Prospect quantitative analysis method and device based on geoscience big data

Abstract

The application relates to a prospecting vision quantitative analysis method and device based on geosciences big data, and belongs to the technical field of mineral resources exploration. The method solves the problem that the level of a prospecting vision area is determined according to some subjective understanding in the existing prospecting prediction scheme, and the circled prospecting vision area is not accurate enough. In the method, the prospecting exploration marks are extracted from a geosciences big data spatial database, and the training area is divided into grid units according to the target deposit type, and a model unit is selected from the grid units. A prediction model is trained according to the prospecting exploration marks of the model unit, the metallogenic probability of each grid unit in the target working area and the prospecting vision area are calculated based on the prediction model, and the prospecting vision area is graded based on the metallogenic probability of each grid unit. Based on the method, the quantitative basis for the delineation of the prospecting vision area can be provided, the influence of subjective factors is avoided, and the delineation accuracy of the prospecting vision area is improved.

Description

Technical Field

[0001] This application relates to the field of mineral resource exploration technology, and in particular to a quantitative analysis method and system for mineral prospecting based on geoscience big data. Background Technology

[0002] Mineral exploration prediction refers to the methods and techniques used to identify potential mineral deposits by applying basic geological metallogenic theories, combining comprehensive information from geology, mineral resources, geophysical exploration, geochemical exploration, and remote sensing, summarizing predictive elements, and making comparative predictions. This guides exploration work and helps discover industrial mineral deposits.

[0003] In mineral exploration prediction, the mineralization probability is often used to characterize the mineralization potential of different grid units within a target exploration area. A threshold value of this mineralization probability is also frequently used to determine the boundary of the prospective exploration area and further define its grade. Traditional mineral exploration prediction schemes employ various calculation methods to determine a threshold value of the mineralization probability as the boundary of the prospective exploration area, such as using 85% of the cumulative probability or the mean plus twice the standard deviation. The grade of the prospective area is often determined subjectively, based on factors such as whether it contains discovered mineral deposits (points), whether it contains exploration markers deemed more important by the individual, or whether the exploration markers covering the prospective exploration area are complete.

[0004] However, the above mineral exploration prediction schemes either determine the boundaries of prospective mineral exploration areas based on the premise that the probability of mineralization (or its transformation) conforms to a normal distribution, or they rely too much on some subjective understanding to determine its level, resulting in the delineation of prospective mineral exploration areas not being accurate enough. Summary of the Invention

[0005] In view of the above analysis, the present invention aims to provide a method and apparatus for quantitative analysis of mineral exploration prospects based on geoscientific big data, in order to solve the problem that existing mineral exploration prediction schemes rely too much on subjective understanding to determine their level, resulting in inaccurate delineation of mineral exploration prospect areas.

[0006] This application provides a method for quantitative analysis of mineral exploration prospects based on geoscientific big data, including the following steps:

[0007] Mineral exploration indicators are extracted from geoscience big data spatial databases, and the training area is divided into grid units according to the target mineral deposit type. Model units are selected from the grid units. The model units are divided into mineralized units and non-mineralized units. The absence or negative anomaly of mineral exploration indicators is the bottom line selection criterion for non-mineralized units.

[0008] The prediction model is trained based on the mineral exploration indicators of the model units; the prediction model determines the boundary of the prospective mineral exploration area based on the set Youden index, in order to delineate the prospective mineral exploration area; the set Youden index is as follows:

[0009] P 最优= [Max(TPR / (FPR+ε))→P];

[0010] Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area, TPR represents the ratio of correctly predicted mineralized model units to all mineralized units, FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units, ε is a given value, and → represents setting the mapping relationship between the Youden index and the mineralization probability.

[0011] The mineralization probability and prospective mineralization areas of each grid cell within the target working area are calculated based on the prediction model, and the prospective mineralization areas are classified based on the mineralization probability of each grid cell.

[0012] This application's embodiment of the quantitative analysis method for mineral exploration prospects based on geoscientific big data extracts mineral exploration indicators from a geoscientific big data spatial database, divides the training area into grid cells according to the target mineral deposit type, and selects model cells from the grid cells. A prediction model is trained based on the mineral exploration indicators of the model cells. Based on the prediction model, the mineralization probability and mineral exploration prospects of each grid cell within the target working area are calculated, and the mineral exploration prospects are classified based on the mineralization probability of each grid cell. This provides a quantitative basis for delineating mineral exploration prospects, avoids the influence of subjective factors, and provides a data model foundation for improving delineation accuracy.

[0013] As one optional embodiment, the process of extracting mineral exploration indicators from a geoscience big data spatial database includes the following steps:

[0014] Prioritize the extraction of mineral exploration indicators based on ore-controlling elements;

[0015] Secondly, mineral exploration indicators are extracted based on the mineral assemblage of the ore.

