S403: Based on the occurrence density level interval value, the occurrence value per unit area of ​​each region is called to determine the level interval to which it belongs, and a corresponding level code identifier is assigned. A corresponding set of numbers and regional spatial positions is constructed, and the code layer is superimposed to generate an occurrence density level label set;

[0146] Apply the set grade interval boundary values ​​to the unit area endowment values ​​of each numbered area, and determine the grade interval to which each value belongs one by one. If the unit area endowment of a number T001 is 0.05, it falls into the second grade interval and the corresponding code is L2. This code is corresponded to the numbered area, and the same judgment logic is performed on all areas to form a complete set of corresponding numbers and grade codes. The grade code identifier is added to the boundary layer of each area through spatial attribute connection to realize the integration of numbered area and grade information. The attribute connection tool can be used to perform code writing operations in the spatial data platform, and then the layer visualization style is set according to the code information, such as L1 is distinguished by light blue, L2 by green, and L3 by yellow. Finally, a set of spatial data layers with grade labels are generated to intuitively present the endowment grade of each area.

[0147] See also Figure 1 , the specific steps of S5 are:

[0148] S501: Call the high-level area number in the density level label set, match the spatial grid number and extract the thickness sequence of the corresponding area, calculate the adjacent thickness difference according to the adjacent grid number sequence, and determine the thickness change direction according to the difference sign to generate the adjacent thickness change trend value;

[0149] When calling the high-level area number in the density level label set, the area number with a level value of not less than 3 should be extracted from the area level data. For example, after extracting the area with level 4, the spatial grid database is searched for spatial units with the same area number, such as G101, G102, and G103. Then, the corresponding thickness values ​​are extracted from the thickness data table based on these numbers. For example, G101 is 6.4m, G102 is 7.1m, and G103 is 6.8m. After arranging them in order of numbers, the thickness difference between adjacent units is calculated. The thickness difference from G101 to G102 is positive, indicating an increase in thickness, and the thickness difference from G102 to G103 is negative, indicating a decrease in thickness. The direction of change can be identified by the positive and negative nature of the difference. The positive value can be set to the trend value 1, the negative value to -1, and the zero difference to 0. Therefore, the trend value sequence is 1 and -1, corresponding to the two groups of positions G101-G102 and G102-G103 respectively.

[0150] S502: Extracting the concentration value of key minerals based on the matching of the corresponding spatial positions according to the adjacent thickness change trend values, calculating the concentration reference value under the area, and comparing the ratio of the point concentration value to the reference value. Based on the trend change direction, select the positions where the concentration ratio is greater than the reference value to generate a concentration trend compliance position set;

[0151] According to the trend value, spatial positions such as G101-G102 and G102-G103 are located respectively, and the mineral concentration value recorded at the center coordinate of each pair of spatial units is obtained. Assuming that G101 is 1.8%, G102 is 2.3%, and G103 is 2.0%, the average benchmark value of the regional concentration is calculated using these three sets of values, which is 2.03%. The concentration value of each group of spatial units is converted to the ratio of this value. For example, the ratio of G102 is 2.3 divided by 2.03 to obtain approximately 1.13. The ratio is combined with the trend direction for screening. For the G101-G102 group, since the concentration ratio is greater than 1 and the trend direction is rising, it meets the screening requirements, so this position is recorded. For the G102-G103 group, the concentration ratio is not enough to be higher than the benchmark value, so this position is eliminated, and the final concentration trend-compliant position set is G101-G102.

[0152] S503: Extracting the occurrence density grade, thickness change trend value and mineral concentration ratio based on the concentration trend matching position set, constructing a feature combination and dividing the combination interval, filling the classification number into the spatial grid field record, and generating the alteration gradient classification field set;

[0153] The spatial data of G101 and G102 were extracted from the concentration trend-compliant locations. On this basis, the occurrence density level of the two grids was extracted to be 4, the thickness change trend value was 1, and the concentration ratio was 1.13. The above three indicators were combined to form a characteristic triplet represented by level 4, trend 1, and ratio 1.13. According to the pre-defined interval rules, if density levels 3 to 5 are high levels, trend value 1 represents rising, and concentration ratios higher than 1.1 are high concentration ratios, then this combination can be classified as the "high level-rising-high ratio" type, and the corresponding classification number is set to 7. The classification number is filled in the corresponding spatial unit record in the field of the spatial database to form a new field "alteration gradient classification". The value of this field in both G101 and G102 units is 7, completing the generation of the classification field set.

