Gold mine prediction method based on deep drilling data

Abstract

The invention discloses a gold mine prediction method based on deep ground drilling data, and relates to the field of mineral resource exploration. The problems that a traditional method depends on artificial experience, ignores a multi-source geological element space coupling relation and lacks three-dimensional probability expression, so that prediction precision is insufficient are solved. Carrying out weight distribution calculation on each key mineralization factor by adopting the coupling strength coefficient, and carrying out comprehensive processing on numerical values of the plurality of mineralization factors through a weighted fusion algorithm to obtain a comprehensive mineralization index distribution diagram layer; based on the numerical value distribution characteristics of the three-dimensional mineralization index field, establishing a mineralization probability prediction model by adopting a Bayesian network algorithm, and if the mineralization index of a certain spatial position is higher than a critical value, calculating a gold ore occurrence probability numerical value of the position; and converting the gold ore occurrence probability value into a probability distribution body under a space coordinate system through a three-dimensional modeling algorithm, and performing rendering processing on the probability distribution body by adopting a volume rendering technology to obtain a three-dimensional visual model of the gold ore occurrence probability.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mineral resources exploration, and particularly relates to a gold mine prediction method based on deep drilling data. BACKGROUND

[0002] With the gradual depletion of shallow gold resources, deep gold exploration has become a strategic focus of mineral resources development. Deep drilling, as a core means of obtaining underground geological information, can provide key data such as core samples, structural features, physical parameters and element content. Current gold mine prediction methods mainly rely on manual interpretation of geological maps and experience-based judgment of ore-forming regularities. These traditional methods have obvious shortcomings when dealing with deep drilling data: manual processing is inefficient and cannot fully exploit the deep ore-forming information contained in large volumes of high-dimensional drilling data; the prediction model fails to effectively integrate the spatial correlation of multiple geological elements such as lithological combination, structural intersection, geochemical anomaly and stratigraphic depth, resulting in limited prediction accuracy in complex geological environments; the experience-driven interpretation method is highly subjective and lacks a unified evaluation standard, making it difficult to meet the objectivity requirements of deep prospecting.

[0003] The core challenges of deep gold mine prediction are concentrated in the effective identification and quantitative modeling of ore-forming factors: due to the complex nonlinear relationship between lithological combination, structural intersection, geochemical anomaly and stratigraphic depth, traditional methods cannot accurately identify and quantify the contribution of key ore-forming factors; prediction models based on a single or a small number of factors cannot fully reflect the complexity of the ore-forming process, leading to deviations in the prediction results and actual ore body distribution; existing technologies lack effective means for spatial modeling and probabilistic prediction of multi-dimensional ore-forming factors, and cannot accurately quantify and display the gold ore occurrence probability in three-dimensional space. Therefore, the above problems are key technical issues that need to be solved. SUMMARY

[0004] To solve the above technical problems, the present application provides a gold mine prediction method based on deep drilling data to solve the problems existing in the prior art.

[0005] In the first aspect, to achieve the above object, the present application provides a gold mine prediction method based on deep drilling data, comprising the following steps:

[0006] Standardizing the deep drilling raw data of the gold mine to be tested to obtain a standardized drilling data set;

[0007] Determining key ore-forming factors from the standardized drilling data set through importance evaluation;

[0008] Calculating the spatial distance and direction angle between the key ore-forming factors, generating a spatial relationship matrix and determining a coupling strength coefficient;

[0009] generating a comprehensive metallogenic index distribution layer based on the coupling strength coefficient weighted fusion of multiple ore-forming factors;

[0010] generating a three-dimensional metallogenic index field by spatial interpolation on the comprehensive metallogenic index distribution layer;

[0011] establishing a metallogenic probability prediction model according to the three-dimensional metallogenic index field, and outputting a gold deposit occurrence probability value;

[0012] converting the gold deposit occurrence probability value into a probability distribution body and performing three-dimensional rendering to generate a visualization model;

[0013] extracting the high-probability metallogenic area boundary based on the probability contour surface distribution of the visualization model, and determining the target area position in combination with the original geological structure data.

[0014] In a second aspect, the present application also provides a gold deposit prediction system based on deep geodrilling data, which is used to implement a gold deposit prediction method based on deep geodrilling data, and the system comprises:

[0015] a data standardization processing module for standardizing the deep geodrilling original data of a gold deposit to be tested to obtain a standardized drilling data set;

[0016] a key ore-forming factor identification module for determining key ore-forming factors from the standardized drilling data set through importance evaluation;

[0017] a spatial coupling analysis module for calculating the spatial distance and direction angle between the key ore-forming factors, generating a spatial relationship matrix and determining a coupling strength coefficient;

[0018] an index fusion generation module for generating a comprehensive metallogenic index distribution layer based on the coupling strength coefficient weighted fusion of multiple ore-forming factors;

[0019] a three-dimensional interpolation modeling module for generating a three-dimensional metallogenic index field by spatial interpolation on the comprehensive metallogenic index distribution layer;

[0020] a probability prediction modeling module for establishing a metallogenic probability prediction model according to the three-dimensional metallogenic index field, and outputting a gold deposit occurrence probability value;

[0021] a visualization rendering module for converting the gold deposit occurrence probability value into a probability distribution body and performing three-dimensional rendering to generate a visualization model;

[0022] a target area positioning module for extracting the high-probability metallogenic area boundary based on the probability contour surface distribution of the visualization model, and determining the target area position in combination with the original geological structure data.

[0023] In a third aspect, the present application further provides a computer terminal device, comprising:

[0024] one or more processors;

[0025] a memory coupled to the processors, for storing one or more programs;

[0026] When the one or more programs are executed by the one or more processors, the one or more processors implement a gold mine prediction method based on deep drilling data.

[0027] In a fourth aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement a gold mine prediction method based on deep drilling data.

[0028] Compared with the prior art, the present application has the following advantages and technical effects:

[0029] The gold mine prediction method based on deep drilling data provided by the present application automatically screens key ore-forming factors in lithological combination, structural density, chemical anomaly intensity and burial depth through a random forest algorithm, quantifies the coupling strength coefficients between multiple factors based on spatial relationship analysis, generates a comprehensive ore-forming index distribution layer using a weighted fusion algorithm, constructs a three-dimensional ore-forming index field in combination with Kriging interpolation, establishes an ore-forming probability prediction model using a Bayesian network, converts the probability values into a three-dimensional visualization model, and finally determines a gold mine prediction target area through probability contour surface segmentation and geological structure superposition analysis. This method realizes spatial coupling modeling and three-dimensional probabilistic expression of multiple source geological elements, eliminates subjective bias of artificial experience, intuitively displays gold ore occurrence trends, and provides accurate decision-making basis for deep exploration. BRIEF DESCRIPTION OF DRAWINGS

[0030] The accompanying drawings, which form a part of the present application, are intended to provide further understanding of the present application, and the illustrative embodiments thereof and their descriptions serve to explain the present application, and do not constitute improper limitations on the present application. In the drawings:

[0031] Figure 1 It is a gold mine prediction method overall flowchart of the present application embodiment;

[0032] Figure 2 It is a key ore-forming factor identification process schematic diagram of the present application embodiment;

[0033] Figure 3 It is a comprehensive ore-forming index space generation process schematic diagram of the present application embodiment. DETAILED DESCRIPTION

[0034] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other in the case of no conflict. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0035] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that here.