[0016] Finally, mineral exploration markers are extracted based on altered mineral assemblages.

[0017] As one optional embodiment, the process of selecting model elements from mesh elements includes the following steps:

[0018] Prioritize grid cells from selected areas of the training region where no mineral deposits have been found as mineral-free cells;

[0019] Secondly, grid cells that cannot form the target deposit type or that have formed other deposit types that are opposed to the target deposit type are used as non-mineralized cells;

[0020] Finally, grid cells that are far from known mineral deposits or have missing or negative anomalies in mineral exploration indicators are considered as mineral-free cells.

[0021] As one optional embodiment, the process of classifying prospective mineral areas based on the mineralization probability of each grid cell is as follows:

[0022]

[0023] in, Let p be the mean probability of mineralization within the prospective mineralization area, n be the total number of grid cells within the prospective mineralization area, and p be the mean probability of mineralization within the prospective mineralization area. i Let be the mineralization probability of the i-th grid cell within the prospective mineral exploration area; classify the prospective mineral exploration area based on interval division of the mean.

[0024] As one optional embodiment, the process of establishing a geoscientific big data spatial database includes the following steps:

[0025] Acquire geoscientific big data, including geological, mineral, geophysical, geochemical, heavy mineral, and remote sensing data;

[0026] Transform geoscientific big data data from different coordinate systems into a unified spatial coordinate system, and convert various data formats in the geoscientific big data data into a unified data type format to generate a geoscientific big data spatial database.

[0027] As one optional embodiment, the process of calculating the mineralization probability of each grid cell within the target working area based on a prediction model includes the following steps:

[0028] Mineralization probability is calculated based on grid cell-based mineral exploration indicators; the prediction model is a random forest model.

[0029] As one optional embodiment, the method further includes the following steps:

[0030] Record the prospective mineral exploration areas and results of the target work area;

[0031] A mapping relationship between prospective mineral exploration areas and exploration results is established based on an AI model.

[0032] The correlation between the mean and the grading is adjusted based on the mapping relationship.

[0033] This application also provides a quantitative analysis device for mineral exploration prospects based on geoscience big data, including:

[0034] The data processing module is used to extract mineral exploration indicators from the geoscience big data spatial database, divide the training area into grid cells according to the target mineral deposit type, and select model cells from the grid cells. Among them, the model cells are divided into mineralized cells and non-mineralized cells, and the absence or negative anomaly of mineral exploration indicators is the bottom line selection criterion for non-mineralized cells.

[0035] The model training module is used to train a prediction model based on the unit mineral exploration indicators of the model. The prediction model determines the boundary of the prospective mineral exploration area based on a set Youden index, thus delineating the prospective mineral exploration area. The Youden index is set as follows:

[0036] P 最优 = [Max(TPR / (FPR+ε))→P];

[0037] Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area, TPR represents the ratio of correctly predicted mineralized model units to all mineralized units, FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units, ε is a given value, and → represents setting the mapping relationship between the Youden index and the mineralization probability.

[0038] The analysis and prediction module is used to calculate the mineralization probability and prospective mineralization areas of each grid cell within the target working area based on the prediction model, and to classify the prospective mineralization areas based on the mineralization probability of each grid cell.

[0039] The quantitative analysis device for mineral exploration prospects based on geoscientific big data in this application extracts mineral exploration indicators from a geoscientific big data spatial database, divides the training area into grid cells according to the target mineral deposit type, and selects model cells from the grid cells. A prediction model is trained based on the mineral exploration indicators of the model cells. Based on the prediction model, the mineralization probability and mineral exploration prospects of each grid cell within the target working area are calculated, and the mineral exploration prospects are classified based on the mineralization probability of each grid cell. This provides a quantitative basis for delineating mineral exploration prospects, avoids the influence of subjective factors, and provides a data model foundation for improving delineation accuracy.

[0040] At least one embodiment of this application also provides a data control device, including:

[0041] One or more memories that store computer-executable instructions non-transitory;

[0042] One or more processors are configured to run computer-executable instructions, wherein the computer-executable instructions are executed by the one or more processors to implement the quantitative analysis method for mineral prospecting based on geoscience big data according to any embodiment of the present application.

[0043] The aforementioned data control device, based on spatial data calculation rules, integrates multi-source land and space data to construct a standardized, data-driven minimum functional statistical standard unit division model. This model can generate minimum functional statistical standard units with unified spatial granularity that are adaptable to multi-scale planning objectives, significantly improving the spatial support accuracy and analytical application foundation for the implementation of land and space planning. This contributes to the scientific, dynamic, and intelligent governance of land and space.

[0044] At least one embodiment of this application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, implement a method for quantitative analysis of mineral prospecting prospects based on geoscientific big data according to any embodiment of this application.

