A method for constructing a multi-scale exploration identification system guided by a mineralization system

Abstract

The application discloses a kind of multi-scale exploration identification system construction method guided by metallogenic system, and the ordered evolution chain of key metallogenic process is established for target metallogenic system;Then each key process is decomposed into component process;Then at least three hierarchical metallogenic unit scale framework is established;Under the framework, the indication element extraction of scale coupling is realized by the preset scale function conversion rule, so that the same geological element obtains different functional roles at different scale levels;Then determine the exploration identification corresponding to each indication element and its exploration method and quantitative threshold;Finally, the conversion result is constrained and verified using the domain knowledge rule set.The application realizes the systematic and computable conversion from metallogenic theory to exploration practice, significantly improves the scientificity, integrity and operability of exploration identification system.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of mineral exploration and geoscience information engineering, specifically to a method for constructing a multi-scale exploration identification system guided by metallogenic systems. Background Technology

[0002] Mineral resource exploration is a crucial foundation for national economic development. Given the current global shift in mineral exploration towards deeper and overburdened areas, how to scientifically and systematically construct an exploration identification system to guide mineral prospecting has become a critical issue that urgently needs to be addressed in the field of mineral exploration.

[0003] Metallogenic system theory provides an important theoretical framework for mineral exploration. In existing technologies, scholars have proposed a four-step translation method to transform the concept of metallogenic systems into exploration target area location: starting with critical processes, then sequentially transforming them into constituent processes, targeting elements, and targeting criteria. This four-level transformation framework provides conceptual guidance for the construction of an exploration identification system.

[0004] However, the existing four-level transformation framework has the following shortcomings: First, there is a lack of scale coupling mechanisms. Existing frameworks perform transformations at a single scale, failing to consider the systematic changes in the functional roles of the same geological element at different exploration scales. For example, tectonic elements primarily function as ore-controlling channels at the regional scale, but transform into ore-hosting spaces at the deposit scale; stratigraphic elements provide basement source area information at the regional scale, but act as mineralization reaction media at the deposit scale. Existing frameworks cannot handle such scale-dependent functional transformations.

[0005] Second, there is a lack of systematic transformation rules. The existing framework only proposes the concept of transformation, but there are no clear constraints and criteria between transformations at each level. The transformation process relies entirely on the implicit experience of individual experts, and different people performing the same transformation may produce drastically different results, lacking repeatability and transferability.

[0006] Third, there is a lack of a quantitative evaluation mechanism for the importance of exploration markers. The existing framework treats all exploration markers equally, failing to distinguish which markers are more distinguishable in terms of mineralization, and also failing to differentiate the importance of markers based on the geological background differences of different metallogenic units.

[0007] Fourth, the transformation chain is a one-way decomposition process, lacking an aggregation mechanism that traces back from observational evidence to geological judgment, and thus cannot achieve closed-loop verification of the identification system from theory to practice.

[0008] Therefore, this invention provides a method for constructing a multi-scale exploration identification system guided by ore-forming systems, aiming to solve the technical problems of existing four-level transformation frameworks lacking scale coupling mechanisms, missing transformation rules, inability to quantify identification importance, and lack of aggregation loops. Summary of the Invention

