Intelligent delineation system for prospecting target area and abnormal component separation method thereof

Abstract

The invention provides a prospecting target area intelligent delineation system and an abnormal component separation method thereof. All metallogenic indication elements in geochemical sampling data are combined into a geochemical data set; taking the information contribution degree of each metallogenic indication element in the geochemical data set and the metallogenic relevance among the metallogenic indication elements as a screening criterion to extract metallogenic response data of various mineral product types; decomposing each piece of response data into a high-frequency data component used for representing direct association with the mineralization process and a low-frequency data component used for representing association with the regional ore control geological background; determining a plurality of first critical thresholds based on the high-frequency data component, and determining a second critical threshold according to the low-frequency data component; and performing quantitative division on the metallogenic correlation abnormal component and the background component based on all the first critical thresholds and the second critical thresholds. By adopting the scheme of the invention, the abnormal features of different metal combinations can be separated from the complex associated geochemical data.

Description

Technical Field

[0001] This application relates to the field of mineral exploration technology, and more specifically, to an intelligent delineation system for mineral exploration target areas and a method for separating abnormal components thereon. Background Technology

[0002] Intelligent delineation of mineral exploration target areas is a key technology in the field of mineral resource exploration. Its core is to rely on multi-source exploration data such as geophysical and geochemical data, and through data processing, anomaly identification and information extraction, to accurately delineate areas with mineralization potential (i.e., mineral exploration target areas), providing a scientific basis for subsequent exploration engineering deployment. In practical applications, this technology needs to integrate information such as the metallogenic geological background and element distribution patterns of the target mineral cluster area, and filter effective signals related to mineralization from massive amounts of data, ultimately achieving quantitative and intelligent delineation of the target area. It is a core means to improve exploration efficiency and reduce mineral exploration costs, and is especially suitable for large-scale exploration work in large mineral cluster areas.

[0003] In existing technologies, only mineralization information of a single mineral type and a certain element is often considered, without evaluating the resource quantity of mineralized elements in the delineated prospective areas. However, in actual operations, there are often varying degrees of correlation between the various elements in the metallogenic system, and the spatial distribution characteristics of geochemical models are affected by various geological processes. The key to geochemical anomaly identification is how to identify the geochemical background and distinguish anomalies. It is difficult to solve the problem of element distribution superposition caused by multi-stage mineralization. For example, in the Jiaodong gold mining area, traditional methods often fail to distinguish the anomaly characteristics of precious metal assemblages such as Au-Ag-Cd from key metal assemblages such as Be-La-Mo, or confuse mineralization signals with regional background, resulting in insufficient accuracy of the delineated target area and difficulty in meeting the needs of deep exploration. Therefore, how to separate the anomaly characteristics of different metal assemblages from complex and correlated geochemical data has become a difficult problem for the industry. Summary of the Invention

[0004] This application provides an intelligent delineation system for mineral exploration target areas and a method for separating anomaly components, which can separate the anomalous features of different metal combinations from complex and correlated geochemical data.

[0005] In a first aspect, this application provides a method for separating abnormal components in an intelligent delineation system for mineral exploration target areas, comprising: Obtain geochemical sampling data of the target mineral cluster area, and combine all mineralization indicator elements in the geochemical sampling data into a geochemical dataset; Based on the information contribution of each ore-forming indicator element in the geochemical dataset and the ore-forming correlation between each ore-forming indicator element, response data of ore-forming elements of various mineral types are extracted from the geochemical dataset. Each response data point is decomposed into a high-frequency data component that characterizes the mineralization process directly and a low-frequency data component that characterizes the regional ore-controlling geological background. Based on the cumulative frequency characteristics of the high-frequency data components, a first critical threshold for local mineralization anomalies of various minerals is determined, and based on the statistical characteristics of the low-frequency data components, a second critical threshold for favorable background data for regional mineralization is determined. The mineralization-related anomaly components and background components are quantitatively classified based on all first and second critical thresholds.

[0006] In some embodiments, combining all mineralization indicator elements from the geochemical sampling data into a geochemical dataset specifically includes: Based on the data distribution patterns of ore-forming indicator elements, ore-forming indicator elements that are potentially related to the ore-forming process are screened from the geochemical sampling data; All mineralization indicator elements were combined into a geochemical dataset.

[0007] In some embodiments, extracting response data for multiple mineral types from the geochemical dataset using the information contribution of each ore-forming indicator element and the ore-forming correlation between each ore-forming indicator element as screening criteria specifically includes: Determine the multiple information contributions of each ore-forming indicator element in the geochemical dataset and the ore-forming correlations among the ore-forming indicator elements; The quantification matrix of the mineralization correlation is orthogonally reconstructed to obtain an orthogonal correlation matrix; Multiple mineralization indicator elements that meet the correlation criteria are selected from the orthogonal correlation matrix; Based on the information contribution of all selected mineralization indicator elements, the selected mineralization indicator elements are classified into response data of mineralization of multiple mineral types.

[0008] In some embodiments, determining the multiple information contributions of each ore-forming indicator element in the geochemical dataset and the ore-forming correlations among the ore-forming indicator elements specifically includes: Logarithmically normalize the concentration data of each mineralization indicator element in the geochemical dataset to obtain the standard concentration data of each mineralization indicator element. Determine the covariance matrix between the concentrations of each element based on the standard concentration data of each mineralization indicator element; The mineralization correlation among the mineralization indicator elements is determined based on the covariance matrix among the concentrations of each element. The information contribution of each mineralization indicator element is determined based on the mineralization correlation among the various mineralization indicator elements.