[0036] Embodiment one

[0037] As Figure 1 shown, the present embodiment provides a gold deposit prediction method based on deep drilling data, comprising:

[0038] standardizing the deep drilling raw data of the gold deposit to be tested to obtain a standardized drilling data set;

[0039] determining key ore-forming factors from the standardized drilling data set through importance evaluation;

[0040] calculate the spatial distance and direction angle between the key ore-forming factors, generate a spatial relationship matrix and determine the coupling strength coefficient;

[0041] Weighted fusion of multiple ore-forming factors based on the coupling strength coefficient to generate a comprehensive ore-forming index distribution layer;

[0042] spatial interpolation of the comprehensive ore-forming index distribution layer to generate a three-dimensional ore-forming index field;

[0043] establishing an ore-forming probability prediction model according to the three-dimensional ore-forming index field and outputting a gold deposit occurrence probability value;

[0044] Convert the gold deposit occurrence probability value to a probability distribution body and perform three-dimensional rendering to generate a visual model;

[0045] Based on the probability contour surface distribution of the visual model, extract the boundary of the high probability ore-forming area, and determine the target area position in combination with the original geological structure data.

[0046] Specifically, a set of deep drilling raw data of a gold deposit to be tested is obtained, the deep drilling raw data is standardized and denoised by a data preprocessing module to obtain a standardized drilling data set;

[0047] The lithology combination, structure density, chemical anomaly intensity and burial depth in the standardized drilling data set are extracted and evaluated by using a random forest algorithm, and the geological elements with importance score exceeding a preset threshold are marked as key ore-forming factors;

[0048] The spatial distance and direction angle between each ore-forming factor are calculated by a spatial relationship analysis module to obtain a spatial relationship matrix of the ore-forming factors, and a coupling strength coefficient between the factors is determined based on the matrix according to the spatial distribution position of the key ore-forming factors;

[0049] The coupling strength coefficient is used for weight distribution calculation of each key ore-forming factor, and a comprehensive ore-forming index distribution layer is obtained by comprehensively processing the numerical values of multiple ore-forming factors through a weighted fusion algorithm.

[0050] According to the numerical variation gradient in the comprehensive ore-forming index distribution layer, a Kriging spatial interpolation algorithm is used to estimate the ore-forming index of unknown areas to obtain a continuous three-dimensional ore-forming index field.

[0051] Based on the numerical distribution characteristics of the three-dimensional ore-forming index field, a Bayesian network algorithm is used to establish a metallogenic probability prediction model to calculate the gold deposit occurrence probability value of the spatial position where the ore-forming index is higher than the critical value.

[0052] The gold deposit occurrence probability value is converted into a probability distribution body in a spatial coordinate system through a three-dimensional modeling algorithm, and the probability distribution body is rendered through volume rendering technology to obtain a three-dimensional visualization model of the gold deposit occurrence probability.

[0053] According to the probability contour surface distribution in the three-dimensional visualization model, a threshold segmentation algorithm is used to extract the spatial boundary of the high-probability ore-forming area, and the boundary information is superimposed and analyzed with the original geological structure data to determine the position of the gold deposit prediction target area.

[0054] As an embodiment in this embodiment, the process of the standardization processing includes:

[0055] The lithological combination information, the structural intersection position coordinates, the geochemical element content values and the stratum depth parameters in the deep drilling original data are obtained.

[0056] The lithological combination information, the element content values and the stratum depth parameters are dimensionally normalized, and the data exceeding the threshold range is adjusted to the standard interval.

[0057] The noise of the lithological combination information and the element content values is filtered through a wavelet transform denoising method, and the abnormal values deviating from the smooth curve are removed.

[0058] The structural intersection position coordinates and the stratum depth parameters are spatially grouped by using a classification algorithm, and the data is structured according to the grouping result.

[0059] Specifically, in step S101, the deep drilling raw data set of the gold mine to be tested is obtained, including lithology combination information, structural intersection position coordinates, geochemical element content values and stratum depth parameters. The deep drilling raw data is standardized and denoised by a data preprocessing module to obtain a standardized drilling data set of the gold mine to be tested.

[0060] The core sample raw data is obtained from the deep drilling equipment of the gold mine to be tested, and the raw data includes lithology combination information, structural intersection position coordinates, geochemical element content values and stratum depth parameters. The data acquisition module is used to convert the format of the raw data to obtain an initial drilling data set in a unified format. For the initial drilling data set, a standardization processing method is used to normalize the lithology combination information, element content values and stratum depth parameters. If the data value exceeds the preset threshold range, it is adjusted to the standard interval to obtain a standardized drilling data set. According to the standardized drilling data set, a wavelet transform denoising method is used to filter the noise of the lithology combination information and element content values. If it is detected that the data points deviate from the preset smooth curve, the abnormal values are removed to obtain a denoised drilling data set. The classification algorithm is used to spatially group the structural intersection position coordinates and stratum depth parameters in the denoised drilling data set. According to the grouping result, the data is structured to obtain a standardized drilling data set of the gold mine to be tested.

[0061] As an optional implementation in this embodiment, in the deep drilling of the gold mine to be tested, the raw data acquisition is the core link of the whole process. The core sample raw data usually includes lithology combination information, structural intersection position coordinates, geochemical element content values and stratum depth parameters. Assuming that the data obtained from a gold mine drilling point shows that the lithology combination is granite and quartz vein, the structural intersection position coordinates are (x: 120.5, y: 230.7, z: 1500.2), the geochemical element content includes gold (Au) 0.8 ppm, copper (Cu) 150 ppm, and the stratum depth is 1500 m. These data are converted into a unified format by a data acquisition module, such as converting heterogeneous data output by different devices into a CSV format, to ensure compatibility for subsequent processing. The initial drilling data set formed in this way is convenient for further analysis and reduces processing errors caused by inconsistent formats.

[0062] Specifically, the standardization processing normalizes the lithology combination information, element content values and stratum depth parameters to eliminate the influence of different dimensions.

[0063] As an optional implementation in this embodiment, the element content numerical range difference is large, and the gold (Au) content can be 0-10 ppm, and the copper (Cu) content can be 0-1000 ppm. The minimum-maximum normalization method is used to map the data to the [0, 1] interval. If the Au content of a certain data point is 12 ppm, which exceeds the preset threshold of 10 ppm, it is adjusted to the upper threshold of 10 ppm, and normalized to 1. This processing ensures that the data is within the standard interval, reduces the interference of outliers on subsequent analysis, and improves the stability of model training.

[0064] In an embodiment, the wavelet transform denoising method is used to process the noise in the standardized drilling data set.

[0065] As an optional implementation in this embodiment, for the fluctuation of element content values, the discrete wavelet transform can be used to decompose the data, and the low-frequency signal can be retained to filter high-frequency noise. Assuming that a certain Au content data sequence is [0.8, 0.9, 2.5, 0.7], and 2.5 deviates from the smooth curve and is judged as an outlier and is removed. The denoised data can better reflect the true mineralization characteristics, avoid noise interference in subsequent classification, and improve data reliability.

[0066] As an optional implementation in this embodiment, the classification algorithm performs spatial grouping on the structure intersection position coordinates and the stratum depth parameters in the denoised drilling data set. The K-means clustering algorithm is used to divide the data into several geological units according to the coordinates and depth.

[0067] As an optional implementation in this embodiment, the sample points with a depth of 1500-1600 m are grouped into one mineralization zone, and the points with coordinates (x: 120-130, y: 220-240, z: 1450-1550) are grouped into another group. After grouping, the structured processing is performed to form a standardized gold mine drilling data set to be tested. This grouping helps to identify the characteristics of different mineralization areas, and is convenient for subsequent ore body modeling and resource evaluation.

[0068] It can be understood that the above process significantly improves the data quality and analysis accuracy through format unification, data standardization, denoising and grouping.

[0069] As an optional implementation in this embodiment, the standardized data set provides reliable input for subsequent machine learning models, and denoising and grouping enhance the geological significance of the data, which helps to accurately locate the gold mineralization zone and improves the efficiency of exploration and the accuracy of resource evaluation.