[0009] The purpose of this invention is to provide a method for constructing a multi-scale exploration identification system guided by ore-forming systems. The method includes the following steps: S1. Establish the metallogenic process evolution chain: For the target metallogenic system, identify the key metallogenic processes and establish an ordered evolution chain between each key metallogenic process according to the geological time sequence. The ordered evolution chain represents the sequence of complete dynamic processes of the metallogenic system from the formation of the source region to the preservation of the mineralization process. S2, Decomposition and Composition Process: Decompose the key mineralization process in S1 into composition processes, correspond each composition process to a specific geodynamic event and physicochemical action, and establish the direct subordinate relationship between the composition process and the key mineralization process. S3. Establish a multi-level metallogenic unit scale framework: Establish a multi-level hierarchical metallogenic unit framework, with strict hierarchical subordinate relationships between metallogenic units at each level. The metallogenic unit framework provides a multi-scale spatial reference system for subsequent transformations. S4: Extraction of indicator elements by scale coupling: Under the multi-level metallogenic unit scale framework, indicator elements are extracted from the composition process for each level of metallogenic unit; among them, the same geological element is assigned different functional roles at different scale levels according to the preset scale function conversion rules, and the indicator elements are associated with the composition process and metallogenic unit. The scale-function conversion rules in S4 include: a dual-function conversion rule for tectonic elements, where tectonic elements primarily function as conduits at the metallogenic domain, metallogenic province, and metallogenic belt scales, but transform into ore-hosting space functions at the ore field, deposit, and ore body scales; a triple-role conversion rule for stratigraphy, where stratigraphic elements function as source region modulation as basement at the metallogenic domain and metallogenic province scales, as situ modulation as ore-hosting space at the metallogenic belt and ore field scales, and as mineralization modulation as reaction medium at the deposit and ore body scales; and a progressive rule for age-constrained precision, where temporal resolution increases as the metallogenic unit scale decreases, progressing from epoch-level precision to instantaneous metallogenic event precision. S5: Determine exploration identifiers: For each indicator element, determine the corresponding exploration identifier. Each exploration identifier is associated with at least one exploration method and corresponding quantitative thresholds and discrimination criteria. S6: Rule Constraints and Verification: The conversion results from S1 to S5 are constrained and verified using a preset domain knowledge rule set. The domain knowledge rule set includes constraints on the functional roles of geological elements, the definition of spatiotemporal constraint ranges, and the evaluation of evidence strength. S61. Construct a domain knowledge rule set: This rule set consists of at least one of the following: scale function transformation rules, temporal constraint rules, mineral assemblage succession rules, and spatial geometric constraint rules. S62. Apply rule checks to the conversion results from S1 to S5 one by one: assign a functional role value to each indicator element and check whether it conforms to the scale function conversion rule; check whether the temporal relationship of the mineralization process meets the temporal constraint rule; check whether the alteration mineral assemblage conforms to the mineral assemblage succession rule; check whether the spatial distribution of the mineralization body meets the spatial geometric constraint rule; if a conversion result that violates the rules is found, mark it as abnormal and perform manual review or automatic correction.

[0010] Furthermore, the specific steps for establishing the ore-forming process evolution chain in S1 are as follows: S11. Based on the geological background data and existing research results of the target metallogenic system, identify the key geodynamic processes and material evolution links directly related to mineralization. S12. Based on geochronological data and the chronological superposition of geological events, the key mineralization processes are sorted in order of geological time from early to late. S13. Using a directed linked list data structure, the key mineralization processes are linked together into an ordered evolution chain. Each node in the chain represents a key mineralization process. The directed edges between nodes represent the temporal progression relationship and the causal driving relationship. The ordered evolution chain covers the complete dynamic process sequence from the formation of the mineral source area, magma migration and emplacement, fluid exsolution and migration, alteration and mineralization to post-mineralization preservation and transformation.

[0011] Furthermore, the specific steps of S2 are as follows: S21. For each key mineralization process, based on the geodynamic sequence and physicochemical reaction mechanism, it is decomposed into several subprocesses that are distinguishable in time and space. Each subprocess is a component process. The decomposition of component processes follows the following principles: each component process should correspond to an independently identifiable geodynamic event or physicochemical action; a one-to-many tree-like hierarchical relationship is established between each component process and the key mineralization process to which it belongs, that is, a key mineralization process contains multiple component processes, and each component process belongs to only one key mineralization process.