[0009] In some embodiments, decomposing each response data into a high-frequency data component directly associated with the mineralization process and a low-frequency data component associated with the regional ore-controlling geological background specifically includes: Select one response data as the selected response data, and integrate the selected response data into an initial distribution map of mineral resources within the target mineral cluster area; The initial distribution map is decomposed into multiple concentration data components; Based on the frequency of all concentration data components, all concentration data components are divided into high-frequency data components used to characterize the mineralization process and low-frequency data components used to characterize the regional ore-controlling geological background. The remaining response data were further decomposed into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional ore-controlling geological background.

[0010] In some embodiments, decomposing the initial distribution map into multiple concentration data components specifically includes: Determine the average trend surface of the concentration distribution in the initial distribution map; The process components are obtained by removing the average trend surface from the initial distribution map; If the process component does not meet the preset termination condition, the average trend surface of the process component is removed from the process component to obtain a new process component until the obtained process component meets the preset termination condition. The obtained process component is then used as a concentration data component. The concentration data components are removed from the initial distribution map to obtain a new initial distribution map. The above steps are repeated until the number of extreme points in the obtained initial distribution map is less than a preset extreme point threshold, thereby obtaining multiple concentration data components.

[0011] In some embodiments, determining the first critical threshold for multiple types of local mineralization anomaly data based on the cumulative frequency characteristics of the high-frequency data components specifically includes: Extreme interference values ​​are removed from all high-frequency data components to obtain multiple denoised high-frequency components; Select a denoised high-frequency component as the selected high-frequency component, and group the selected high-frequency component according to logarithmic intervals to obtain multiple grouped data; Based on the cumulative frequency of each group of data, a cumulative frequency curve of the concentration distribution of ore-forming elements within the target ore cluster area is plotted. The first critical threshold of local mineralization anomaly data in the mineral type corresponding to the selected high-frequency component is determined by the cumulative frequency curve. Continue to determine the first critical threshold of the local mineralization anomaly data in the mineral type corresponding to the remaining denoised high-frequency components.

[0012] In some embodiments, determining the second critical threshold for favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components specifically includes: Remove extreme interference values ​​from all low-frequency data components to obtain multiple denoised low-frequency components; The second critical threshold for favorable background data for regional mineralization was determined by statistical characteristics of the concentration data in all denoised low-frequency components.

[0013] In some embodiments, the geochemical sampling data is geochemical data of stream sediments collected within the target mineralized area.

[0014] Secondly, this application provides an intelligent delineation system for mineral exploration target areas, which includes an anomaly component separation unit, the anomaly component separation unit comprising: The acquisition module is used to acquire geochemical sampling data of the target mineral cluster area and combine all mineralization indicator elements in the geochemical sampling data into a geochemical dataset. The processing module is used to extract response data of mineralization of various mineral types from the geochemical dataset based on the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element as screening criteria. The processing module is also used to decompose each response data into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional ore-controlling geological background. The processing module is also used to determine the first critical threshold of local mineralization anomaly data of multiple types of minerals based on the cumulative frequency characteristics of the high-frequency data components, and to determine the second critical threshold of favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components. The execution module is used to quantitatively classify the mineralization-related anomaly components and background components based on all first and second critical thresholds.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The intelligent delineation system for mineral exploration target areas and its anomaly component separation method provided in this application first acquires geochemical sampling data of the target mineral cluster area, and combines all mineralization indicator elements in the geochemical sampling data into a geochemical dataset; using the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element as screening criteria, response data of mineralization for multiple mineral types are extracted from the geochemical dataset; each response data is decomposed into a high-frequency data component used to characterize the direct correlation with the mineralization process and a low-frequency data component used to characterize the correlation with the regional mineralization-controlling geological background; based on the cumulative frequency characteristics of the high-frequency data component, a first critical threshold for local mineralization anomaly data of multiple mineral types is determined, and based on the statistical characteristics of the low-frequency data component, a second critical threshold for favorable regional mineralization background data is determined; based on all the first critical thresholds and the second critical thresholds, the mineralization-related anomaly component and the background component are quantitatively divided.

[0016] Therefore, this application first screens potential related elements and constructs a dataset based on the data distribution patterns of ore-forming indicator elements, avoiding the abnormal interference caused by the inclusion of irrelevant elements in traditional methods, thus improving the effectiveness of ore-forming information from the data source. Secondly, by determining the contribution of elemental information and its ore-forming correlation, response data for multiple mineral types is extracted using a dual-criteria approach, effectively solving the problem of omission or redundancy of ore-forming information caused by single data screening, ensuring a high degree of matching between the response data and the specific mineralization process. Furthermore, the response data is decomposed into high-frequency (mineralization correlation) and low-frequency (regional background) components, accurately separating the confused local anomalies and regional background signals in traditional methods, laying the foundation for subsequent threshold division. Finally, based on the cumulative frequency characteristics of the high-frequency components, a first critical threshold for local ore-forming anomalies is determined, and combined with the statistical characteristics of the low-frequency components, a second critical threshold for favorable regional background is determined. This dual-threshold approach achieves quantitative division between ore-forming correlation anomalies and background. In summary, the scheme of this application can separate the anomalous features of different metal assemblages from complexly correlated geochemical data. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of an abnormal component separation method in a smart delineation system for mineral exploration target areas, as shown in some embodiments of this application. Figure 2 This is an exemplary flowchart illustrating the extraction of response data according to some embodiments of this application; Figure 3 This is an exemplary flowchart illustrating the determination of a first critical threshold according to some embodiments of this application; Figure 4 This is a schematic diagram of the abnormal component separation unit according to some embodiments of this application; Figure 5This is a schematic diagram of the structure of a computer device for implementing an abnormal component separation method of an intelligent delineation system for mineral exploration target areas, according to some embodiments of this application. Detailed Implementation

[0018] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] refer to Figure 1 The figure is an exemplary flowchart of an anomaly component separation method of a smart delineation system for mineral exploration target areas according to some embodiments of this application. The anomaly component separation method of the smart delineation system for mineral exploration target areas mainly includes the following steps: In step 101, geochemical sampling data of the target mineral cluster area is obtained, and all mineralization indicator elements in the geochemical sampling data are combined into a geochemical dataset.