[0045] The aforementioned non-transient computer-readable storage medium, based on spatial data calculation rules, integrates multi-source territorial spatial data to construct a standardized, data-driven minimum functional statistical standard unit division model. This model can generate minimum functional statistical standard units with unified spatial granularity that adapt to multi-scale planning objectives, significantly improving the spatial support accuracy and analytical application foundation for the implementation of territorial spatial planning, and contributing to the scientific, dynamic, and intelligent governance of territorial space. Attached Figure Description

[0046] Figure 1 A flowchart illustrating a quantitative analysis method for mineral exploration prospects based on geoscience big data, according to an embodiment of the application.

[0047] Figure 2 A schematic diagram of the quantitative analysis process for an example of quantitative delineation of a prospective mineral deposit of a certain type in a certain area, provided by the present invention;

[0048] Figure 3 This is a structural diagram of a mineral exploration prospect quantitative analysis device module based on geoscience big data, according to an embodiment of the application.

[0049] Figure 4 A schematic block diagram of a data control device provided by the present invention;

[0050] Figure 5 This is a schematic diagram of a non-transitory computer-readable storage medium provided by the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the described embodiments of this application without creative effort are within the scope of protection of this application.

[0052] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0053] To keep the following description of the embodiments of this application clear and concise, detailed descriptions of some known functions and components have been omitted.

[0054] This application provides a method for quantitative analysis of mineral exploration prospects based on geoscience big data.

[0055] Figure 1 Here is a flowchart of a quantitative analysis method for mineral exploration prospects based on geoscience big data, as described in one embodiment of the application. Figure 1 As shown, a quantitative analysis method for mineral exploration prospects based on geoscientific big data according to one embodiment of the application includes steps S100 to S102:

[0056] S100 extracts mineral exploration indicators from the geoscience big data spatial database, divides the training area into grid units for the target mineral deposit type, and selects model units from the grid units; among them, the model units are divided into mineralized units and non-mineralized units, and the absence or negative anomaly of mineral exploration indicators is the bottom line selection criterion for non-mineralized units.

[0057] S101, a prediction model is trained based on the mineral exploration indicators of the model units; wherein, the prediction model determines the boundary of the prospective mineral exploration area based on a set improved Youden index, so as to delineate the prospective mineral exploration area; the set improved Youden index is as follows:

[0058] P 最优 = [Max(TPR / (FPR+ε))→P];

[0059] Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area, TPR represents the ratio of correctly predicted mineralized model units to all mineralized units, FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units, ε is a given value, and → represents setting the improved Youden index and the mapping relationship between the mineralization probability.

[0060] S102, calculate the mineralization probability and prospective mineralization areas of each grid cell within the target working area based on the prediction model, and classify the prospective mineralization areas based on the mineralization probability of each grid cell within the prospective mineralization area.

[0061] In this embodiment, a training region and a target working region are provided. The training region is used to provide a training dataset for the prediction model, and the target working region is used to delineate and classify prospective mineral exploration areas using the prediction model. The data source for the training region is a geoscientific big data spatial database.

[0062] As one preferred embodiment, the training area is acquired including geological, mineral, geophysical, geochemical, heavy mineral, and remote sensing big data; the geoscience big data data with different coordinate systems are transformed into a unified spatial coordinate system; various data formats are converted into a unified data type format, and finally a geoscience big data spatial database is generated.

[0063] Preferably, in the extraction of mineral exploration markers, sequential extraction replaces the traditional targeted extraction, establishing a mineral exploration marker system. Selecting mineral exploration markers from geoscientific big data spatial databases is a crucial step in predicting mineral exploration in target deposit types. This application embodiment is based on the three-in-one characteristics of target deposit types—"ore-controlling factors, ore mineral assemblages, and alteration mineral assemblages"—and sequentially extracts mineral exploration markers from geoscientific big data spatial databases to construct a mineral exploration marker system. It should be noted that the sequential extraction is necessary due to the chronological logic of deposit formation. Ore-controlling factors are prerequisites for deposit formation, ore mineral assemblages are the result of deposit formation, and alteration mineral assemblages are associated geological phenomena. This application embodiment uses a porphyry-skarn-low-temperature hydrothermal vein-type gold-iron-copper-lead-zinc polymetallic deposit in a certain area as an example to illustrate the construction process of the mineral exploration marker system. It should be particularly noted that other target deposit types can also follow this approach, but the difference lies in the different characteristics of the target deposit types, which will lead to different mineral exploration marker systems.

[0064] As one preferred embodiment, the process of extracting mineral exploration indicators from a geoscience big data spatial database includes the following steps:

[0065] Prioritize the extraction of mineral exploration indicators based on ore-controlling elements;

[0066] Secondly, mineral exploration indicators are extracted based on the mineral assemblage of the ore.

[0067] Finally, mineral exploration markers are extracted based on altered mineral assemblages.