[0012] Furthermore, the multi-level metallogenic unit scale framework in S3 includes five levels: metallogenic domain, metallogenic province, metallogenic belt, ore field, ore deposit, and ore body. These levels have a strict hierarchical relationship from top to bottom. The typical area of ​​a metallogenic domain is on the order of millions of square kilometers, a metallogenic province on the order of hundreds of thousands of square kilometers, a metallogenic belt on the order of tens of thousands of square kilometers, an ore field on the order of hundreds to thousands of square kilometers, and ore deposits and ore bodies on the order of square kilometers to sub-square kilometers. The specific steps are as follows: S31. Based on the spatial distribution range of the target metallogenic system and the hierarchical nature of geological ore-controlling factors, determine the number of metallogenic unit levels, with at least three levels. S32. Metallogenic units are divided into levels from top to bottom. The division of each level of metallogenic unit is based on the geological criteria of structural framework, metallogenic geological background, consistency of ore source, area scale, and spatial nesting relationship. Strict hierarchical subordinate constraints are established between metallogenic units at each level. That is, the metallogenic unit at the next higher level completely contains the metallogenic unit at the next lower level in space, and the metallogenic unit at the next lower level must not cross the boundary of the metallogenic unit at the next higher level.

[0013] Furthermore, the specific steps of S4 are as follows: S41. Under the multi-level metallogenic unit scale framework, for each level of metallogenic unit, traverse each component process determined in S2, determine the functional role that the geological elements involved in the component process should play at the current scale level according to the preset scale function conversion rules, and extract the corresponding indicator elements accordingly. S411, whereby an association is established between the indicator element and the composition process and ore-forming unit: the source composition process and scale level of each indicator element are recorded in the form of a ternary, forming a structured association index of the indicator element, so that the indicator element can be traced back to its corresponding composition process and ore-forming unit. The ternary includes the composition process, the ore-forming unit level, and the indicator element.

[0014] Furthermore, the specific steps of S5 are as follows: S51. For each indicator element, based on its geological attributes and observability, determine the exploration identifier for detecting the indicator element under field and laboratory conditions; the determination of each exploration identifier must be associated with the following information: at least one structured exploration method and the corresponding quantification threshold and discrimination criteria. The quantification threshold is determined based on the statistical data and geochemical benchmark values ​​of known mineral deposits, and its accuracy increases as the scale of the metallogenic unit decreases, progressing step by step from existence criteria, qualitative classification criteria, semi-quantitative interval criteria to continuous numerical quantification thresholds; S511. The structured exploration method includes the definition of input data, operators, and output results. Operators include feature operators bound to a single exploration identifier and comprehensive operators that dynamically combine the detection results of exploration identifiers. The specific definition of feature operators is as follows: for each exploration identifier, the source of input data, data processing rules, and output result format of the identifier are clearly defined, thereby solidifying the complete extraction process from the original exploration data to the detection result of the exploration identifier into a reproducible rule definition. The specific definition of comprehensive operators is as follows: the input is a layer of detection results of multiple exploration identifiers, the processing rules are weighted superposition, evidence weight accumulation, and machine learning fusion algorithms, and the output is a comprehensive evaluation layer of combined identifiers. Comprehensive operators include evidence weight method, comprehensive information content method, random forest model, and deep learning model.

[0015] Furthermore, the detection results of the exploration markers in S511 are aggregated hierarchically. This hierarchical aggregation corresponds to the decomposition hierarchy of S1 to S5, including indicator feature level combinations, component process level combinations, key process level combinations, and prediction level combinations. The specific steps are as follows: S511.1, the indicator element level combination performs weighted fusion of exploration marker detection results under the same indicator element, with the weights determined based on the information content and evidence weight of each marker; the component process level combination fuses cross-indicator results associated with the same component process using the comprehensive information content method and evidence weight method to generate a component process level comprehensive evaluation; the key process level combination fuses cross-component process results associated with the same key mineralization process to generate a key process level comprehensive evaluation; the prediction level combination performs global nonlinear fusion of all key mineralization process level results using a random forest model or a deep learning model to output the final mineral exploration target area probability map.

[0016] Furthermore, the quantitative precision of exploration markers in S5 increases as the scale of metallogenic units decreases: at the scale from metallogenic domain to metallogenic province, exploration markers adopt existence criteria or qualitative classification criteria; at the scale from metallogenic belt to ore field, semi-quantitative classification criteria or quantitative interval criteria are adopted; at the scale of ore deposit / ore body, continuous numerical quantification thresholds are adopted.