[0070] As an implementation in this embodiment, the process of determining the key ore-forming factors includes:

[0071] extracting the distribution characteristics of lithological assemblage, structural density, chemical anomaly intensity and burial depth from the standardized drilling data set;

[0072] dividing the region by spatial analysis, marking the region where the structural density or chemical anomaly intensity exceeds the threshold value as a potential ore-forming region;

[0073] extracting the characteristics of lithological assemblage and burial depth in the potential ore-forming region, and classifying the characteristics that meet the ore-forming condition range as the key attention objects;

[0074] importance evaluation of chemical anomaly and structural density of key attention objects, marking the geological elements with weight value exceeding the threshold value as key ore-forming factors.

[0075] Specifically, in step S102, according to the distribution characteristics of geological elements in the standardized drilling data set of the to-be-tested gold mine, the random forest algorithm is used to extract and evaluate the characteristics of lithological assemblage, structural density, chemical anomaly intensity and burial depth, and if the importance score of a certain geological element exceeds the preset threshold value, it is marked as a key ore-forming factor.

[0076] As shown in Figure 2 , the geological element information is obtained from the standardized drilling data set of the to-be-tested gold mine, and the information is classified and arranged by using a data screening tool for lithological assemblage, structural density, chemical anomaly intensity and burial depth. According to the classified geological element data set, the data set is divided into regions by using a spatial analysis tool for distribution characteristics and spatial distribution. If the structural density or chemical anomaly intensity in a certain region exceeds the preset threshold value, it is marked as a potential ore-forming region, and the divided regional distribution data is obtained. Through the divided regional distribution data, the characteristics of lithological assemblage and burial depth in the potential ore-forming region are extracted by using a data comparison tool. If the extracted characteristic value meets the preset ore-forming condition range, it is classified as a key attention object, and the attention data set after feature extraction is obtained. The attention data set after feature extraction is obtained, and the ore-forming conditions of the data are evaluated for importance by using a weight calculation tool for chemical anomaly and structural density. If the weight value of a certain geological element exceeds the preset threshold value, it is marked as a key ore-forming factor, and the final key factor set is determined.

[0077] As an optional implementation in this embodiment, when obtaining geological element information from the standardized drilling data set of the gold mine to be tested, lithological combination, structural density, chemical anomaly intensity, and burial depth are the core analysis objects. Lithological combination reflects the main rock types in the mining area, such as granite, quartz vein, or gneiss, which are closely related to gold mineralization. Structural density refers to the number of faults or folds per unit area, usually expressed as the number of faults per square kilometer, such as 5 faults per square kilometer in a certain area, indicating strong tectonic activity. Chemical anomaly intensity is determined by the content of geochemical elements, such as gold content exceeding 2 ppm being considered high anomaly. Burial depth refers to the vertical distance of the ore body from the surface, such as 1500 meters deep, which may indicate deep mineralization. The data screening tool can use classification algorithms, such as decision trees, to group data by lithological type or anomaly intensity.

[0078] As an optional implementation in this embodiment, a data set contains 100 sample points, and after screening, samples with quartz vein lithology and gold content greater than 1 ppm are classified into one category to form a classified geological element data set. This classification facilitates subsequent spatial analysis.

[0079] In a possible implementation, for the classified geological element data set, the spatial analysis tool can use geographic information systems for regional division. Based on the distribution characteristics of structural density and chemical anomaly intensity, threshold values are set, such as regions with structural density greater than 3 faults per square kilometer or gold content greater than 1.5 ppm being marked as potential mineralization areas.

[0080] As an optional implementation in this embodiment, a certain mining area is divided into 10 regions, of which 3 regions are marked as potential mineralization areas due to structural density of 4 faults per square kilometer and average gold content of 2.2 ppm. This division clearly reflects the concentrated mineralization area, facilitating the focus on exploration resources.

[0081] As an optional implementation in this embodiment, for the lithological combination and burial depth in the potential mineralization area, the data comparison tool can identify key mineralization conditions through feature extraction. Assuming that the lithology in a certain area is quartz vein and the burial depth is 1400-1600 meters, by comparing the pre-set conditions such as quartz vein proportion greater than 60% and depth less than 2000 meters, the feature values that meet the conditions are extracted and classified as key objects of attention.

[0082] As an optional implementation in this embodiment, a sample point with 70% quartz vein lithology and 1450 meters depth meets the conditions and is included in the attention data set. This method ensures that the selected area has high mineralization potential.

[0083] In a possible implementation, for the chemical anomaly and structural density in the attention data set, the weight calculation tool can use the analytic hierarchy process to evaluate the importance of mineralization conditions.

[0084] As an optional implementation in this embodiment, the weight of gold content anomaly is set to 0.6, and the weight of tectonic density is 0.4. If the gold content of a certain area is 2.5ppm, and the tectonic density is 5 strips per square kilometer, the calculated comprehensive weight value is 0.85, which exceeds the threshold value 0.7, so the gold content and tectonic density are marked as key ore-forming factors. This evaluation method highlights the contribution of main geological factors to mineralization, which is convenient for optimizing exploration strategy.

[0085] It can be understood that the above process extracts key information from complex data through classification and arrangement, regional division, feature extraction and weight evaluation, layer by layer. Each step supports each other to form a complete analysis chain.

[0086] As an optional implementation in this embodiment, classification and arrangement provide the basis for regional division, and feature extraction and weight evaluation further focus on key factors. This logically rigorous analysis process helps to accurately identify gold ore-forming areas and improve exploration efficiency.

[0087] As an implementation in this embodiment, the process of generating a spatial relationship matrix includes:

[0088] Extracting the spatial distribution position coordinates of the key ore-forming factors;

[0089] Measuring the spatial distance and direction angle between each ore-forming factor by a spatial calculation tool;

[0090] Building an ore-forming factor spatial relationship matrix containing spatial distance and direction angle;

[0091] Quantitative processing of the factor correlation in the spatial relationship matrix, and marking the factor pairs with correlation values exceeding the threshold value as high coupling factor pairs.

[0092] Step S103, the key ore-forming factors are the data of the gold mine to be measured. For the spatial distribution position of the key ore-forming factors, the spatial distance and direction angle between each ore-forming factor are calculated by a spatial relationship analysis module to obtain an ore-forming factor spatial relationship matrix, and the coupling strength coefficient between the factors is determined based on the matrix.

[0093] The distribution position and position coordinates are obtained from the key ore-forming factor data of the to-be-tested gold mine. The spatial distance and direction angle between factors are measured by using a spatial calculation tool to obtain a preliminary spatial relationship data set. According to the preliminary spatial relationship data set, the spatial distance and direction angle are used to construct a data matrix by using a data integration tool to obtain a spatial relationship matrix of ore-forming factors. According to the spatial relationship matrix of ore-forming factors, the geological characteristics and ore-forming conditions are quantitatively processed by using a weight calculation tool according to the factor correlation and coupling strength. If the correlation value between certain factors exceeds a preset threshold, it is marked as a high coupling factor pair to obtain a preliminary coupling strength result. According to the preliminary coupling strength result, the data are classified and arranged by using a spatial division tool according to the distribution density and regional division of the high coupling factor pair to determine the final coupling strength coefficient and spatial relationship distribution map.

[0094] As an optional implementation in this embodiment, the distribution position and position coordinates are obtained for the key ore-forming factor data of the to-be-tested gold mine. The spatial calculation tool can extract coordinates and measure distances through a geographic information system. Assuming that a gold mine area contains 100 sample points, each point records the longitude and latitude coordinates of factors such as lithological assemblage, structural density, and chemical anomaly strength. The spatial calculation tool first extracts these coordinates and calculates the spatial distance between factors. For example, the distance between a quartz vein sample point A and a high gold content anomaly point B is 2.5 kilometers, and the direction angle is northeast 45 degrees. This measurement forms a preliminary spatial relationship data set containing distance and angle information, providing a basis for subsequent analysis.