[0154] See also Figure 2 , a reserve exploration area evaluation system based on classification algorithm, including:

[0155] The structural classification module obtains spatial trend data of structural lines and lithologic zones, calculates the directional angle between structural joints and lithologic boundaries, uses spatial partitioning information to divide structural combinations by number, extracts the directional difference between each structural and lithologic combination, determines whether the fluctuation range of the directional difference sequence meets the classification criteria, incorporates the combination relationships that meet the criteria into the sample set, extracts the trend mean of the combination and binds it to a number to generate a structural trend classification label set;

[0156] The enrichment extraction module calls the number information in the construction trend classification label set, extracts the concentrations of multiple mineralized elements in the corresponding area, calculates the spatial distance between the element distribution center and the regional centroid, selects the areas that meet the distance threshold as the consistency type, and generates the enrichment type field set;

[0157] The focused recognition module calls the coded information in the enrichment type field set, extracts the coordinate data of the abnormal body in the corresponding area, calculates the average distance and standard deviation of the abnormal point set to the center position, selects the numbered areas with strong convergence and records the number of abnormal points, and generates a focused classification feature set;

[0158] The density assessment module calls the target area number in the focused classification feature set, extracts the regional occurrence quantity and spatial area, calculates the unit area ratio, and divides the levels in a segmented manner, assigns level identifiers, and generates a density level label set;

[0159] The alteration classification module calls the high-level number in the occurrence density level label set, extracts the grid thickness sequence of the corresponding area, calculates the adjacent change trend and determines the directional stability, extracts the concentration of key minerals and performs ratio analysis with the background value, screens the spatial locations that meet the trend and concentration difference characteristics, and generates the alteration gradient classification field set.

[0160] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for evaluating reserve exploration areas based on a classification algorithm, characterized in that: The following steps are involved: S1: Acquire structural line and lithologic zone trend data, extract the directional angle between structural joints and lithologic boundaries, divide structural trend combinations according to spatial position, extract the combination relationships within the matching range of directional differences, establish structural and lithologic trend classification samples, and generate a structural trend classification label set; S2: calling the spatial number in the structural trend classification label set, extracting the concentration data of multiple mineralized elements at corresponding positions, calculating the spatial distance between the element distribution center and the centroid of the geological area, extracting areas within the distance threshold and marking them as feature-consistent types, and generating an enrichment type field set; S3: Calling the coded information in the enrichment type field set, extracting the coordinate data of the abnormal body in the target area, calculating the average distance and distribution standard deviation of the point to the center position, identifying the distance convergence area, and establishing a spatial classification field based on the number of abnormal points to generate a focused classification feature set; S4: Call the target area number corresponding to the focused classification feature set, obtain the occurrence quantity and spatial area data in the area, calculate the unit area ratio, divide the results into multiple levels according to the natural segmentation method and encode and label them to generate an occurrence density level label set.

2. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 1, characterized in that: The structural trend classification label set includes structural strike combination type, direction matching interval, and lithologic trend category; the enrichment type field set includes element distribution center coordinates, centroid distance threshold, and feature consistent area identifier; the focused classification feature set includes average distance value of outliers, distribution standard deviation parameter, spatial convergence area identifier, and outlier quantity classification; the occurrence density level label set includes unit area ratio data, natural segmentation level division, and density level coding.

3. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 1, characterized in that: The specific steps of S1 are: S101: Acquire spatial coordinate information of the structural line and the lithologic zone boundary, extract the structural line direction and the lithologic zone strike based on the path direction relationship between the line segments, calculate the angle between the structural and lithologic boundaries by using the node direction difference, and obtain a structural and lithologic direction angle sequence; S102: clustering the structural lines according to the spatial distribution position in the structural and lithologic direction angle sequence, and screening them based on the angle value and the set interval to extract structural and lithologic combinations that meet the conditions, and obtain a matching trend combination comparison table; S103: extracting the direction difference set and combination identifier of structure and lithology based on the combination information in the matching trend combination comparison table, calculating the difference variation and establishing a classification map to generate a structural lithology trend classification label set.

4. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 3, characterized in that: The specific calculation formula for calculating the difference variability is: Calculate the difference variation index, establish a classification map, and generate a structural lithology trend classification label set; Among them, H represents the difference variability index of the structure and lithology direction difference set, represents the absolute difference of the eth element in the set of structural and lithological direction differences, w e is the normalized weight factor corresponding to the combination identifier, μ Δ is the weighted arithmetic mean of the direction difference set, k is the lithologic trend adjustment coefficient, λ g is the matching strength parameter of the g-th combination identifier, max(Δ d ) and min(Δ d ) represent the maximum and minimum values ​​of the direction difference set, is the unbiased variance of the directional difference, z is the total number of elements in the directional difference set, and y is the total number of combined identification categories.

5. The reserve exploration area evaluation method based on classification algorithm according to claim 3 is characterized in that: The specific steps of S2 are: S201: calling the spatial number in the structural lithologic trend classification label set, extracting the mineralized element concentration data at the corresponding position, associating the concentration value with the number according to the spatial coordinate, establishing a spatial distribution set of the mineralized elements, and generating an element distribution value matrix; S202: Based on the element distribution value matrix, extract the distribution center of the mineralized elements, calculate the spatial distance between the distribution center of the mineralized elements and the center of mass of the geological area, extract the offset distance data of all elements, and generate an element offset distance value set; S203: Based on the element offset distance value set, the spatial distance threshold in the trend classification label set is called to filter the numbered areas within the range, mark the corresponding numbers as feature consistent types, add a mark field, and generate an enriched type field set.

6. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 5, characterized in that: The specific steps of S3 are: S301: calling the coding information and anomaly distribution data in the enrichment type field set, extracting the coordinate values ​​of the anomalies in the target area, calculating the distance from the point to the center position, summarizing the average distance and standard deviation of all points, and generating a point distance distribution parameter set; S302: Calculate the difference between the point and the center position based on the average distance and standard deviation in the point distance distribution parameter set, perform screening, select a concentrated point set, calculate the spatial density characteristics, and generate a coefficient set for the outlier dense area; S303: Calling the outlier dense area coefficient set and outlier quantity data, summarizing the number of outliers in the spatial area, performing classification judgment based on the outlier proportion and density conditions, and generating a focused classification feature set.

7. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 6, characterized in that: The specific calculation formula for the number of outliers in the summary spatial area is: Calculate the summary value of spatial outliers and generate a focused classification feature set; Among them, N k Represents the total number of outliers in the k-th spatial region, w i Represents the coefficient of the ith outlier dense area, d i represents the number of abnormal points in the corresponding area, c i represents the spatial radius of the ith dense area, α represents the smoothing constant, β represents the cross-region influence weight factor, and p j represents the outlier impact factor of the jth adjacent area, γ represents the benchmark adjustment amount, ∈ j Represents the distance attenuation coefficient of the jth adjacent area.

8. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 6, characterized in that: The specific steps of S4 are: S401: Call the target area number corresponding to the focused classification feature set, extract the area boundary and attribute field information under the number, extract the target objects in the area based on the boundary, count the occurrence quantity and spatial area in each area, calculate the area ratio between the two data, and generate the occurrence value per unit area; S402: Based on the occurrence value per unit area, the ratio data of all regions are summarized, the grade interval boundaries are established by sorting, and the data are divided into multiple grade segments according to the value range to generate the occurrence density grade interval value; S403: According to the occurrence density level interval value, the unit area occurrence value of each area is called to determine the level interval to which it belongs, and a corresponding level coding identifier is assigned. A corresponding set of numbers and regional spatial positions is constructed, and the coding layer is superimposed to generate an occurrence density level label set.

9. The method for evaluating reserve exploration areas based on a classification algorithm according to claim 1, characterized in that: The method further comprises: S5: calling the high-level area number in the occurrence density level label set, extracting the grid thickness sequence and calculating the adjacent change trend, obtaining the concentration data of key minerals and calculating the ratio with the background concentration, extracting the spatial position that meets the change trend and concentration difference characteristics, and generating the alteration gradient classification field set; The alteration gradient classification field set includes grid thickness variation sequence, concentration difference ratio, and gradient trend matching area; The specific steps of S5 are: S501: calling the high-level area number in the density level label set, matching the spatial grid number and extracting the thickness sequence of the corresponding area, calculating the adjacent thickness difference according to the adjacent grid number sequence, and judging the thickness change direction according to the difference sign, and generating the adjacent thickness change trend value; S502: extracting the concentration value of key minerals from the corresponding spatial positions according to the adjacent thickness change trend values, calculating the concentration reference value of the area, and comparing the ratio of the point concentration value to the reference value. Based on the trend change direction, select the positions where the concentration ratio is greater than the reference value to generate a concentration trend compliance position set; S503: Extract the occurrence density grade, thickness change trend value and mineral concentration ratio according to the concentration trend compliance position set, construct a feature combination and divide the combination interval, fill the classification number into the spatial grid field record, and generate the alteration gradient classification field set.

10. A reserve exploration area evaluation system based on a classification algorithm, characterized in that: According to a method for evaluating a reserve exploration area based on a classification algorithm according to any one of claims 1 to 9, the system comprises: The structural classification module obtains spatial trend data of structural lines and lithologic zones, calculates the directional angle between structural joints and lithologic boundaries, uses spatial partitioning information to divide structural combinations by number, extracts the directional difference between each structural and lithologic combination, determines whether the fluctuation range of the directional difference sequence meets the classification criteria, incorporates the combination relationships that meet the criteria into the sample set, extracts the trend mean of the combination and binds it to a number to generate a structural trend classification label set; The enrichment extraction module calls the number information in the structural trend classification label set, extracts the concentrations of multiple mineralized elements in the corresponding area, calculates the spatial distance between the element distribution center and the regional centroid, selects the areas that meet the distance threshold as consistency types, and generates an enrichment type field set; The focus identification module calls the coded information in the enrichment type field set, extracts the coordinate data of the abnormal body in the corresponding area, calculates the average distance and standard deviation of the abnormal point set to the center position, selects the numbered areas with strong convergence and records the number of abnormal points, and generates a focus classification feature set; The density assessment module calls the target area number in the focused classification feature set, extracts the regional occurrence quantity and spatial area, calculates the unit area ratio, and divides the levels in a segmented manner, assigns level identifiers, and generates an occurrence density level label set; The alteration classification module calls the high-level number in the occurrence density level label set, extracts the grid thickness sequence of the corresponding area, calculates the adjacent change trend and determines the directional stability, extracts the concentration of key minerals and performs ratio analysis with the background value, screens the spatial locations that meet the trend and concentration difference characteristics, and generates the alteration gradient classification field set.

Images