[0017] The present invention has the following beneficial effects: The present invention establishes an ordered evolution chain of key mineralization processes and couples it with a multi-level mineralization unit scale framework, thereby realizing a systematic and structured transformation from mineralization theory to exploration practice, so that the construction of the exploration identification system no longer depends entirely on the implicit experience of individual experts, and has repeatability and transferability.

[0018] This invention introduces a scale-function conversion rule, which enables the same geological element to be assigned differentiated functional roles that conform to geological realities at different exploration scales, so that the exploration identification system can adapt to the full-scale exploration needs from regional area selection to ore body location.

[0019] This invention establishes a hierarchical aggregation mechanism corresponding to the decomposition chain, thereby forming a complete closed loop from theoretical decomposition to evidence aggregation. This enables the exploration identification system to not only guide exploration deployment but also verify the rationality of mineralization theory judgments through evidence backtracking.

[0020] This invention establishes a mechanism for the quantitative accuracy of identification to increase with scale, and forms a dual progression with the time-constrained accuracy progression. This enables the exploration identification system to adaptively match the full-scale exploration needs from regional general surveys (existence criteria) to detailed ore body surveys (continuous numerical thresholds), avoiding the mismatch problem of requiring excessively high quantitative accuracy at large scales or using overly coarse criteria at small scales. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the evolution chain of the five key mineralization processes in the porphyry copper deposit mineralization system of the present invention; Figure 3 This is a schematic diagram of the five-level metallogenic unit scale framework and membership constraint relationship of the present invention; Figure 4 This is a schematic diagram of the scale coupling of the construction of the dual-function conversion rule of the present invention; Figure 5 This is a schematic diagram of the scale coupling of the stratigraphic triple role transformation rule of the present invention; Figure 6 This is a schematic diagram showing the mirror correspondence between the four-level decomposition and the four-level aggregation of the present invention; Figure 7 This is a schematic diagram illustrating the structured definition and operator classification of the exploration method of the present invention. Detailed Implementation

[0022] This invention provides a method for constructing a multi-scale exploration identification system guided by ore-forming systems.

[0023] Example 1 uses a porphyry copper deposit mineralization system as the target mineralization system and the porphyry copper deposit mineralization domain of the Central Asian Orogenic Belt as the spatial scope to describe the steps of the present invention in detail. It should be noted that although this example uses a porphyry copper deposit, the method described in this invention is applicable to any type of mineralization system.

[0024] S1. Establish the evolution chain of key mineralization processes: For the porphyry copper deposit mineralization system, identify five key mineralization processes and establish an orderly evolution chain; S11. Identification of Metallogenic Processes: Based on regional geological surveys of the target metallogenic system, reports, ore deposit research papers, and ore deposit genetic models systematically identify key geodynamic processes directly related to metallogenesis. Identification criteria include: whether the process directly participates in the generation, migration, enrichment, or preservation of ore-forming materials, and whether the process has an identifiable geological record (such as specific rock assemblages, mineral assemblages, geochemical anomalies, etc.). S12. Chronological Sequence: The chronological order of key mineralization processes is determined based on the following geochronological evidence: (a) isotopic dating data, including zircon U-Pb ages (magma emplacement age), molybdenite Re-Os ages (mineralization age), and sericite Ar-Ar ages (alteration age); (b) the interpenetration and superposition relationships of geological bodies; and (c) the succession sequence of mineral assemblages. When the results of multiple dating methods are consistent, the chronological sequence is confirmed as reliable; when there are contradictions, direct dating data is preferred, and uncertainties are noted. S13. Construction of the evolution chain: The sorted key mineralization processes are linked together in the form of a directed linked list. Each node in the chain contains the following attributes: process name, time interval, dominant geodynamic mechanism, and product characteristics. The directed edges between nodes represent the temporal progression relationship (the product of the previous process serves as the input or condition of the next process), forming a complete sequence of dynamic processes.