[0095] It should be noted that the spatial distance and direction angle reflect the geometric relationship between ore-forming factors, which helps to reveal the potential rules of mineralization distribution.

[0096] In a possible implementation, based on the preliminary spatial relationship data set, the data integration tool can arrange data by a matrix construction method to form a spatial relationship matrix of ore-forming factors. Assuming that the data set contains 50 sample points involving two types of factors, lithological assemblage and chemical anomaly strength, the tool integrates the distance and angle between each pair of factors into a matrix form.

[0097] As an optional implementation in this embodiment, the distance between point A and point B is 2.5 kilometers, and the angle is 45 degrees. The distance between point A and point C is 3.2 kilometers, and the angle is 60 degrees. This matrix structure is clear and convenient for subsequent correlation analysis.

[0098] It should be noted that matrix construction ensures data standardization and facilitates quantitative analysis of the spatial coupling strength between factors.

[0099] As an optional implementation in this embodiment, for the ore-forming factor spatial relationship matrix, the weight calculation tool can use factor analysis method to evaluate the correlation and coupling strength between factors. Assuming that the matrix shows that the average distance between quartz veins and high gold content anomaly points is less than 3 kilometers, the direction angle is concentrated in the range of 30-60 degrees, and the tool calculates the correlation value as 0.82, which exceeds the preset threshold value 0.75, then it is marked as a high coupling factor pair. This quantitative processing highlights the interaction between key factors, which helps to identify the core conditions of mineralization.

[0100] It should be noted that weight calculation reflects the coupling relationship between factors by quantifying the correlation strength, which provides a basis for subsequent regional division.

[0101] Step S104, using the coupling strength coefficient to calculate the weight distribution of each key ore-forming factor, and using a weighted fusion algorithm to comprehensively process the numerical values of multiple ore-forming factors to obtain a comprehensive ore-forming index distribution layer.

[0102] As shown in Figure 3 , the spatial distribution position and coordinates of the key ore-forming factor data are obtained, the spatial distance and direction angle between each ore-forming factor are measured using a spatial calculation tool to obtain a spatial relationship data set. According to the spatial relationship data set, a data integration tool is used to construct an ore-forming factor spatial relationship matrix to judge the coupling strength between factors and obtain a coupling strength coefficient. For the coupling strength coefficient, a weight distribution tool is used to calculate the weight value of each ore-forming factor, and the numerical data is processed by a weighted fusion algorithm to obtain a weighted factor data set. According to the weighted factor data set, a spatial mapping tool is used to classify and arrange the data to generate a comprehensive ore-forming index distribution layer to obtain a distribution layer data.

[0103] As an optional implementation in this embodiment, for the spatial distribution position and coordinate acquisition of key ore-forming factor data, a geographic information system can be used as a spatial calculation tool to extract the latitude and longitude coordinates of sample points. Assuming that a certain gold mine area contains 200 sample points, the coordinates of factors such as magmatic rock distribution, fault zone density and gold element geochemical anomaly are recorded. The spatial calculation tool measures the spatial distance and direction angle between factors through coordinate extraction. For example, the distance between a magmatic rock sample point P and a high gold anomaly point Q is 1.8 kilometers, and the direction angle is southeast 30 degrees. These data constitute a spatial relationship data set, which contains distance and angle information of all sample point pairs, providing a basis for subsequent matrix construction.

[0104] It should be noted that the spatial distance and angle data reflect the geometric distribution characteristics of ore-forming factors, which helps to reveal the mineralization regularity.

[0105] In a possible implementation, the data integration tool constructs a spatial relationship matrix of ore-forming factors based on the spatial relationship dataset. Assuming that the dataset involves 150 sample points and contains two types of factors, i.e., fault zone density and gold element anomaly, the tool arranges the distance and angle between each pair of factors into a matrix form. For example, the matrix records that the distance between point P and point Q is 1.8 kilometers and the angle is 30 degrees, and the distance between point P and point R is 2.7 kilometers and the angle is 45 degrees. This matrix structure facilitates the analysis of the spatial correlation between factors.

[0106] It should be noted that the matrix normalization processing ensures data consistency and provides a reliable basis for the coupling strength judgment.

[0107] As an optional implementation in this embodiment, for the spatial relationship matrix, the weight allocation tool judges the coupling strength between factors by statistical analysis method and generates a coupling strength coefficient. Assuming that the matrix shows that the average distance between the fault zone density and the gold element anomaly is less than 2 kilometers and the angle is concentrated in 20-50 degrees, the tool calculates the coupling coefficient as 0.78, which exceeds the threshold value 0.70 and is marked as a high-coupling factor pair. This quantitative analysis clearly reflects the correlation strength between factors.

[0108] It should be noted that the coupling coefficient reflects the spatial synergy between factors and provides a basis for weight allocation.

[0109] In a possible implementation, based on the coupling strength coefficient, the weight allocation tool calculates the weight value of each factor by using the entropy method. Assuming that the weights of magmatic rock distribution, fault zone density, and gold element anomaly are 0.25, 0.35, and 0.40 respectively, the factor data is integrated into a weighted factor dataset through a weighted fusion algorithm. This weighted processing highlights the contribution of key factors.

[0110] It should be noted that the weighted fusion improves the comprehensiveness of the data and helps to accurately assess the ore-forming potential.

[0111] As an optional implementation in this embodiment, for the weighted factor dataset, the spatial mapping tool classifies and arranges the data through cluster analysis and generates a comprehensive ore-forming index distribution layer. Assuming that the mining area is divided into 10 sub-regions, and 3 of them are marked as high-potential areas due to high weighted index (average 0.85). The distribution layer shows that these areas are concentrated in the southern part of the mining area, and the area of the high-value index area accounts for 15% of the total area. This visualization result intuitively reflects the distribution of ore-forming potential and facilitates exploration planning.

[0112] It should be noted that the distribution layer data converts complex information into spatial expression and supports accurate decision-making.

[0113] As an implementation in this embodiment, the process of generating a three-dimensional ore-forming index field includes:

[0114] analyzing the numerical gradient of the known regions in the integrated metallogenic index distribution layer;

[0115] estimating the metallogenic index of the unknown regions based on the numerical gradient through a spatial interpolation algorithm;

[0116] associating and mapping the spatial coordinates with the interpolated metallogenic index to generate a continuous three-dimensional metallogenic index field;

[0117] optimizing the regions in the three-dimensional metallogenic index field that are lower than the threshold value by fusing the geological data and the interpolation results.

[0118] Specifically, in step S105, the metallogenic index of the unknown regions is estimated according to the numerical variation gradient in the integrated metallogenic index distribution layer, and a continuous three-dimensional metallogenic index field is obtained through interpolation calculation, which is aimed at gold mines.

[0119] The geological data and the spatial coordinates are obtained from the integrated metallogenic index distribution layer, the numerical gradient of the known regions is processed using a grid division tool to obtain the numerical gradient dataset. The metallogenic index of the unknown regions is calculated using a spatial interpolation tool for the numerical gradient dataset to obtain the interpolated metallogenic index dataset. According to the interpolated metallogenic index dataset, the spatial coordinates and the metallogenic index are associated and mapped using a three-dimensional modeling tool to generate the continuous three-dimensional metallogenic index field. If the index values of some regions in the three-dimensional metallogenic index field are lower than the preset threshold value, the geological data and the interpolation results are integrated and processed using a data fusion tool to obtain the optimized three-dimensional metallogenic index field.