[0020] In practice, the geochemical sampling data of the target mineralization area can be obtained in the following way: First, collect 1:20000 stream sediment geochemical data in the target mineralization area, with a sampling density of one sampling point per square kilometer. Collect stream sediment geochemical data at each sampling point. Finally, use the collected stream sediment geochemical data as the geochemical sampling data of the target mineralization area. The 1:20000 stream sediment geochemical data collected in the sampling area can be the 1:20000 stream sediment geochemical data collected by the regional geochemical exploration national sweep plan. The stream sediment geochemical data includes, but is not limited to, the concentration values ​​of ore-forming elements such as Ag, Au, La, Mo, Pb, W, Sb, Zn, and Th.

[0021] In some embodiments, combining all ore-forming indicator elements from the geochemical sampling data into a geochemical dataset can be achieved using the following steps: Based on the data distribution patterns of ore-forming indicator elements, ore-forming indicator elements that are potentially related to the ore-forming process are screened from the geochemical sampling data; All mineralization indicator elements were combined into a geochemical dataset.

[0022] In practice, based on the data distribution patterns of ore-forming indicator elements, ore-forming indicator elements with potential correlation to the mineralization process are screened from the geochemical sampling data. This can be achieved in the following way: First, calculate the average concentration of each ore-forming element in the geochemical sampling data within the target ore-bearing area. Simultaneously, obtain the background average concentration value of each ore-forming element within the corresponding tectonic unit of the target ore-bearing area (this can be obtained through literature or publicly available regional geochemical survey results, for example, the literature "Chemical composition of the continental crust as revealed by studies in..." by GAO S et al.). The study, titled *East China*, investigated the geochemical characteristics of various elements in eastern China. If the target ore-bearing area corresponds to a tectonic unit in eastern China, the background average concentration values ​​of each ore-forming element can be directly obtained from this literature. The enrichment coefficient of each ore-forming element is calculated using the formula: "Enrichment coefficient = Average element concentration in the ore-bearing area / Background average element concentration in the tectonic unit". Ore-forming elements with enrichment coefficients ≥ a preset enrichment coefficient threshold are initially selected as candidate ore-forming indicator elements. Subsequently, statistical analysis tools are used to verify the distribution patterns of the concentration values ​​of the candidate ore-forming indicator elements. This involves plotting frequency distribution histograms and log-normal probability plots of each concentration value, and determining whether each candidate ore-forming indicator element follows a log-normal distribution or... The Pareto distribution is used to retain candidate mineralization indicator elements that follow a log-normal or Pareto distribution. Then, the coefficient of variation (coefficient of variation = standard deviation of concentration / average concentration) of each retained candidate mineralization indicator element is calculated. Candidate mineralization indicator elements with a coefficient of variation ≥ a preset coefficient of variation threshold are preferentially retained, that is, candidates mineralization indicator elements with uneven distribution that may be affected by mineralization. Finally, each of the finally retained candidate mineralization indicator elements is regarded as a mineralization indicator element that has a potential association with the mineralization process. The enrichment coefficient threshold and the coefficient of variation threshold can be preset by referring to the "Regional Geochemical Exploration Specification". For example, in this application, the enrichment coefficient threshold is preset to 1.2 and the coefficient of variation threshold is preset to 0.5.

[0023] In practice, combining all mineralization indicator elements into a geochemical dataset can be achieved in the following way: the concentration values ​​of all mineralization indicator elements are integrated into a matrix, where each row of the matrix corresponds to a sampling point and each column corresponds to an element, that is, the i-th row and j-th column corresponds to the concentration value of the j-th element at the i-th sampling point, and this matrix is ​​used as the geochemical dataset.

[0024] In step 102, response data of mineralization for various mineral types are extracted from the geochemical dataset based on the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element.

[0025] In some embodiments, reference Figure 2 The figure is an exemplary flowchart illustrating the extraction of response data according to some embodiments of this application. In this application, the extraction of response data for multiple mineral types from the geochemical dataset using the information contribution of each ore-forming indicator element and the ore-forming correlation between them as screening criteria can be achieved in the following manner: In step 1021, the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element are determined. In step 1022, the quantization matrix of the mineralization correlation is orthogonally reconstructed to obtain an orthogonal correlation matrix; In step 1023, multiple mineralization indicator elements that meet the correlation degree criteria are selected from the orthogonal correlation matrix; In step 1024, based on the information contribution of all selected mineralization indicator elements, all selected mineralization indicator elements are classified into response data of mineralization of multiple mineral types.

[0026] In some embodiments, determining the multiple information contributions of each ore-forming indicator element in the geochemical dataset and the ore-forming correlations among the ore-forming indicator elements can be achieved by the following steps: Logarithmically normalize the concentration data of each mineralization indicator element in the geochemical dataset to obtain the standard concentration data of each mineralization indicator element. Determine the covariance matrix between the concentrations of each element based on the standard concentration data of each mineralization indicator element; The mineralization correlation among the mineralization indicator elements is determined based on the covariance matrix among the concentrations of each element. The information contribution of each mineralization indicator element is determined based on the mineralization correlation among the various mineralization indicator elements.