[0025] (1) Source region formation: This involves processes such as dehydration and / or melting of the subducting slab, metasomatism and enrichment of the mantle wedge, and MASH (melting-assimilation-storage-homogenization) in the lower crust, which form the source region of ore-bearing magma. The time scale corresponding to this stage is from millions to tens of millions of years.

[0026] (2) Magma migration and emplacement: Ore-bearing magma migrates upward from the source region along deep faults and weak zones in the lithosphere and emplaces in the middle and upper crust, forming porphyry stocks or stock ensembles. During the migration process, assimilation and contamination of the wall rocks and magma differentiation may occur.

[0027] (3) Fluid exsolution and migration: When magma evolves to a certain stage, fluid exsolution occurs, in which high-temperature, high-salinity fluids containing metals separate from the magma body and migrate outward. Fluid migration is controlled by the tectonic fracture network and is accompanied by changes in the temperature-pressure gradient.

[0028] (4) Alteration and mineralization: The ore-bearing fluid reacts with the surrounding rock to produce alteration zones such as potassium silicate alteration, sericite alteration, argillaceization, and cyanite alteration. At the same time, metal sulfides precipitate at the points where physical and chemical conditions change abruptly to form ore bodies.

[0029] (5) Post-mineralization preservation: Supergene weathering and leaching after mineralization can form secondary enrichment zones. Tectonic activity can lead to faulting, displacement, tilting, rotation, or secondary activation and enrichment of the ore body, which determines whether the deposit can be preserved to this day and has economic mining value.

[0030] The five key mineralization processes constitute a strict temporal evolution chain: source region formation → magma migration and emplacement → fluid exsolution and migration → alteration and mineralization → post-mineralization preservation. This sequence is irreversible and cannot be skipped.

[0031] The chronological arrangement of the five key mineralization processes is not a simple linear assumption, but rather based on constraints from a large amount of geochronological data. Taking the porphyry copper deposits in the Central Asian orogenic belt as an example, the time constraint corresponding to the formation of the source region is the onset time of subduction-related magmatism (e.g., zircon U-Pb age approximately 340-330 Ma); the time constraint for magma migration and emplacement is the crystallization age of the porphyry body (e.g., 329±3 Ma); the time constraint for fluid exsolution and migration is the formation age of high-temperature alteration minerals (e.g., molybdenite Re-Os age 327±2 Ma); the time constraint for alteration mineralization is the formation age of low-temperature alteration minerals (e.g., sericite Ar-Ar age 325±2 Ma); and the time constraint for post-mineralization preservation continues to the present day. The chronological order and partial overlap of the time intervals of each stage jointly verify the rationality of the evolutionary chain.

[0032] S2, Decomposition of Component Processes: Each component process is associated with its superior key mineralization process through a tree-like hierarchical relationship. That is, each component process records the key mineralization process number to which it belongs (such as Stage 1, Stage 2, etc.), forming a two-level hierarchical process tree structure. S21. The decomposition of alteration and mineralization involves the following processes: S21.1, Decompression boiling and physical rapid unloading: Rapid precipitation of metal sulfides triggered by sudden changes in physicochemical conditions such as decompression boiling in microcrack spaces or fluid mixing. S21.2, Two-way contact assimilation reaction: The intense contact metasomatism and material replacement that occurs when a metal-bearing hydrothermal fluid encounters a calcium-magnesium carbonate rock or a siliceous-calcium stratum.