[0120] As an optional implementation in this embodiment, the process of obtaining geological data and spatial coordinates from the integrated metallogenic index distribution layer can be realized through the spatial management of geographic information system tools. Assuming that a gold mine area contains 300 sampling points, records information such as mineralization alteration zone, gold element geochemical anomaly, and stratigraphic contact zone, and each sampling point is associated with latitude and longitude coordinates. The geographic information system tool extracts these coordinates to form a dataset containing the positions of the sampling points and geological properties.

[0121] As an optional implementation in this embodiment, a sampling point A records a high gold anomaly value with coordinates of 108.5 degrees east longitude and 34.2 degrees north latitude, and another sampling point B records the characteristics of the alteration zone with coordinates of 108.7 degrees east longitude and 34.3 degrees north latitude. These data provide a basis for subsequent grid division.

[0122] It can be understood that accurate coordinate extraction ensures the spatial consistency of geological data, laying a foundation for gradient analysis.

[0123] In a possible implementation, the grid division tool processes the numerical gradients of the known area to generate a numerical gradient dataset. Assuming that the gold mine area is divided into 1000 grid cells of 50m x 50m, the tool calculates the gradient change of each grid according to the ore-forming index of the sampling points.

[0124] As an optional implementation in this embodiment, the ore-forming index of grid G1 is 0.65, and the index of the adjacent grid G2 is 0.72. The tool calculates the gradient between the two grids as 0.07 / 50m. This gradient data reflects the spatial variation trend of the ore-forming index.

[0125] It can be understood that the grid division refines the spatial resolution of the data, which helps to capture the local variation characteristics.

[0126] As an optional implementation in this embodiment, for the numerical gradient dataset, the spatial interpolation tool can use the Kriging interpolation method to estimate the ore-forming index of the unknown area. Assuming that there is an unsampled area in the north of the mine area, the tool calculates the ore-forming index of the unsampled point C as 0.68, with coordinates of 108.6 degrees east and 34.4 degrees north, according to the gradient data of the known grids. The interpolation result forms an ore-forming index dataset covering the entire mine area.

[0127] It can be understood that the interpolation method fills in the data gaps through mathematical models, improving the data integrity.

[0128] In a possible implementation, the three-dimensional modeling tool associates the interpolated ore-forming index with spatial coordinates to generate a continuous three-dimensional ore-forming index field. Assuming that the depth range of the mine area is 0 to 500m, the tool constructs a three-dimensional grid model according to the sampling points and interpolation data.

[0129] As an optional implementation in this embodiment, a certain grid cell shows an ore-forming index of 0.75 at a depth of 200m, indicating that this area has a higher mineralization potential.

[0130] It can be understood that the three-dimensional model intuitively presents the three-dimensional distribution of the ore-forming index, facilitating the analysis of deep mineralization characteristics.

[0131] As an optional implementation in this embodiment, if the index of certain areas in the three-dimensional ore-forming index field is lower than the preset threshold value of 0.60, the data fusion tool can integrate geological data and interpolation results for optimization. Assuming that the index of a certain area is 0.55, the fusion tool adjusts the index of this area to 0.62 by combining the distribution of alteration zones and gold anomaly data in the adjacent areas. The optimized three-dimensional ore-forming index field more accurately reflects the distribution of mineralization potential.

[0132] It can be understood that data fusion enhances the reliability of the model by integrating multi-source information, providing a more accurate basis for exploration decisions.

[0133] As an embodiment in the present embodiment, the process of establishing the metallogenic probability prediction model comprises:

[0134] extracting numerical distribution data and spatial position coordinates from the three-dimensional metallogenic index field;

[0135] extracting structural features from the numerical distribution data by a grid division tool;

[0136] training a prediction model of the probability relationship between the metallogenic index and gold deposit by a Bayesian network algorithm;

[0137] outputting a gold deposit probability numerical value by the prediction model when the metallogenic index is higher than a critical value.

[0138] Specifically, in step S106, a metallogenic probability prediction model is established based on the numerical distribution characteristics of the three-dimensional metallogenic index field by a Bayesian network algorithm. If the metallogenic index of a certain spatial position is higher than a critical value, the gold deposit probability numerical value of the position is calculated.

[0139] According to the three-dimensional metallogenic index field, numerical distribution data and spatial position coordinates are obtained. Data features are extracted by a grid division tool to obtain a structured data feature set. For the data feature set, a Bayesian network algorithm is used to train the probability relationship between the metallogenic index and gold deposit to obtain a pre-established probability prediction model. If the metallogenic index of a certain spatial position is higher than a preset critical value, the gold deposit probability of the position is calculated by the probability prediction model to obtain a probability numerical value data set. A three-dimensional mapping tool is used to associate the spatial position and probability numerical value in the probability numerical value data set to generate a three-dimensional distribution field of gold deposit probability.

[0140] As an optional embodiment in the present embodiment, the process of obtaining numerical distribution data and spatial position coordinates from the three-dimensional metallogenic index field can be realized by a data extraction tool. Assuming that the three-dimensional metallogenic index field of a certain gold mine area has been constructed, the data extraction tool reads the metallogenic index and corresponding spatial coordinates of each grid cell from the model.

[0141] For example, a certain grid cell is located at 108.6 degrees east longitude, 34.3 degrees north latitude, and a depth of 300 meters, with a metallogenic index of 0.72. Another grid cell is located at 108.5 degrees east longitude, 34.2 degrees north latitude, and a depth of 250 meters, with a metallogenic index of 0.68. The data extraction tool integrates these numerical values and coordinates into a structured data set, containing attribute information of each grid cell.

[0142] It can be understood that such data organization method facilitates subsequent feature extraction and analysis.

[0143] In one possible implementation, the grid division tool processes the extracted numerical distribution data and spatial position coordinates to generate a structured data feature set.

[0144] Specifically, the tool can divide the gold mine area into a plurality of 50m x 50m x 50m cubic grid units. For each grid unit, the tool analyzes the difference between its metallogenic index and that of adjacent units, extracts features such as index change rate and spatial correlation.

[0145] As an optional implementation in this embodiment, if the metallogenic index of a grid unit is 0.70 and the index of adjacent units is 0.75, the tool can identify the feature of an upward trend in this area. In another example, if the metallogenic index of a grid unit is significantly higher than that of surrounding units, the tool can mark it as a feature point of a high-potential area.

[0146] It can be understood that this feature extraction method helps to reveal the spatial distribution of the metallogenic index.

[0147] As an optional implementation in this embodiment, the process of training the relationship between the metallogenic index and the gold occurrence probability using the Bayesian network algorithm for the data feature set can be implemented by a probability modeling tool.

[0148] Specifically, the Bayesian network algorithm uses the metallogenic index and spatial position information in the data feature set to analyze its correlation with gold occurrence.

[0149] In one possible implementation, the tool first collects feature data of known gold occurrence areas, assuming that the metallogenic index of a certain area is 0.78 and that gold has been confirmed to exist historically. The tool learns from these data to build a probability model reflecting the relationship between the metallogenic index and the possibility of gold occurrence.

[0150] As an optional implementation in this embodiment, a grid unit with a metallogenic index of 0.80 can be associated by the model to a higher probability of gold occurrence.

[0151] It can be understood that this method improves the relevance of prediction by combining historical data and features.

[0152] In one possible implementation, if the metallogenic index of a spatial position is higher than a preset threshold, the probability prediction model can calculate the probability of gold occurrence at this position.