[0027] In specific implementation, the standard concentration data of each mineralization indicator element in the geochemical dataset is logarithmically standardized to obtain the standard concentration data of each mineralization indicator element. This can be achieved in the following way: First, a central logarithmic ratio transformation is performed on the concentration data of each mineralization indicator element. That is, the geometric mean of the concentration data of all mineralization indicator elements in each sampling point in the geochemical dataset is calculated first. Then, for each sampling point, the concentration data of each mineralization indicator element in each sampling point is divided by the geometric mean of each sampling point, and the natural logarithm of the quotient is taken. The values ​​obtained after taking the natural logarithm are used as the standard concentration data of each mineralization indicator element in each sampling point. If there are concentration data of 0 in the geochemical dataset, the concentration data of all sampling points containing 0 should be logarithmically standardized beforehand to eliminate the 0 values.

[0028] In practice, the covariance matrix between the concentrations of each element can be determined based on the standard concentration data of each mineralization indicator element in the following way: First, the standard concentration data of each mineralization indicator element are re-integrated into a matrix, where each row of the matrix corresponds to a sampling point and each column corresponds to an element, that is, the i-th row and j-th column corresponds to the concentration value of the j-th element of the i-th sampling point. Then, the covariance matrix of this matrix is ​​calculated and used as the covariance matrix between the concentrations of each element.

[0029] In specific implementation, the mineralization correlation between each mineralization indicator element can be determined based on the covariance matrix between the concentrations of each element in the following way: First, calculate multiple eigenvalues ​​of the covariance matrix between the concentrations of each element and the eigenvector corresponding to each eigenvalue. Then, take all eigenvectors with eigenvalues ​​greater than a preset screening threshold (usually set to 1) as principal components related to mineralization. Then, for the a-th mineralization indicator element, take the product of the a-th element in each principal component and the square root of the eigenvalue of the principal component as the correlation weight of the a-th mineralization indicator element on each principal component. Then, integrate all the correlation weights into a matrix, where each row of the matrix corresponds to a mineralization indicator element and each column corresponds to a principal component. That is, the k-th row and t-th column correspond to the correlation weight of the k-th mineralization indicator element on the t-th principal component. This matrix is ​​used as the quantification matrix of the mineralization correlation between each mineralization indicator element.

[0030] It should be noted that in this application, mineralization correlation refers to the synergistic correlation formed by the concentration changes of mineralization indicator elements due to their joint participation in the same mineralization process (such as magmatic hydrothermal activity and tectonic mineralization). The quantitative matrix of mineralization correlation can directly provide a basis for classifying the mineralization element combinations corresponding to different mineral types (such as distinguishing different element combinations such as Au-Ag-Cd gold deposit combination and Be-La-Mo-Nb-Th-UY key metal combination).

[0031] In specific implementation, the determination of multiple information contributions of each mineralization indicator element based on the mineralization correlation between the mineralization indicator elements can be achieved in the following way: First, obtain all principal components related to mineralization and calculate the information contribution of each principal component. Then, multiply the correlation weight of each column in the mineralization correlation quantification matrix by the information contribution of the corresponding principal component. Finally, use the product of each row in the resulting matrix as the multiple information contributions of each mineralization indicator element.

[0032] It should be noted that, in this application, the information contribution degree is a characterization of the degree to which the mineralization indicator elements contribute to the mineralization information carried by each mineralization-related principal component.

[0033] In specific implementation, the orthogonal reconstruction of the quantification matrix of the mineralization correlation can be achieved in the following way: the quantification matrix of the mineralization correlation can be orthogonally rotated using the orthogonal rotation algorithm in the prior art, and the iteration is performed with the goal of maximizing the variance of all correlation weights corresponding to each principal component in the quantification matrix, and the matrix obtained after orthogonal rotation is used as the orthogonal correlation matrix.

[0034] It should be noted that the orthogonal correlation matrix in this application refers to the matrix obtained by orthogonally rotating the quantification matrix of mineralization correlation. The element in the b-th row and c-th column of the orthogonal correlation matrix is ​​the orthogonal correlation weight of the b-th mineralization indicator element on the c-th principal component. This orthogonal correlation weight is the correlation weight after orthogonal rotation.

[0035] In specific implementation, the following method can be used to select multiple mineralization indicator elements that meet the correlation degree from the orthogonal correlation matrix: First, a threshold value is preset (usually set to 1), and elements greater than or equal to the threshold value are used as the criteria for meeting the criteria. Then, each element in the orthogonal correlation matrix is ​​traversed, and all mineralization indicator elements whose absolute values ​​of orthogonal correlation weights on at least one principal component meet the criteria for meeting the criteria are selected. The element in the b-th row and c-th column of the orthogonal correlation matrix is ​​the orthogonal correlation weight of the b-th mineralization indicator element on the c-th principal component.

[0036] In practice, the response data of all selected mineralization indicator elements can be classified into multiple mineral types based on the information contribution of all selected mineralization indicator elements. This can be achieved by classifying the concentration data of mineralization indicator elements with an information contribution of ≥1% (a known low threshold to ensure that the element has an actual information contribution to the principal component) into the response data of the same mineral type.

[0037] It should be noted that the response data in this application refers to the concentration data of ore-forming indicator elements belonging to the same mineral type.

[0038] In step 103, each response data is decomposed into a high-frequency data component that characterizes the mineralization process and a low-frequency data component that characterizes the regional mineralization geological background.

[0039] In some embodiments, decomposing each response data point into a high-frequency data component that characterizes the mineralization process and a low-frequency data component that characterizes the regional ore-controlling geological background can be achieved by the following steps: Select one response data as the selected response data, and integrate the selected response data into an initial distribution map of mineral resources within the target mineral cluster area; The initial distribution map is decomposed into multiple concentration data components; Based on the frequency of all concentration data components, all concentration data components are divided into high-frequency data components used to characterize the mineralization process and low-frequency data components used to characterize the regional ore-controlling geological background. The remaining response data were further decomposed into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional ore-controlling geological background.