[0033] S22. The decomposition of the source region consists of the following processes: S22.1 Dehydration of subducting slabs: When oceanic slabs subduct to a depth of 60-120 km, they undergo metamorphism and dehydration, releasing metal-containing fluids that replace the overlying mantle wedge. S22.2 Partial melting of the mantle wedge: The replaced and enriched mantle wedge undergoes partial melting after the solidus temperature is lowered, generating primitive magma rich in ore-forming metals such as copper; S22.3 Lower crustal MASH process: Primitive magma undergoes melting-assimilation-storage-homogenization (MASH) in the lower crust, further enriching ore-bearing metals and adjusting magma composition. The decomposition is based on the fact that each subprocess is distinguishable in physical space (mantle vs. lower crust) and can be independently identified in terms of action mechanism (dehydration vs. melting vs. mixing). S23. The decomposition of magma migration and emplacement consists of the following processes: S23.1 Magma ascends and migrates along deep faults: Ore-bearing magma migrates from the lower crust to the middle and upper crust along deep fault systems at the lithosphere scale; S23.2, Assimilation and Contamination of Wall Rocks: During the migration process, magma exchanges materials with the surrounding rocks; S23.3 Shallow emplacement: Magma is emplaced in tectonically favorable locations in the middle and upper crust (such as fault intersections or the core of anticlines) to form porphyry stocks; S24. The decomposition of fluid exsolution transport consists of the following processes: S24.1 Magma fluid exsolution: After magma crystallizes and evolves to the point of fluid saturation, high-temperature, high-salinity fluids containing metals exsolution separate from the top of the magma body; S24.2 Fluid migration along fracture network: Exsolution of mineral-bearing fluids is driven by temperature and pressure gradients to migrate and diffuse outwards along structural fractures and vein networks; S25. The decomposition process of post-mining preservation consists of the following components: S25.1, Surface weathering and leaching and secondary enrichment: After the ore body is exposed to the surface, weathering and leaching can form an oxidation zone and a secondary sulfide enrichment zone; S25.2, Tectonic Modification and Preservation: Later tectonic activities cause the ore body to be displaced, tilted, or buried, which determines whether the ore body can be preserved to this day and has economic value; S3. Establish a five-level metallogenic unit scale framework: Level 1 – Metallogenic Domain: This domain covers an area of ​​approximately one million square kilometers and corresponds to the maximum spatial extent of an orogenic belt or metallogenic region. For example, the Central Asian porphyry copper metallogenic domain (CAPCMD) covers an area of ​​approximately 6.82 million square kilometers.

[0034] The second level is the metallogenic province, which covers an area of ​​approximately 100,000 square kilometers and corresponds to a metallogenic subregion controlled by a unified geodynamic regime. Examples include the Kazakhstan-Gobi Metallogenic Province (KGMP), the Tuva-Mongolia Metallogenic Province (TMMP), and the Northeast China Metallogenic Province (NEMP).

[0035] The third level – metallogenic belts: These cover an area of ​​approximately tens of thousands of square kilometers and correspond to linear or arc-shaped metallogenic sections controlled by specific tectonic frameworks, such as the Eastern Tianshan metallogenic belt, the Southern Gobi metallogenic belt, and the Circum-Balkhash metallogenic belt.

[0036] Level 4 – Mining Fields: These cover areas ranging from hundreds to thousands of square kilometers and correspond to localized areas where mineral deposits cluster, such as the Oyu Tolgoi mining field.

[0037] Level 5 – Deposits / Ore Bodies: These have an area ranging from square kilometers to sub-square kilometers and correspond to specific mineralized bodies, such as the Hugo Dummett ore body.

[0038] There are strict top-down hierarchical constraints among the five levels: metallogenic domains contain several metallogenic provinces, each metallogenic province contains several metallogenic belts, each metallogenic belt contains several ore fields, and each ore field contains several ore deposits / ore bodies. For example: CAPCMD → KGMP → South Gobi Belt → Oyu Tolgoi Ore Field → Hugo Dummett Ore Body.

[0039] S4. Extraction of indicator elements of scale coupling: By introducing scale function transformation rules, the specific implementation process of constructing dual function transformation rules is illustrated by taking construction elements as an example. At the metallogenic domain to metallogenic belt scale (levels 1-3), deep faults and suture zones mainly serve as conduit systems for magma and fluid transport. For example, the various suture zones and strike-slip fault systems in the Central Asian orogenic belt control the transport path of ore-bearing magma from the mantle source region to the middle and upper crust, and their indicative element is the regional deep fault / suture zone-magma conduit.

[0040] At the scale from mining field to deposit / ore body (level 4-5), the function of the same structural element changes: faults are no longer long-distance migration channels, but provide spatial locations for mineralization. For example, in the Oyu Tolgoi mining field, the intersection of faults and the structural detachment zone control the location of the ore body, and their indicative elements are transformed into ore-controlling fault-hosting spaces.