[0068] First, mineral exploration markers are extracted based on ore-controlling elements. Ore-controlling elements refer to the geological elements that control the formation of mineral deposits, mainly including strata, structures, and igneous rocks. Different target deposit types require different ore-controlling elements. Taking porphyry-skarn-low-temperature hydrothermal vein type deposits as an example, they are mainly controlled by porphyry intrusive bodies and carbonate strata. Therefore, porphyry bodies and carbonate strata from geological maps in the geoscience big data spatial database are extracted as igneous rock and stratigraphic markers for mineral exploration. In addition, this type of deposit is also controlled by intrusive contact structures. Therefore, intrusive contact structures from geological maps in the geoscience big data spatial database are extracted as structural markers for mineral exploration.

[0069] Secondly, prospecting indicators are extracted based on ore mineral assemblages. Ore mineral assemblages refer to the aggregates of useful minerals in an ore deposit (ore body). Different target deposit types have different ore mineral assemblages. Taking porphyry-skarn-low-temperature hydrothermal vein type deposits as an example, the ore mineral assemblages they produce mainly include pyrite, magnetite, chalcopyrite, galena, and sphalerite, etc. One key characteristic of magnetite is its strong magnetism compared to the surrounding rock. Therefore, magnetic anomalies in geochemical data extracted from geochemical databases can be used as geophysical indicators for mineral exploration. Ore-forming elements such as gold, copper, lead, and zinc, which contain pyrite, chalcopyrite, galena, and sphalerite, can be relatively enriched in stream sediments due to surface weathering. Therefore, anomalies of these elements can be extracted from geochemical data in geochemical databases as geochemical indicators for mineral exploration. Furthermore, gold is often very stable in its supergene state, forming natural heavy minerals. Therefore, gold-bearing natural heavy minerals can be extracted from natural heavy mineral data in geochemical databases as indicators of natural heavy minerals for mineral exploration.

[0070] Finally, mineral exploration markers are extracted based on alteration mineral assemblages. An alteration mineral assemblage refers to a specific mineral aggregate formed after the exchange of materials between the protolith and fluids under hydrothermal action. Different target deposit types have different alteration mineral assemblages. Taking a porphyry-skarn-low-temperature hydrothermal vein type deposit as an example, during the formation of the deposit, hydrothermal alteration also forms water- and iron-rich alteration minerals (such as sericite, chlorite, epidote, etc.). Therefore, hydroxyl groups or iron stains representing alteration minerals can be extracted from remote sensing data (such as ETM or ASTER) in geoscientific big data spatial databases as remote sensing markers for mineral exploration.

[0071] By integrating the above-mentioned mineral exploration indicators, a multi-dimensional exploration indicator system was successfully constructed.

[0072] Based on dividing the training area into grid cells according to the target ore deposit type, a sufficient number of model cells need to be selected for training the prediction model. Model cells include ore-bearing cells and ore-free cells. The selection method for ore-bearing cells is relatively mature, that is, selecting grid cells where ore deposits have been discovered through detailed exploration as ore-bearing cells. If there are not enough ore-bearing cells, the number of ore-bearing cells can be increased through various expansion methods, such as buffer analysis or conditional generative adversarial networks.

[0073] However, selecting mineral-free units is a complex task. Traditionally, only grid cells that have undergone detailed exploration and yielded no mineral deposits can be identified as mineral-free units. However, such grid cells are very limited and often fail to meet the quantity requirements of prediction models for mineral-free units, necessitating a broader approach to selecting them. This application proposes a "three-level selection method" to select mineral-free units, including the following steps:

[0074] Prioritize grid cells from selected areas of the training region where no mineral deposits have been found as mineral-free cells;

[0075] Secondly, grid cells that cannot form the target deposit type or that have formed other deposit types that are opposed to the target deposit type are used as non-mineralized cells;

[0076] Finally, grid cells that are far from known mineral deposits or have missing or negative anomalies in mineral exploration indicators are considered as mineral-free cells.

[0077] The first level prioritizes grid cells with no discovered mineral deposits in areas with a high level of exploration (generally reaching the level of general survey or above) as the mineral-free cells in the model.

[0078] The second level, when the first level cannot meet the requirement of having enough mineral-free units, selects grid units that are geologically considered unlikely to form the target deposit type (e.g., sedimentary mineral deposits cannot form in granite areas) or grid units containing other deposit types that are opposed to the target deposit type (if a basic-ultrabasic rock type deposit has already formed, it is impossible for a granite-related deposit to form again) as mineral-free units in the model unit.

[0079] At the third level, if the number of non-mineralized units cannot be met even at the second level, non-mineralized units can be selected based on the fundamental fact that the formation of mineral deposits is of low probability and their output exhibits clustering phenomena. These units are located in grid cells far from known mineral deposits or overlapping areas where all mineral exploration indicators are missing (or have negative anomalies).