[0040] In practice, the initial distribution map of mineral distribution within the target mineral cluster area can be integrated from the selected response data in the following way: all concentration data in the selected response data are integrated into a matrix, with each row of the matrix corresponding to a sampling point and each column corresponding to a mineralization indicator element.

[0041] It should be noted that the initial distribution map in this application is a matrix diagram used to visually present the distribution of element concentrations related to a specified mineral type within the target mineral cluster area.

[0042] In some embodiments, decomposing the initial distribution map into multiple concentration data components can be achieved using the following steps: Determine the average trend surface of the concentration distribution in the initial distribution map; The process components are obtained by removing the average trend surface from the initial distribution map; If the process component does not meet the preset termination condition, the average trend surface of the process component is removed from the process component to obtain a new process component until the obtained process component meets the preset termination condition. The obtained process component is then used as a concentration data component. The concentration data components are removed from the initial distribution map to obtain a new initial distribution map. The above steps are repeated until the number of extreme points in the obtained initial distribution map is less than a preset extreme point threshold, thereby obtaining multiple concentration data components.

[0043] In specific implementation, the average trend surface of the concentration distribution in the initial distribution map can be determined in the following way: First, extract all the maximum points and all the minimum points in the initial distribution map. Then, use a two-dimensional interpolation method, such as radial basis function, thin plate spline method or linear interpolation after triangulation, to fit all the maximum points into a plane. Similarly, use a two-dimensional interpolation method to fit all the minimum points into another plane. Then, calculate the average plane of the two planes, that is, add the two planes together and divide by two. The resulting plane is used as the average trend surface of the concentration distribution in the initial distribution map.

[0044] It should be noted that the average trend surface in this application is a plane used to characterize the overall trend characteristics of the element concentration distribution related to a specified mineral type in the target mineral cluster area.

[0045] In a specific implementation, the process component can be obtained by removing the average trend surface from the initial distribution map in the following way: subtract the average trend surface from the initial distribution map and use the resulting matrix as the process component.

[0046] It should be noted that the process components in this application are only process quantities in the iterative decomposition of response data, and are only for the purpose of description and have no actual meaning.

[0047] Additionally, it should be noted that the termination condition in this application refers to the ratio of the number of maximum points to the number of minimum points in the process component being within the range of 0.8-1.2, that is, ensuring that the fluctuation of the process component does not have a significant unidirectional bias.

[0048] In specific implementation, the mean trend surface of the process component is further removed from the process component. The new process component can be implemented in the following way: First, extract all the maximum points and all the minimum points of the process component. Then, fit all the maximum points into a plane using a two-dimensional interpolation method, such as radial basis function, thin plate spline method or linear interpolation after triangulation. Similarly, fit all the minimum points into another plane using the same two-dimensional interpolation method. Then, calculate the mean plane of the two planes, that is, add the two planes and divide by two. Subtract the mean plane from the process component. Finally, take the matrix obtained after subtracting the mean plane as the new process component.

[0049] In a specific implementation, removing the concentration data component from the initial distribution map to obtain a new initial distribution map can be achieved in the following way: subtracting the concentration data component from the initial distribution map to obtain the matrix as the new initial distribution map.

[0050] In specific implementation, dividing all concentration data components into high-frequency data components directly related to the mineralization process and low-frequency data components related to the regional ore-controlling geological background can be achieved as follows: First, for each concentration data component, the two-dimensional discrete Fourier transform (DFT) is used to transform each component from the spatial domain to the frequency domain, and the power spectral density matrix of each frequency domain matrix is ​​calculated. Then, the coordinates (u, v, where u corresponds to the frequency domain index in the row direction and v corresponds to the frequency domain index in the column direction) of each power spectral density matrix are recorded, and all coordinates are transformed back to the dominant frequency in the spatial domain. The maximum value is taken as the dominant frequency of each concentration data component. Finally, all concentration data components are divided using a frequency threshold. To characterize high-frequency data components directly related to mineralization processes and low-frequency data components related to regional ore-controlling geological backgrounds, all concentration data components with a dominant frequency greater than or equal to a frequency threshold are considered high-frequency data components directly related to mineralization processes, while all concentration data components with a dominant frequency less than the frequency threshold are considered low-frequency data components related to regional ore-controlling geological backgrounds. The frequency threshold can be preset based on geological survey experience. For example, regional ore-controlling geological backgrounds (such as craton basements and large magmatic rock belts) typically correspond to a large-scale distribution with a spatial scale of ≥10km, and their dominant frequency is generally ≤0.1 cycles / km. In contrast, mineralization processes (such as ore bodies and mineralized zones) are mostly local features with a spatial scale of ≤2km, and their dominant frequency is usually ≥0.5 cycles / km. Therefore, 0.3 cycles / km can be set as the frequency division threshold.

[0051] It should be noted that in this application, the high-frequency data component is the concentration data component used to characterize local anomalies directly related to the mineralization process, and the low-frequency data component is the concentration data component used to characterize the regional ore-controlling geological background. The mineralization process (such as ore body and mineralization zone) is a local small-scale geological phenomenon with a small spatial range and drastic changes in element concentration, corresponding to a high signal frequency; the regional ore-controlling geological background (such as Linglong granite body and Guojialing granodiorite body) is a large-scale macroscopic geological body with a large spatial span and gradual changes in element concentration, corresponding to a low signal frequency. Therefore, the high-frequency data component is directly related to the mineralization process, and the low-frequency data component is related to the regional ore-controlling background.

[0052] In step 104, a first critical threshold for local mineralization anomaly data of multiple types of minerals is determined based on the cumulative frequency characteristics of the high-frequency data components, and a second critical threshold for favorable background data for regional mineralization is determined based on the statistical characteristics of the low-frequency data components.