[0153] Specifically, assuming that the preset threshold is 0.70, a grid cell is located at 108.7 degrees east longitude, 34.4 degrees north latitude, and a depth of 200 meters, and the metallogenic index is 0.76. According to the training results, the probability prediction model calculates the gold deposit occurrence probability of the grid cell to be 0.85 based on the feature data of the grid cell. In another example, the metallogenic index of a grid cell is 0.72, and the model calculates the probability to be 0.78. The tool integrates these probability values with the corresponding spatial positions to form a probability numerical data set.

[0154] It can be understood that this probability quantification method provides a data basis for subsequent distribution analysis.

[0155] As an optional implementation in this embodiment, the process of associating the spatial positions in the probability numerical data set with the probability numerical values by using the three-dimensional mapping tool to generate a three-dimensional distribution field of the gold deposit occurrence probability can be realized by a visualization tool.

[0156] Specifically, the tool associates each probability value in the probability numerical data set with the corresponding spatial coordinates to construct a three-dimensional model.

[0157] Illustratively, a grid cell is located at 108.6 degrees east longitude, 34.3 degrees north latitude, and a depth of 300 meters, and the gold deposit occurrence probability of the grid cell is 0.82. The tool marks the point in the three-dimensional space. Another grid cell is located at 108.5 degrees east longitude, 34.2 degrees north latitude, and a depth of 350 meters, and the probability is 0.79. The tool also maps it to the model.

[0158] It can be understood that this three-dimensional distribution field intuitively reflects the spatial variation of the gold deposit occurrence probability and facilitates the identification of high-potential areas.

[0159] In a possible implementation, the three-dimensional distribution field of the gold deposit occurrence probability can be further analyzed for its spatial characteristics.

[0160] As an optional implementation in this embodiment, the tool can identify that the areas with high probability values are concentrated in the southern part of the mining area, in the range of 200 meters to 300 meters in depth, and the probability values are generally above 0.80. In another example, the probability values in the northern part of the mining area are low, generally below 0.60, indicating that the gold deposit occurrence probability in this area is small.

[0161] It can be understood that this analysis method helps to focus on key areas and provides targeted basis for exploration.

[0162] In step S107, the gold deposit occurrence probability numerical values are converted into a probability distribution body in a spatial coordinate system by a three-dimensional modeling algorithm, the probability distribution body is rendered by a volume rendering technique, and a three-dimensional visualization model of the gold deposit occurrence probability is obtained.

[0163] According to the association data of the probability value and the spatial coordinates, a coordinate conversion tool is used to perform a spatial mapping operation to generate an initial spatial distribution body structure, and determine a spatial distribution form of the probability data. For the initial spatial distribution body structure, a volume rendering method tool is used to perform surface texture enhancement and detail filling processing to generate a probability mapping body with a stereoscopic level, and obtain a stereoscopic expression of the probability distribution. If the area data of the probability mapping body is lower than a preset threshold value, a distribution rendering tool is used to perform color layering processing on a low probability area to generate a distribution body type rendering result with visual differentiation, and determine the probability distribution characteristics of the rendering area. Through the distribution body type rendering result, a three-dimensional visualization graph generation tool is used to integrate the spatial distribution characteristics of the gold ore probability to generate a three-dimensional visualization graph that meets the spatial coordinate and probability mapping requirements, and determine the spatial distribution effect of the gold ore probability.

[0164] For example, for the association data of the probability value and the spatial coordinates, a coordinate conversion tool is used to perform a spatial mapping operation to generate an initial spatial distribution body structure, which can be realized by a spatial data processing tool.

[0165] As an optional implementation in this embodiment, the probability data of a certain gold mine area includes spatial coordinates and corresponding gold ore occurrence probability. A coordinate conversion tool maps these data to a three-dimensional space to form a preliminary distribution body. Assuming that a point is located at 108.8 degrees east longitude, 34.5 degrees north latitude, and a depth of 400 meters, and the probability value is 0.75, the tool will locate it in a three-dimensional grid to generate a spatial framework containing all probability points.

[0166] It can be understood that this mapping method ensures the accurate correspondence between the probability data and the spatial position, and lays a foundation for subsequent processing.

[0167] In a possible implementation, for the initial spatial distribution body structure, a volume rendering method tool is used to perform surface texture enhancement and detail filling to generate a probability mapping body.

[0168] As an optional implementation in this embodiment, the volume rendering tool adds smooth texture to the high probability area by analyzing the gradient change of the probability value, and fills the low probability area with rough texture. Assuming that the probability value of a certain area gradually changes from 0.70 to 0.80, the tool draws a surface with smooth transition in the area to enhance the stereoscopic effect. This processing method intuitively reflects the spatial change trend of the probability, and is convenient for identifying the key area. For the area in the probability mapping body that is lower than the preset threshold value, a distribution rendering tool is used to perform color layering processing to generate a rendering result with visual differentiation.

[0169] As an optional implementation in the embodiment, assuming that the threshold value is 0.65, the probability value of a certain area is 0.60, the tool assigns a blue tone to it, and the area with a probability value of 0.80 is assigned a red tone. After rendering, the low-probability area and the high-probability area are clearly distinguished by color difference.

[0170] It can be understood that such color layering facilitates quick judgment of probability distribution characteristics and highlights high-potential areas.

[0171] In a possible implementation, the tool integrates the spatial distribution characteristics of gold ore probability by using a three-dimensional visualization diagram generation tool to generate a three-dimensional visualization diagram.

[0172] As an optional implementation in the embodiment, the tool combines data points in the probability map with spatial coordinates to generate an interactive three-dimensional model. Assuming that a high-probability area is concentrated at 108.7 degrees east longitude, 34.4 degrees north latitude, and a depth of 250 meters, with a probability value of 0.85, the area is displayed in bright red in the model, and the northern low-probability area is displayed in blue. Such a visualization diagram intuitively shows the spatial distribution law of probability, facilitating analysts to focus on high-potential areas.

[0173] It can be understood that this approach improves data interpretation efficiency and provides a basis for exploration decisions.

[0174] As an optional implementation in the embodiment, for the implementation of color layering processing, the tool can further analyze the spatial connectivity of the low-probability area to determine its distribution characteristics. Assuming that the probability value of the northern low-probability area is between 0.50 and 0.60 and presents a continuous distribution, the tool marks it as a low-potential area. Such analysis helps to exclude unnecessary exploration areas and optimize resource allocation.

[0175] It can be understood that through multi-level rendering and visualization, technical personnel can more efficiently identify potential areas of gold ore occurrence.

[0176] Step S108, according to the probability contour surface distribution in the three-dimensional visualization model, the spatial boundary of the high-probability ore-forming area is extracted by a threshold segmentation algorithm, and the boundary information is superimposed and analyzed with the original geological structure data to determine the location of the final gold ore prediction target area.

[0177] The probability contour surface data is obtained from the three-dimensional visualization model, a threshold segmentation tool is used to filter the areas with probability values higher than a preset threshold, a preliminary spatial range of the high-probability ore-forming area is generated, and the distribution characteristics of the high-probability area are determined. For the preliminary spatial range, an edge detection tool is used to perform edge detection processing on the high-probability area to generate spatial boundary data of the high-probability area, and an accurate expression of the spatial boundary is obtained. The original geological structure data is obtained, a superposition analysis tool is used to perform spatial registration and fusion of the spatial boundary data and the geological structure data, boundary integration data containing geological features are generated, and the geological correlation of the boundary area is determined. If the boundary integration data meets a preset ore-forming probability condition, a data integration tool is used to map the boundary integration data and the ore-forming probability distribution, spatial position data of a gold mine prediction target area are generated, and the position of the prediction target area is determined.