[0080] Based on this, mineralized and non-mineralized units are labeled, and the prediction model is trained. Specifically, the optimal random forest prediction model is selected and trained. There are various methods for calculating the probability of mineralization, such as feature analysis, weighted evidence method, logistic regression, ensemble learning, and deep learning. This embodiment uses the random forest algorithm, based on the mineral exploration indicators of the model units, for training and optimization.

[0081] Simultaneously, a threshold (optimal mineralization probability) is determined based on a modified Youden index to delineate prospective mineral exploration areas. After obtaining the optimal model, a reasonable threshold is determined as the boundary value of the prospective mineral exploration area. This application proposes using a modified Youden index to determine the boundary (threshold) of the prospective mineral exploration area and delineate it. The formula for setting the modified Youden index is as follows:

[0082] P 最优 = [Max(TPR / (FPR+ε))→P];

[0083] Among them, P 最优 This refers to determining the optimal mineralization probability of the prospective exploration area boundary (threshold). TPR is the true positive rate, specifically representing the ratio of correctly predicted mineralized model units to all mineralized model units in this embodiment. FPR is the false positive rate, the ratio of incorrectly predicted mineralized model units to all non-mineralized units. ε is a very small value, typically 0.001 to prevent the denominator from being zero. Max(TPR / (FPR+ε)) is the improved Youden index, P is the mineralization probability corresponding to the improved Youden index, and → represents the mapping relationship between the improved Youden index and the mineralization probability. The mineralization probability corresponding to the maximum value of the ratio of true positive rate to false positive rate is the optimal threshold for determining the prospective exploration area boundary. Setting the improved Youden index to a range of 1 to +∞ further highlights the Youden index, preventing the difficulty in determining the corresponding optimal mineralization probability when the traditional Youden index value changes slightly. This amplifies the difference between the sequential extraction system of mineral exploration markers and the fine acquisition of non-mineralized units, thereby more accurately determining the prospective exploration area boundary.

[0084] As one optional embodiment, the process of calculating the mineralization probability of each grid cell within the target working area using a predictive model includes the following steps:

[0085] Mineralization probability is calculated based on grid cell-based mineral exploration indicators; the prediction model is a random forest model.

[0086] Prospective mineral exploration areas are delineated within the target working area. Based on the obtained random forest model, the mineralization probability of all grid cells within the target working area is calculated according to the mineral exploration indicators of the grid cells, and prospective mineral exploration areas are delineated based on the optimal mineralization probability.

[0087] Mineral exploration areas are classified based on the average mineralization probability, with the level representing the quality of mineralization conditions. However, after delineating mineral exploration areas, since each area consists of multiple grid units, and each grid unit often has a different mineralization probability value, there is a lack of a unified indicator for determining the level of the mineral exploration area. Therefore, this application uses the average mineralization probability of all grid units within the mineral exploration area as the quantitative basis for classifying the level of the exploration area, and it is also a quantitative indicator for selecting priority exploration targets. The calculation formula is as follows:

[0088]

[0089] in, Let p be the mean probability of mineralization within the prospective mineralization area, n be the total number of grid cells within the prospective mineralization area, and p be the mean probability of mineralization within the prospective mineralization area. i Let represent the mineralization probability of the i-th grid cell within the prospective mineral exploration area.

[0090] This involves dividing the area into multiple intervals based on the range in which the mean value falls, with each interval corresponding to a level of prospective mineral exploration area. In actual engineering, the mean value can be divided into a set number of levels based on the granularity of mineral exploration; this is not limited here.

[0091] As one optional embodiment, the method further includes the following steps:

[0092] Record the prospective mineral exploration areas and results of the target work area;

[0093] A mapping relationship between prospective mineral exploration areas and exploration results is established based on an AI model.

[0094] The correlation between the mean and the grading is adjusted based on the mapping relationship.

[0095] Based on historical analysis of prospective mineral exploration areas and exploration results, a mapping relationship between prospective mineral exploration areas and exploration results is established. The correlation between the mean and the classification is adjusted. The number of classification levels is adjusted through an AI model to quantify the accuracy standard of classification, which facilitates the integrated and automated deployment of quantitative analysis data models for prospective mineral exploration areas.

[0096] Preferably, the prospective mineral exploration area and the results of mineral exploration are input into the AI ​​model as graphic results, and a mapping relationship is established based on the similarity of the graphics.