[0053] In some embodiments, reference Figure 3The figure is an exemplary flowchart illustrating the determination of a first critical threshold according to some embodiments of this application. The determination of the first critical threshold for multiple types of local mineralization anomaly data based on the cumulative frequency characteristics of the high-frequency data components in this application can be achieved using the following steps: In step 1041, extreme interference values ​​in all high-frequency data components are removed to obtain multiple denoised high-frequency components; In step 1042, a denoised high-frequency component is selected as the selected high-frequency component, and the selected high-frequency component is grouped according to logarithmic intervals to obtain multiple grouped data. In step 1043, a cumulative frequency curve of the concentration distribution of ore-forming elements in the target ore cluster area is plotted based on the cumulative frequency of each group of data. In step 1044, the first critical threshold of the local mineralization anomaly data in the mineral type corresponding to the selected high-frequency component is determined by the cumulative frequency curve. In step 1045, the first critical threshold of the local mineralization anomaly data in the mineral type corresponding to the remaining denoised high-frequency components is determined.

[0054] In practice, removing extreme interference values ​​from all high-frequency data components to obtain multiple denoised high-frequency components can be achieved in the following way: extreme interference values ​​can be removed by using the three-times-standard-deviation criterion of the normal distribution. That is, the average value and standard deviation of the concentration data in each high-frequency data component are calculated. Then, the concentration data that exceeds the average value ± three times the standard deviation are removed from each high-frequency data component, and the resulting components are all used as denoised high-frequency components.

[0055] It should be noted that the denoised high-frequency component in this application is the high-frequency data component after removing sampling errors.

[0056] In practice, the selected high-frequency components are grouped according to logarithmic intervals to obtain multiple grouped data. This can be achieved in the following way: First, take the natural logarithm of the minimum value of all concentration data in the selected high-frequency components as the starting point, divide multiple intervals according to a distribution interval of 0.5, and finally divide the natural logarithms of all concentration data in the selected high-frequency components into each interval, and take the concentration data in each interval as a group of grouped data.

[0057] In practice, the cumulative frequency curve of the concentration distribution of ore-forming elements in the target ore-forming area based on the cumulative frequency of each group of data can be implemented in the following way: First, calculate the cumulative frequency of each group of data. For the nth group of data, divide the total length of the first to the nth group of data by the total length of all group data to obtain the cumulative frequency of the nth group of data. Finally, plot all the cumulative frequencies on the same coordinate system, with the vertical axis representing the cumulative frequency and the horizontal axis representing the interval range of the group of data. Plot each cumulative frequency at the midpoint of the interval range to obtain multiple data points. Finally, fit all the data points into a curve using the Lagrange interpolation method. The obtained curve is used as the cumulative frequency curve of the concentration distribution of ore-forming elements in the target ore-forming area.

[0058] It should be noted that the cumulative frequency curve in this application is a statistical curve that reflects the distribution law of ore-forming element concentration in the target ore cluster area.

[0059] In specific implementation, the first critical threshold of local mineralization anomaly data in the mineral type corresponding to the selected high-frequency component can be determined by the following method: take the horizontal axis corresponding to the inflection point in the cumulative frequency curve and take the value of the natural exponential function of the horizontal axis as the first critical threshold of local mineralization anomaly data in the mineral type corresponding to the selected high-frequency component.

[0060] It should be noted that the first critical threshold in this application is the critical value of ore-forming element concentration used to distinguish between local mineralization anomaly data and regional background data in the mineral type corresponding to the selected high-frequency component. The cumulative frequency curve of geochemical data usually presents two distinct segments. The first segment of the curve, i.e., the data points with a cumulative frequency of 0% to 90%, is roughly distributed along a straight line, corresponding to the background data of ore-forming elements in the region (mostly normally distributed non-mineralized background values). The second segment of the curve, i.e., the data points with a cumulative frequency of 90% to 100%, deviates significantly from the straight line, corresponding to local anomaly data affected by mineralization.

[0061] In some embodiments, determining the second critical threshold for favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components can be achieved through the following steps: Remove extreme interference values ​​from all low-frequency data components to obtain multiple denoised low-frequency components; The second critical threshold for favorable background data for regional mineralization was determined by statistical characteristics of the concentration data in all denoised low-frequency components.

[0062] In practice, removing extreme interference values ​​from all low-frequency data components to obtain multiple denoised low-frequency components can be achieved in the following way: extreme interference values ​​can be removed by using the three-times-standard-deviation criterion of the normal distribution. That is, calculate the average value and standard deviation of the concentration data in each low-frequency data component, and then remove the concentration data that exceeds the average value ± three times the standard deviation from each low-frequency data component, and use the resulting components as denoised low-frequency components.

[0063] It should be noted that the denoised low-frequency component in this application is the low-frequency data component after removing sampling errors.

[0064] In specific implementation, the second critical threshold for determining favorable background data for regional mineralization based on the statistical characteristics of concentration data in all denoised low-frequency components can be achieved as follows: First, after integrating all denoised low-frequency components, extract the minimum value. Using the natural logarithm of this minimum value as the starting point, divide the data into multiple intervals at 0.5 intervals. Then, divide all concentration data in all denoised low-frequency components into the corresponding intervals according to their natural logarithm values. Calculate the cumulative frequency of concentration data in each interval. Subsequently, fit the obtained cumulative frequency into a curve, with the vertical axis representing the cumulative frequency and the horizontal axis representing the corresponding interval. Then, extract the inflection point on the curve and use the value of the natural exponential function of the horizontal axis corresponding to the inflection point as the second critical threshold for favorable background data for regional mineralization.

[0065] It should be noted that the second critical threshold in this application is a concentration critical value used to define the favorable background data for regional mineralization and the background data for ordinary regions.