[0041] Similarly, the implementation of the three-stage role transformation rule of strata is as follows: at the scale of metallogenic domain and metallogenic province, the composition and evolution history of the basement strata determine the metallogenic potential of the source area (source area modulation); at the scale of metallogenic belt and ore field, the lithological combination of specific strata provides favorable ore-hosting space for mineralization (in-situ modulation); at the scale of deposit / ore body, the surrounding rock strata directly participate in the mineralization process as a medium that reacts with fluids (mineralization modulation).

[0042] The implementation of the progressive rule of age-constrained precision is as follows: at the metallogenic domain scale, the age constraint is at the epoch level (e.g., Late Paleozoic metallogenic domain); at the metallogenic province scale, it is at the period level (e.g., Late Carboniferous-Permian main metallogenic period); at the metallogenic belt scale, it is at the interdecadal level (e.g., the concentrated metallogenic period of 323-280 Ma); at the ore field scale, it is at the precise period level (e.g., zircon U-Pb age 329±3 Ma); and at the deposit / ore body scale, it is at the instantaneous event level (e.g., single magmatic pulse).

[0043] S4.1 Specific operational process for extracting indicator elements and establishing correlation relationships: Taking the decompression boiling process of alteration mineralization as an example, at the deposit / orebody scale (level 5), according to the scale function conversion rule, the structural elements in the decompression boiling process are given the function of ore-hosting space. Therefore, the extracted indicator element is the densely developed microfracture zone (as the spatial carrier of decompression boiling). At the ore field scale (level 4), the structural elements in the same composition process are also given the function of ore-hosting space, but the spatial granularity is upgraded to the ore-controlling fault intersection. Therefore, the extracted indicator element is the fault intersection structural detachment zone. Each of the above indicator elements records the correlation relationship in the form of a ternary set: (decompression boiling, deposit / orebody level, densely developed microfracture zone) and (decompression boiling, ore field level, fault intersection structural detachment zone), thereby realizing the bidirectional tracing of the indicator element to its source composition process and the scale level to which it belongs.

[0044] The extraction process is similar for the three-fold transformation of stratigraphic elements: at the scale of metallogenic domains and metallogenic provinces, stratigraphic elements are given the role of source region modulation, and the extracted indicator elements are the rock composition and evolutionary history of the basement strata (such as the existence of Archean-Paleoproterozoic metamorphic basement); at the scale of metallogenic belts and ore fields, stratigraphic elements are given the role of in-situ modulation, and the extracted indicator elements are the lithological assemblage and physical and mechanical properties of specific strata (such as carbonate-clastic interbedded rocks being favorable host lithology); at the scale of deposits / ore bodies, stratigraphic elements are given the role of mineralization modulation, and the extracted indicator elements are the chemical reactivity of the surrounding rock strata (such as the contact metasomatic reaction capacity of calcium-magnesium carbonate strata).

[0045] S5. Determine exploration markers: Taking the indicator elements of the alteration and mineralization stage (potassium silicate alteration zone) as an example, determine its corresponding exploration markers as follows: Exploration Mark 1: Short-wave infrared (SWIR) absorption peak wavelength of muscovite Al-OH < 2200 nm, exploration method is Class B - mineralogical method, quantization threshold is wavelength < 2200 nm.

[0046] Exploration Mark 2: Anomaly in apparent polarization of induced polarization. The exploration method is Class C - geophysical method, and the quantification threshold is apparent polarization > 3%.

[0047] Exploration Identifier 3: Soil geochemical Cu-Mo-Au combination anomaly, exploration method is Class D-geochemical method, quantification threshold is Cu > 3 times the standard deviation of the background value.

[0048] The structured exploration method is defined using exploration identifier 1 as an example: the input is ASTER or WorldView-3 multispectral remote sensing image, the feature operators are spectral band ratio calculation and endmember decomposition, and the output is the abundance distribution map of altered minerals and the wavelength value of Al-OH absorption peak.