[0097] Figure 2 This invention provides a schematic diagram of the quantitative analysis process for an example of quantitative delineation of a prospective mineral deposit of a certain type in a certain region, as shown in the following figure. Figure 2 As shown, this invention illustrates the quantitative delineation of prospective mineral exploration areas based on grading. Taking a specific mineral deposit type in a certain area as an example, the labeled map of the training area (real mineralized units and non-mineralized units) is shown below. Figure 2Figure a in the diagram; the mineralization probability of the training region is predicted based on the optimal random forest model, see [link / reference]. Figure 2 Figure b in the diagram; the improved Youden index as a function of mineralization probability, obtained from the label map and mineralization probability map of the training area, can be found in [reference]. Figure 2 In Figure c, the maximum Yangen index of 603.3777 corresponds to a mineralization probability of 0.9686. 0.9686 is the optimal mineralization probability and can be used as a threshold for delineating prospective mineral exploration areas. Based on the optimal random forest model, predictions are made in the target prediction area, and the calculated mineralization probability map is shown in [reference]. Figure 2 The d-plot in the diagram; the prospective mineral exploration area delineated based on the optimal mineralization probability and the mineralization probability map of the target prediction area, see [reference]. Figure 2 Figure e in the diagram; the mean value of the mineralization probability of all grid cells within each prospective mineralization area of ​​the target prediction region is determined by calculation, see [reference]. Figure 2 The f-plot in the diagram; based on the above averages, using 0.98 as the boundary, and determining the level according to the magnitude of the average, areas with a mineralization probability greater than or equal to 0.98 are classified as Class A prospective areas, and those less than 0.98 are classified as Class B prospective areas. See [reference needed]. Figure 2 The g-graph in the image.

[0098] It is important to clarify that the primary purpose of classifying prospective areas is to facilitate further mineral exploration, prioritizing limited human, financial, and time resources in the most promising regions. Relatively speaking, Category A prospective areas have a higher probability of mineralization and better prospects than Category B areas, so exploration is generally prioritized in Category A areas. However, this does not mean that Category B prospective areas have no prospective potential. If there are still remaining human, financial, and time resources after the Category A areas have been completed, exploration can continue in Category B areas. Therefore, the distinction between Category A and Category B is relative.

[0099] This application's embodiment of the quantitative analysis method for mineral exploration prospects based on geoscientific big data extracts mineral exploration indicators from a geoscientific big data spatial database, divides the training area into grid cells according to the target mineral deposit type, and selects model cells from the grid cells. A prediction model is trained based on the mineral exploration indicators of the model cells. Based on the prediction model, the mineralization probability and mineral exploration prospects of each grid cell within the target working area are calculated, and the mineral exploration prospects are classified based on the mineralization probability of each grid cell. This provides a quantitative basis for delineating mineral exploration prospects, avoids the influence of subjective factors, and provides a data model foundation for improving delineation accuracy.

[0100] This application also provides a device for quantitative analysis of mineral exploration prospects based on geoscience big data.

[0101] Figure 3This is a structural diagram of a module for a quantitative analysis device for mineral exploration prospects based on geoscience big data, as described in one embodiment of the application. Figure 3 As shown, an embodiment of the application includes a quantitative analysis device for mineral exploration prospects based on geoscientific big data, comprising:

[0102] The data processing module 100 is used to extract mineral exploration indicators from the geoscience big data spatial database, divide the training area into grid units according to the target mineral deposit type, and select model units from the grid units; among them, the model units are divided into mineralized units and non-mineralized units, and the absence or negative anomaly of mineral exploration indicators is the bottom line selection criterion for non-mineralized units.

[0103] Model training module 101 is used to train a prediction model based on the mineral exploration indicators of the model units; wherein, the prediction model determines the boundary of the prospective mineral exploration area based on a set Youden index, so as to delineate the prospective mineral exploration area; the improved Youden index is set as follows:

[0104] P 最优 = [Max(TPR / (FPR+ε))→P];

[0105] Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area, TPR represents the ratio of correctly predicted mineralized model units to all mineralized units, FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units, ε is a given value, and → represents setting the mapping relationship between the Youden index and the mineralization probability.

[0106] The analysis and prediction module 102 is used to calculate the mineralization probability and prospective mineralization areas of each grid cell within the target working area based on the prediction model, and to classify the prospective mineralization areas based on the mineralization probability of each grid cell.

[0107] The quantitative analysis device for mineral exploration prospects based on geoscientific big data in this application extracts mineral exploration indicators from a geoscientific big data spatial database, divides the training area into grid cells according to the target mineral deposit type, and selects model cells from the grid cells. A prediction model is trained based on the mineral exploration indicators of the model cells. Based on the prediction model, the mineralization probability and mineral exploration prospects of each grid cell within the target working area are calculated, and the mineral exploration prospects are classified based on the mineralization probability of each grid cell. This provides a quantitative basis for delineating mineral exploration prospects, avoids the influence of subjective factors, and provides a data model foundation for improving delineation accuracy.

[0108] At least one embodiment of this application also provides a data control device. Figure 4 This is a schematic block diagram of a data control device provided for at least one embodiment of this application. For example, such as... Figure 4As shown, the data control device 20 may include one or more memories 200 and one or more processors 201. The memories 200 are used to store computer-executable instructions non-transitory; the processors 201 are used to run the computer-executable instructions, which, when run by the processors 201, can cause the processors 201 to perform one or more steps in the quantitative analysis method for mineral prospecting based on geoscientific big data according to any embodiment of this application.