[0066] In step 105, the mineralization-related anomaly components and background components are quantitatively divided based on all first critical thresholds and second critical thresholds.

[0067] In specific implementation, the quantitative division of mineralization-related anomaly components and background components based on all first and second critical thresholds can be achieved in the following way: for each high-frequency data component, the concentration data in each high-frequency data component that is higher than the first critical threshold corresponding to each high-frequency data component is taken as the mineralization-related anomaly component corresponding to each high-frequency data component. Subsequently, the concentration data in each high-frequency data component that is lower than the first critical threshold corresponding to each high-frequency data component but higher than the second critical threshold is taken as the favorable background component related to regional mineralization.

[0068] It should be noted that the components in this application that directly characterize the local mineralization anomalies related to the mineralization process (such as the formation of ore bodies and mineralization zones) within the target mineralization area are the components that reflect the favorable background conditions (such as proximity to the ore-controlling rock body or being located in a favorable tectonic position) within the target mineralization area.

[0069] Furthermore, in another aspect of this application, in some embodiments, this application provides an intelligent delineation system for mineral exploration target areas, which includes an anomaly component separation unit, with reference to... Figure 4 The figure is a schematic diagram of the structure of an abnormal component separation unit according to some embodiments of this application. The abnormal component separation unit 400 includes: a data acquisition module 401, a processing module 402, and an execution module 403, which are described below: Acquisition module 401, in this application, is mainly used to acquire geochemical sampling data of the target mineral cluster area and combine all mineralization indicator elements in the geochemical sampling data into a geochemical dataset. Processing module 402, in this application, is mainly used to extract response data of mineralization of various mineral types from the geochemical dataset based on the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element as screening criteria. It should be noted that the processing module 402 in this application is also used to decompose each response data into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional mineralization geological background. It should be noted that the processing module 402 in this application is also used to determine the first critical threshold of local mineralization anomaly data of multiple types of minerals based on the cumulative frequency characteristics of the high-frequency data components, and to determine the second critical threshold of favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components. The execution module 403 in this application is mainly used to quantitatively divide the mineralization-related anomaly component and the background component based on all first critical thresholds and second critical thresholds.

[0070] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described method for separating abnormal components in a smart delineation system for mineral exploration target areas.

[0071] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device for implementing an anomaly component separation method of an intelligent delineation system for mineral exploration target areas, according to some embodiments of this application. The anomaly component separation method of the intelligent delineation system for mineral exploration target areas in the above embodiments can be achieved through… Figure 5 The computer device shown is used to implement this, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0072] Processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0073] The communication bus 502 can be used to transmit information between the aforementioned components.

[0074] Memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CDROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 503 may exist independently and be connected to processor 501 via communication bus 502. Memory 503 may also be integrated with processor 501.

[0075] The memory 503 stores program code for executing the scheme of this application, and its execution is controlled by the processor 501. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. The abnormal component separation method of the intelligent delineation system for mineral exploration target areas in the above embodiments can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0076] Communication interface 504 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0077] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single CPU) processor or a multi-core (multi CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0078] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0079] In addition, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for separating abnormal components in an intelligent delineation system for mineral exploration target areas.

[0080] In summary, the intelligent delineation system for mineral exploration target areas and its anomaly component separation method disclosed in this application firstly acquires geochemical sampling data of the target mineral cluster area and combines all mineralization indicator elements in the geochemical sampling data into a geochemical dataset; using the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element as screening criteria, response data of mineralization for multiple mineral types are extracted from the geochemical dataset; each response data is decomposed into a high-frequency data component used to characterize the direct correlation with the mineralization process and a low-frequency data component used to characterize the correlation with the regional mineralization-controlling geological background; based on the cumulative frequency characteristics of the high-frequency data component, a first critical threshold for local mineralization anomaly data of multiple mineral types is determined, and based on the statistical characteristics of the low-frequency data component, a second critical threshold for favorable regional mineralization background data is determined; based on all the first critical thresholds and the second critical threshold, the mineralization-related anomaly component and the background component are quantitatively divided.

[0081] Therefore, this application first screens potential related elements and constructs a dataset based on the data distribution patterns of ore-forming indicator elements, avoiding the abnormal interference caused by the inclusion of irrelevant elements in traditional methods, thus improving the effectiveness of ore-forming information from the data source. Secondly, by determining the contribution of elemental information and its ore-forming correlation, response data for multiple mineral types is extracted using a dual-criteria approach, effectively solving the problem of omission or redundancy of ore-forming information caused by single data screening, ensuring a high degree of matching between the response data and the specific mineralization process. Furthermore, the response data is decomposed into high-frequency (mineralization correlation) and low-frequency (regional background) components, accurately separating the confused local anomalies and regional background signals in traditional methods, laying the foundation for subsequent threshold division. Finally, based on the cumulative frequency characteristics of the high-frequency components, a first critical threshold for local ore-forming anomalies is determined, and combined with the statistical characteristics of the low-frequency components, a second critical threshold for favorable regional background is determined. This dual-threshold approach achieves quantitative division between ore-forming correlation anomalies and background. In summary, the scheme of this application can separate the anomalous features of different metal assemblages from complexly correlated geochemical data.

[0082] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0083] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for separating abnormal components in an intelligent delineation system for mineral exploration target areas, characterized in that, include: Obtain geochemical sampling data of the target mineral cluster area, and combine all mineralization indicator elements in the geochemical sampling data into a geochemical dataset; Based on the information contribution of each ore-forming indicator element in the geochemical dataset and the ore-forming correlation between each ore-forming indicator element, response data of ore-forming elements of various mineral types are extracted from the geochemical dataset. Each response data point is decomposed into a high-frequency data component that characterizes the mineralization process directly and a low-frequency data component that characterizes the regional ore-controlling geological background. Based on the cumulative frequency characteristics of the high-frequency data components, a first critical threshold for local mineralization anomalies of various minerals is determined, and based on the statistical characteristics of the low-frequency data components, a second critical threshold for favorable background data for regional mineralization is determined. The mineralization-related anomaly components and background components are quantitatively classified based on all first and second critical thresholds.