[0178] As an optional implementation in this embodiment, when obtaining probability contour surface data from a three-dimensional visualization model, a special analysis tool can be used to extract spatial distribution information of a specific probability value. Assuming that in a gold mine exploration project, the preset probability threshold is 0.70, the tool will filter out all area data with a probability value higher than 0.70 to form a preliminary high-probability ore-forming area range. Such an area may be concentrated in a certain spatial depth, such as a range of 300 meters to 500 meters in depth, and the distribution characteristics show an irregular ellipsoid. This screening method helps to quickly lock the potential ore-forming area and provides a data basis for subsequent analysis.

[0179] As an optional implementation in this embodiment, for the boundary extraction of the preliminary spatial range, an edge detection tool can be used to accurately identify the edge of the high-probability area. Assuming that in the above-mentioned area, the tool identifies a boundary line where the probability value drops from 0.70 to 0.50 by analyzing the spatial variation of the probability value, forming a closed spatial boundary data. This boundary may exhibit characteristics of being relatively smooth in the south and relatively tortuous in the north, reflecting the complexity of the underground structure. Such boundary data provides an accurate spatial reference for subsequent geological correlation analysis.

[0180] As an optional implementation in this embodiment, when obtaining original geological structure data and performing superposition analysis, the spatial boundary data can be spatially registered with the geological structure information. Assuming that the geological data of a certain gold mine area shows that there is a fault zone in the boundary area, the tool performs fusion analysis of the fault zone and the high-probability area boundary, and finds that the boundary in the south highly coincides with the fault zone, indicating that this area may be controlled by the fault to form ore. The boundary integration data generated by this fusion can reveal the potential relationship between geological features and ore-forming probability, providing an important basis for further judging ore-forming conditions.

[0181] As an optional implementation in this embodiment, after judging whether the boundary integrated data meets the metallogenic probability condition, if the condition is met, the spatial position data of the gold mine prediction target area can be generated by the data integration tool. Assuming that the analysis result shows that the area with a probability value higher than 0.75 in the boundary integrated data has a coincidence degree of more than 80% with the fault zone, the tool maps this area as the prediction target area, and the specific position can be located at east longitude 108.6 degrees, north latitude 34.3 degrees and depth 350 meters. This target area position data provides a clear guidance direction for subsequent exploration work, and helps to optimize the allocation and planning of exploration resources.

[0182] It can be understood that each link above forms a complete analysis chain from probability distribution to prediction target area through step-by-step screening, extraction and fusion of data. The processing of each link is closely related to the spatial characteristics of gold metallogenic probability, ensuring the pertinence and reliability of the analysis. This method can quickly locate the potential metallogenic area in a complex geological environment and provide scientific support for exploration decision-making.

[0183] Based on this, the embodiment of the present application provides a gold mine prediction method based on deep drilling data. The present application automatically screens key metallogenic factors in lithological combination, structural density, chemical anomaly intensity and burial depth by using a random forest algorithm, quantifies the coupling strength coefficient between multiple factors based on spatial relationship analysis, generates a comprehensive metallogenic index distribution layer by using a weighted fusion algorithm, constructs a three-dimensional metallogenic index field by combining Kriging interpolation, establishes a metallogenic probability prediction model by using a Bayesian network, converts the probability value into a three-dimensional visualization model, and finally determines the gold mine prediction target area by probability contour surface segmentation and geological structure superposition analysis. This method realizes spatial coupling modeling and three-dimensional probabilistic expression of multiple source geological elements, eliminates subjective bias of artificial experience, intuitively displays the gold ore occurrence trend, and provides accurate decision-making basis for deep exploration.

[0184] Embodiment two

[0185] In this embodiment, a computer terminal device is provided, comprising:

[0186] one or more processors;

[0187] a memory coupled to the processor, for storing one or more programs;

[0188] When the one or more programs are executed by the one or more processors, the one or more processors implement the method in the above embodiments.

[0189] In this embodiment, a computer readable storage medium having a computer program stored thereon is also provided, and the computer program is executed by a processor to implement the method in the above embodiments.

[0190] In the embodiment, an electronic device is also provided, comprising a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute the method in the above embodiment.

[0191] The above program can be run in the processor, or can also be stored in the memory (or called computer readable medium), the computer readable medium includes permanent and non-permanent, removable and non-removable media, which can realize information storage by any method or technology. Information can be computer readable instructions, data structure, program module or other data. Examples of computer storage medium include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape disk storage or other magnetic storage device or any other non-transmission medium that can be used to store information accessible by a computing device.

[0192] These computer programs can also be loaded into a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce computer implemented processing, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flow Figure 1 One flow or multiple flows and / or blocks Figure 1 One block or multiple blocks, and different steps can be realized by different modules.

[0193] Such a device or system is provided in the embodiment. The system is called gold mine prediction system based on deep drilling data, which comprises:

[0194] The data standardization processing module is configured to perform standardization processing on the deep drilling original data of the to-be-tested gold mine to obtain a standardized drilling data set.

[0195] The key ore-forming factor identification module is configured to determine key ore-forming factors from the standardized drilling data set through importance evaluation.

[0196] The spatial coupling analysis module is configured to calculate spatial distances and direction angles between the key ore-forming factors, generate a spatial relationship matrix and determine a coupling strength coefficient.

[0197] The index fusion generation module is configured to fuse multiple ore-forming factors based on the coupling strength coefficient to generate a comprehensive ore-forming index distribution layer.

[0198] a three-dimensional interpolation modeling module configured to perform spatial interpolation on the comprehensive metallogenic index distribution layer to generate a three-dimensional metallogenic index field;

[0199] a probability prediction modeling module configured to establish a metallogenic probability prediction model according to the three-dimensional metallogenic index field and output a gold deposit occurrence probability value;

[0200] a visualization rendering module configured to convert the gold deposit occurrence probability value into a probability distribution body and perform three-dimensional rendering to generate a visualization model;

[0201] a target area positioning module configured to extract a high-probability metallogenic region boundary based on a probability contour surface distribution of the visualization model and determine a target area position in combination with original geological structure data.

[0202] As an implementation manner in the embodiment, the data standardization processing module comprises:

[0203] an original data acquisition unit configured to acquire lithological combination information, structural intersection position coordinates, geochemical element content values and stratum depth parameters in deep drilling original data;

[0204] a dimension normalization unit configured to perform dimension normalization on the lithological combination information, the element content values and the stratum depth parameters and adjust data exceeding a threshold range to a standard interval;

[0205] a noise filtering unit configured to filter noise of the lithological combination information and the element content values by a wavelet transform denoising method and eliminate abnormal values deviating from a smooth curve;

[0206] a spatial grouping unit configured to perform spatial grouping on the structural intersection position coordinates and the stratum depth parameters by using a classification algorithm and structure data according to a grouping result.

[0207] As an implementation manner in the embodiment, the key metallogenic factor identification module comprises:

[0208] a feature extraction unit configured to extract distribution features of lithological combination, structural density, chemical anomaly intensity and burial depth from the standardized drilling data set;

[0209] a region marking unit configured to mark a region in which the structural density or the chemical anomaly intensity exceeds a threshold value as a potential metallogenic region by spatial analysis;

[0210] a condition screening unit configured to extract features of the lithological combination and the burial depth in the potential metallogenic region and classify features meeting a metallogenic condition range as key objects of attention;

[0211] A weight evaluation unit is configured to evaluate the importance of the chemical anomaly and the tectonic density of the object of interest, and mark geological elements with a weight value exceeding a threshold value as key ore-forming factors.

[0212] As an embodiment of the present embodiment, the spatial coupling analysis module comprises:

[0213] A coordinate extraction unit is configured to extract the spatial distribution position coordinates of the key ore-forming factors.

[0214] A spatial measurement unit is configured to measure the spatial distance and direction angle between the ore-forming factors by using a spatial calculation tool.

[0215] A matrix construction unit is configured to construct an ore-forming factor spatial relationship matrix comprising the spatial distance and direction angle.