[0109] The specific implementation and explanation of each step of the quantitative analysis method for mineral exploration prospects based on geoscientific big data can be found in the relevant content of the above-mentioned embodiment of the quantitative analysis method for mineral exploration prospects based on geoscientific big data, and will not be repeated here. It should be noted that... Figure 4 The components of the data control device 20 shown are merely exemplary and not limiting. The data control device 20 may have other components depending on the actual application requirements.

[0110] In one embodiment, the processor 201 and the memory 200 can communicate directly or indirectly with each other. For example, the processor 201 and the memory 200 can communicate via a network connection. The network can include a wireless network, a wired network, and / or any combination of wireless and wired networks; this application does not limit the type and function of the network. Alternatively, the processor 201 and the memory 200 can also communicate via a bus connection. The bus can be a Peripheral Component Interconnect Standard (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. For example, the processor 201 and the memory 200 can be located at a remote data server (cloud) or a distributed energy system (local), or at a client (e.g., a mobile device such as a mobile phone). For example, the processor 201 can be a central processing unit (CPU), a tensor processor (TPU), or a graphics processing unit (GPU), etc., with data processing and / or instruction execution capabilities, and can control other components in the data control device 20 to perform desired functions. The central processing unit (CPU) can be an x86 or ARM architecture, etc.

[0111] In one embodiment, memory 200 may include any combination of one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, erasable programmable read-only memory (EPROM), portable compact disc read-only memory (CD-ROM), USB memory, flash memory, etc. One or more computer-executable instructions may be stored on the computer-readable storage medium, and processor 201 may execute these computer-executable instructions to implement various functions of data control device 20. Various application programs and various data, as well as various data used and / or generated by the application programs, may also be stored in memory 200.

[0112] It should be noted that the data control device 20 can achieve similar technical effects to the aforementioned quantitative analysis method for mineral prospecting based on geoscience big data, and the repetitions will not be repeated.

[0113] At least one embodiment of this application also provides a non-transitory computer-readable storage medium. Figure 5 This is a schematic diagram of a non-transitory computer-readable storage medium provided for at least one embodiment of this application. For example, such as... Figure 5 As shown, one or more computer-executable instructions 301 may be stored non-transitory on the non-transitory computer-readable storage medium 30. For example, when the computer-executable instructions 301 are executed by a computer, the computer may perform one or more steps in the quantitative analysis method for mineral prospecting prospects based on geoscientific big data according to any embodiment of this application.

[0114] In one embodiment, the non-transitory computer-readable storage medium 30 can be applied to the data control device 20 described above, for example, it can be the memory 200 in the data control device 20.

[0115] In one embodiment, the description of the non-transitory computer-readable storage medium 30 can be found in the description of the memory 200 in the embodiment of the data control device 20, and will not be repeated hereafter.

[0116] It should be noted that the memory 200 stores different non-transient computer-executable instructions, and the data control device 20 corresponds to the firmware upgrade device. When the computer-executable instructions are run by the processor 201, the processor 201 can perform one or more steps in the quantitative analysis method for mineral prospecting based on geoscience big data according to any embodiment of this application.

[0117] The following points should be noted regarding this application:

[0118] (1) The accompanying drawings of the embodiments of this application only involve the structures involved in the embodiments of this application. Other structures can be referred to the general design.

[0119] (2) For clarity, the thickness and dimensions of layers or structures are enlarged in the accompanying drawings used to describe embodiments of the invention. It will be understood that when an element such as a layer, film, region, or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element, or there may be intermediate elements present.

[0120] (3) Where there is no conflict, the embodiments and features in the embodiments of this application can be combined with each other to obtain new embodiments. The above are only specific implementations of this application, but the protection scope of this application is not limited thereto, and the protection scope of this application shall be determined by the protection scope of the claims.

[0121] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0122] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A quantitative analysis method for mineral exploration prospects based on geoscientific big data, characterized in that, Including the following steps: Mineral exploration indicators are extracted from the geoscience big data spatial database, and the training area is divided into grid units according to the target mineral deposit type. Model units are selected from the grid units. The model units are divided into mineralized units and non-mineralized units. The absence or negative anomaly of the mineral exploration indicators is the bottom line selection criterion for the non-mineralized units. A prediction model is trained based on the mineral exploration indicators of the model unit; wherein, the prediction model determines the boundary of the prospective mineral exploration area based on a set Youden index, so as to delineate the prospective mineral exploration area; the set Youden index is as follows: P 最优 =[Max(TPR / (FPR+ε))→P]; Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area; TPR represents the ratio of correctly predicted mineralized model units to all mineralized units; FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units; ε is a given value; → represents setting the mapping relationship between the Youden index and the mineralization probability; Based on the prediction model, the mineralization probability and prospective mineralization areas of each grid cell within the target working area are calculated, and the prospective mineralization areas are classified based on the mineralization probability of each grid cell. The process of extracting mineral exploration indicators from a geoscience big data spatial database includes the following steps: Prioritize the extraction of mineral exploration indicators based on ore-controlling elements; Secondly, mineral exploration indicators are extracted based on the mineral assemblage of the ore. Finally, mineral exploration markers are extracted based on altered mineral assemblages.