2. The method as described in claim 1, characterized in that, Combining all mineralization indicator elements from the geochemical sampling data into a geochemical dataset specifically includes: Based on the data distribution patterns of ore-forming indicator elements, ore-forming indicator elements that are potentially related to the ore-forming process are screened from the geochemical sampling data; All mineralization indicator elements were combined into a geochemical dataset.

3. The method as described in claim 1, characterized in that, Using the information contribution of each ore-forming indicator element in the geochemical dataset and the ore-forming correlation between each ore-forming indicator element as screening criteria, response data for various mineral types are extracted from the geochemical dataset. Specifically, this includes: Determine the multiple information contributions of each ore-forming indicator element in the geochemical dataset and the ore-forming correlations among the ore-forming indicator elements; The quantification matrix of the mineralization correlation is orthogonally reconstructed to obtain an orthogonal correlation matrix; Multiple mineralization indicator elements that meet the correlation criteria are selected from the orthogonal correlation matrix; Based on the information contribution of all selected mineralization indicator elements, the selected mineralization indicator elements are classified into response data of mineralization of multiple mineral types.

4. The method as described in claim 3, characterized in that, Determining the multiple information contributions of each ore-forming indicator element in the geochemical dataset and the ore-forming correlations among the ore-forming indicator elements specifically includes: Logarithmically normalize the concentration data of each mineralization indicator element in the geochemical dataset to obtain the standard concentration data of each mineralization indicator element. Determine the covariance matrix between the concentrations of each element based on the standard concentration data of each mineralization indicator element; The mineralization correlation among the mineralization indicator elements is determined based on the covariance matrix among the concentrations of each element. The information contribution of each mineralization indicator element is determined based on the mineralization correlation among the various mineralization indicator elements.

5. The method as described in claim 1, characterized in that, Each response data point is decomposed into high-frequency data components directly related to the mineralization process and low-frequency data components related to the regional ore-controlling geological background. Specifically, these include: Select one response data as the selected response data, and integrate the selected response data into an initial distribution map of mineral resources within the target mineral cluster area; The initial distribution map is decomposed into multiple concentration data components; Based on the frequency of all concentration data components, all concentration data components are divided into high-frequency data components used to characterize the mineralization process and low-frequency data components used to characterize the regional ore-controlling geological background. The remaining response data were further decomposed into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional ore-controlling geological background.

6. The method as described in claim 5, characterized in that, Decomposing the initial distribution map into multiple concentration data components specifically includes: Determine the average trend surface of the concentration distribution in the initial distribution map; The process components are obtained by removing the average trend surface from the initial distribution map; If the process component does not meet the preset termination condition, the average trend surface of the process component is removed from the process component to obtain a new process component until the obtained process component meets the preset termination condition. The obtained process component is then used as a concentration data component. The concentration data component is removed from the initial distribution map to obtain a new initial distribution map. The above steps are repeated until the number of extreme points in the obtained initial distribution map is less than a preset extreme point threshold, thereby obtaining multiple concentration data components.

7. The method as described in claim 1, characterized in that, Determining the first critical threshold for various types of local mineralization anomalies based on the cumulative frequency characteristics of the high-frequency data components specifically includes: Extreme interference values ​​are removed from all high-frequency data components to obtain multiple denoised high-frequency components; Select a denoised high-frequency component as the selected high-frequency component, and group the selected high-frequency component according to logarithmic intervals to obtain multiple grouped data; Based on the cumulative frequency of each group of data, a cumulative frequency curve of the concentration distribution of ore-forming elements within the target ore cluster area is plotted. The first critical threshold of local mineralization anomaly data in the mineral type corresponding to the selected high-frequency component is determined by the cumulative frequency curve. Continue to determine the first critical threshold of the local mineralization anomaly data in the mineral type corresponding to the remaining denoised high-frequency components.

8. The method as described in claim 1, characterized in that, The second critical threshold for determining favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components specifically includes: Remove extreme interference values ​​from all low-frequency data components to obtain multiple denoised low-frequency components; The second critical threshold for favorable background data for regional mineralization was determined by statistical characteristics of the concentration data in all denoised low-frequency components.

9. The method as described in claim 1, characterized in that, The geochemical sampling data refers to the geochemical data of stream sediments collected within the target mineralization area.

10. A smart delineation system for mineral exploration target areas, the system comprising an anomaly component separation unit, characterized in that, The abnormal component separation unit includes: The acquisition module is used to acquire geochemical sampling data of the target mineral cluster area and combine all mineralization indicator elements in the geochemical sampling data into a geochemical dataset. The processing module is used to extract response data of mineralization of various mineral types from the geochemical dataset based on the information contribution of each mineralization indicator element in the geochemical dataset and the mineralization correlation between each mineralization indicator element as screening criteria. The processing module is also used to decompose each response data into high-frequency data components that characterize the mineralization process and low-frequency data components that characterize the regional ore-controlling geological background. The processing module is also used to determine the first critical threshold of local mineralization anomaly data of multiple types of minerals based on the cumulative frequency characteristics of the high-frequency data components, and to determine the second critical threshold of favorable background data for regional mineralization based on the statistical characteristics of the low-frequency data components. The execution module is used to quantitatively classify the mineralization-related anomaly components and background components based on all first and second critical thresholds.

Images