[0216] A correlation quantization unit is configured to quantize the factor correlation in the spatial relationship matrix, and mark factor pairs with a correlation value exceeding a threshold value as high coupling factor pairs.

[0217] As an embodiment of the present embodiment, the three-dimensional interpolation modeling module comprises:

[0218] A gradient analysis unit is configured to analyze the numerical gradient of the known region in the comprehensive ore-forming index distribution layer.

[0219] A spatial estimation unit is configured to estimate the ore-forming index of the unknown region by using a spatial interpolation algorithm based on the numerical gradient.

[0220] A three-dimensional mapping unit is configured to associate and map the spatial coordinates and the interpolated ore-forming index to generate a continuous three-dimensional ore-forming index field.

[0221] A data optimization unit is configured to optimize the regions with a value lower than a threshold value in the three-dimensional ore-forming index field by fusing geological data and interpolation results.

[0222] As an embodiment of the present embodiment, the probability prediction modeling module comprises:

[0223] A data extraction unit is configured to extract numerical distribution data and spatial position coordinates from the three-dimensional ore-forming index field.

[0224] A feature structuring unit is configured to extract structured features from the numerical distribution data by using a grid division tool.

[0225] A model training unit is configured to train a prediction model of the probability relationship between the ore-forming index and gold occurrence by using a Bayesian network algorithm.

[0226] A probability output unit is configured to output a gold occurrence probability value by using the prediction model when the ore-forming index is higher than a critical value.

[0227] The system or device is used to realize the functions of the methods in the above-mentioned embodiments, each module in the system or device corresponds to each step in the method, which has been described in the method and will not be repeated here.

[0228] Through the above embodiments, the problem of gold mine prediction based on deep drilling data in the related art is solved, thereby being able to ensure solving the problems in the prior art.

[0229] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A gold deposit prediction method based on deep drilling data, characterized in that, Includes the following steps: The raw deep-earth drilling data of the gold deposit to be tested are standardized to obtain a standardized drilling dataset. Key mineralization factors were identified from the standardized drilling dataset through importance assessment; Calculate the spatial distance and directional angle between the key mineralization factors, generate a spatial relationship matrix, and determine the coupling strength coefficient; Based on the weighted fusion of multiple mineralization factors using the aforementioned coupling strength coefficient, a comprehensive mineralization index distribution layer is generated. Spatial interpolation is performed on the comprehensive metallogenic index distribution layer to generate a three-dimensional metallogenic index field; A mineralization probability prediction model is established based on the three-dimensional metallogenic index field, and the probability value of gold deposit occurrence is output. The gold deposit occurrence probability value is converted into a probability distribution volume and then rendered in three dimensions to generate a visualization model; Based on the probability isosurface distribution of the visualization model, the boundaries of high-probability mineralization areas are extracted, and the target area location is determined by combining the original geological structure data.

2. The method according to claim 1, characterized in that, The standardization process includes: To obtain lithological assemblage information, coordinates of tectonic intersections, geochemical element content values, and stratigraphic depth parameters from raw deep-earth drilling data; Dimensional normalization was performed on lithological assemblage information, elemental content values, and formation depth parameters, and data exceeding the threshold range were adjusted to the standard range. The wavelet transform denoising method is used to filter noise from lithological combination information and element content values, and to remove outliers that deviate from the smooth curve. A classification algorithm was used to spatially group the coordinates of the tectonic intersections and the stratigraphic depth parameters, and the data was then processed in a structured manner based on the grouping results.

3. The method according to claim 1, characterized in that, The process of determining key ore-forming factors includes: Distribution characteristics of lithological assemblage, tectonic density, chemical anomaly intensity, and burial depth were extracted from standardized drilling datasets. By dividing the region through spatial analysis, areas with tectonic density or chemical anomaly intensity exceeding the threshold are marked as potential mineralization areas; Extract the characteristics of lithological assemblage and burial depth within potential mineralization areas, and classify the characteristics that meet the mineralization conditions as key focus objects; The importance of chemical anomalies and tectonic density of key targets is assessed, and geological elements with weight values ​​exceeding the threshold are marked as key mineralization factors.

4. The method according to claim 1, characterized in that, The process of generating the spatial relation matrix includes: Extract the spatial distribution coordinates of key ore-forming factors; Spatial distances and directional angles between various mineralization factors are measured using spatial calculation tools; Construct a spatial relationship matrix of ore-forming factors that includes spatial distance and directional angle; The correlation of factors in the spatial relationship matrix is ​​quantified, and factor pairs with correlation values ​​exceeding a threshold are marked as highly coupled factor pairs.

5. The method according to claim 1, characterized in that, The process of generating the three-dimensional metallogenic index field includes: Analyze the numerical gradient of known regions in the comprehensive metallogenic index distribution layer; The mineralization index of an unknown region is estimated using a spatial interpolation algorithm based on numerical gradients. A continuous three-dimensional metallogenic index field is generated by mapping spatial coordinates to interpolated metallogenic indices. For regions in the three-dimensional metallogenic index field that are below the threshold, the geological data and interpolation results are integrated for optimization.

6. The method according to claim 1, characterized in that, The process of establishing the mineralization probability prediction model includes: Numerical distribution data and spatial coordinates were extracted from the three-dimensional metallogenic index field; A structured feature set is obtained by extracting features from numerical distribution data using a grid partitioning tool; A predictive model is generated by training the probabilistic relationship between the mineralization index and gold deposit occurrence using a Bayesian network algorithm. When the mineralization index is higher than the critical value, the prediction model outputs a gold deposit occurrence probability value.

7. A gold deposit prediction system based on deep drilling data, characterized in that, The system includes: The data standardization processing module is used to standardize the raw deep-ground drilling data of the gold mine to be tested, so as to obtain a standardized drilling dataset. A key mineralization factor identification module is used to determine key mineralization factors from the standardized drilling dataset through importance assessment; The spatial coupling analysis module is used to calculate the spatial distance and directional angle between the key mineralization factors, generate a spatial relationship matrix, and determine the coupling strength coefficient. The index fusion generation module is used to generate a comprehensive mineralization index distribution layer by weighted fusion of multiple mineralization factors based on the coupling strength coefficient. The three-dimensional interpolation modeling module is used to perform spatial interpolation on the comprehensive metallogenic index distribution layer to generate a three-dimensional metallogenic index field. The probability prediction modeling module is used to establish a mineralization probability prediction model based on the three-dimensional mineralization index field and output a gold deposit occurrence probability value. The visualization rendering module is used to convert the gold ore occurrence probability value into a probability distribution and perform three-dimensional rendering to generate a visualization model. The target area localization module is used to extract the boundary of high-probability mineralization area based on the probability isosurface distribution of the visualization model, and determine the target area location by combining it with the original geological structure data.

8. The system according to claim 7, characterized in that, The data standardization processing module includes: The raw data acquisition unit is used to acquire lithological combination information, coordinates of structural intersections, geochemical element content values, and stratigraphic depth parameters from the raw data of deep drilling. The dimensionless normalization unit is used to normalize the lithological assemblage information, elemental content values ​​and formation depth parameters, and adjust data that exceeds the threshold range to the standard range. The noise filtering unit is used to filter noise from lithological combination information and element content values ​​through wavelet transform denoising method, and remove outliers that deviate from the smooth curve. Spatial grouping units are used to spatially group the coordinates of tectonic intersections and stratigraphic depth parameters using a classification algorithm, and then process the data in a structured manner based on the grouping results.

9. A computer terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the gold ore prediction method based on deep drilling data as described in any one of claims 1-6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the gold ore prediction method based on deep drilling data as described in any one of claims 1-6.

Images