2. The method for quantitative analysis of mineral exploration prospects based on geoscientific big data according to claim 1, characterized in that, The process of selecting model cells from the mesh cells includes the following steps: Grid cells from selected areas of the training region where no mineral deposits were found are preferentially designated as mineral-free cells. Secondly, grid cells that cannot form the target deposit type or that have formed other deposit types that are opposed to the target deposit type are used as non-mineralized cells; Finally, grid cells that are far from known mineral deposits or have missing or negative anomalies in mineral exploration indicators are considered as mineral-free cells.

3. The quantitative analysis method for mineral exploration prospects based on geoscientific big data according to claim 1 or 2, characterized in that, The process of classifying prospective mineral areas based on the mineralization probability of each grid cell is as follows: in, Let p be the mean probability of mineralization within the prospective mineralization area, n be the total number of grid cells within the prospective mineralization area, and p be the mean probability of mineralization within the prospective mineralization area. i The mineralization probability of the i-th grid cell within the prospective mineral exploration area is given; the prospective mineral exploration area is classified into different levels based on the interval division of the mean value.

4. The quantitative analysis method for mineral exploration prospects based on geoscientific big data according to claim 1, characterized in that, The process of establishing the aforementioned geoscientific big data spatial database includes the following steps: Acquire geoscientific big data, including geological, mineral, geophysical, geochemical, heavy mineral, and remote sensing data; Transform geoscientific big data data from different coordinate systems into a unified spatial coordinate system, and convert various data formats in the geoscientific big data data into a unified data type format to generate a geoscientific big data spatial database.

5. The method for quantitative analysis of mineral exploration prospects based on geoscientific big data according to claim 1, characterized in that, The process of calculating the mineralization probability of each grid cell within the target working area based on the prediction model includes the following steps: Mineralization probability is calculated based on grid cell-based mineral exploration indicators; wherein, the prediction model is a random forest model.

6. The method for quantitative analysis of mineral exploration prospects based on geoscientific big data according to claim 4, characterized in that, It also includes the following steps: Record the prospective mineral exploration areas and results of the target work area; A mapping relationship between the prospective mineral exploration area and the mineral exploration results is established based on an AI model; The correlation between the mean and the grading is adjusted based on the aforementioned mapping relationship.

7. A quantitative analysis device for mineral exploration prospects based on geoscientific big data, characterized in that, include: The data processing module is used to extract mineral exploration indicators from the geoscience big data spatial database, divide the training area into grid cells according to the target mineral deposit type, and select model cells from the grid cells; wherein, the model cells are divided into mineralized cells and non-mineralized cells, and the absence or negative anomaly of the mineral exploration indicators is the bottom line selection criterion for the non-mineralized cells; The model training module is used to train a prediction model based on the mineral exploration indicators of the model units; wherein, the prediction model determines the boundary of the prospective mineral exploration area based on a set Youden index, so as to delineate the prospective mineral exploration area; the set Youden index is as follows: P 最优 =[Max(TPR / (FPR+ε))→P]; Among them, P 最优 To determine the optimal mineralization probability at the boundary of a prospective mineralization area, TPR represents the ratio of correctly predicted mineralized model units to all mineralized units, FPR represents the ratio of incorrectly predicted mineralized model units to all non-mineralized units, ε is a given value, and → represents setting the mapping relationship between the Youden index and the mineralization probability. The analysis and prediction module is used to calculate the mineralization probability and prospective mineralization areas of each grid cell within the target working area based on the prediction model, and to classify the prospective mineralization areas based on the mineralization probability of each grid cell. The process of extracting mineral exploration indicators from a geoscience big data spatial database includes the following steps: Prioritize the extraction of mineral exploration indicators based on ore-controlling elements; Secondly, mineral exploration indicators are extracted based on the mineral assemblage of the ore. Finally, mineral exploration markers are extracted based on altered mineral assemblages.

8. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, implement the quantitative analysis method for mineral prospecting based on geoscience big data as described in any one of claims 1 to 6.

9. A data control device, characterized in that, include: One or more memories that store computer-executable instructions non-transitory; One or more processors configured to run computer-executable instructions, wherein the computer-executable instructions are executed by the one or more processors to implement the quantitative analysis method for mineral prospecting based on geoscience big data as described in any one of claims 1 to 6.

Images