[0049] The specific details of the method for determining the quantitative threshold and the criteria for judging exploration markers are as follows: (1) Existence criteria (metallogenic domain to metallogenic province scale): The criteria are binary (yes / no). For example, to determine whether there is large-scale intermediate-acidic magmatic activity or regional deep fault system in the target area, the method is to identify the presence or absence of geological elements based on remote sensing interpretation, regional geological maps and gravity anomaly data, with the threshold being the presence or absence of geological elements.

[0050] (2) Qualitative classification criteria (from metallogenic province to metallogenic belt scale): The criteria are in the form of categories. For example, to determine whether an igneous rock series belongs to the calc-alkaline, high-potassium calc-alkaline, or alkaline series, the method is as follows: based on rock geochemical analysis data, the rock series are classified according to the TAS diagram (total alkali-silica diagram) or K2O-SiO2 diagram, and the classification boundary is based on the internationally accepted rock classification standard.

[0051] (3) Semi-quantitative interval criteria (metallogenic belt to ore field scale): The criterion is in the form of numerical intervals, such as determining whether the Sr / Y ratio falls within the adakite affinity range of 20-40. The determination method is as follows: statistically analyze the distribution of geochemical parameters of known ore-bearing porphyry bodies and non-ore-bearing porphyry bodies, and use the 25th percentile to the 75th percentile of the parameter distribution of the ore-bearing porphyry body as the semi-quantitative criterion interval.

[0052] (4) Continuous numerical quantification threshold (from mining field to deposit / ore body scale): The criterion is a precise numerical threshold, such as Sr / Y > 40, La / Yb > 20, SiO2 = 56-70 wt.%, oxygen fugacity higher than the NNO+1 buffer line, SWIR muscovite Al-OH absorption peak wavelength < 2200 nm, apparent polarizability > 3%, soil Cu > 3 times the standard deviation of the background value, etc. The determination method is as follows: based on the statistical analysis of known deposits, calculate the optimal discrimination threshold of each index between ore-bearing and non-ore-bearing samples, and use the receiver operating characteristic curve (ROC) or Youden index to determine the optimal classification threshold.

[0053] Implementation of scale-dependent quantification precision for the same indicator: Taking intermediate-acidic intrusive rocks as an example, this illustrates the progression of quantification precision for the same indicator at different scales: At the metallogenic domain scale (Level 1), the exploration indicator is the existence criterion: whether there is large-scale intermediate-acidic magmatic activity in the area. The criterion is in binary form (yes / no) and is used for preliminary screening in the regional selection stage.

[0054] At the scale of metallogenic provinces to metallogenic belts (levels 2-3), exploration identification criteria are upgraded to qualitative classification criteria to semi-quantitative criteria: whether the magmatic rock series is calc-alkaline or high-potassium calc-alkaline, and whether it has adakitic affinity (approximate range of Sr / Y ratio). The criteria are in the form of categories or intervals.

[0055] At the scale from mining field to deposit / orebody (level 4-5), exploration identification is further upgraded to continuous numerical quantification thresholds: Sr / Y > 40, La / Yb > 20, SiO2 = 56-70 wt.%, oxygen fugacity higher than the NNO+1 buffer line, etc. The criterion is in the form of precise numerical thresholds, which are directly used for borehole target area location.

[0056] The aforementioned quantitative precision progression and the era-constrained precision progression constitute a dual progression mechanism: the time dimension progresses from the epoch level to instantaneous events, and the identification dimension progresses from existence criteria to continuous numerical thresholds. Together, they ensure that the exploration identification system can adaptively match the full-scale exploration needs from regional general surveys to detailed ore body surveys.

[0057] The specific implementation rules of the above-mentioned quantitative precision progressive mechanism are summarized as follows: For each indicator element, exploration markers and corresponding exploration methods with different precision levels are configured at different scale levels. At the metallogenic domain scale, large-area coverage methods such as remote sensing interpretation and regional geological mapping are used, combined with existence criteria for preliminary screening. At the metallogenic province and metallogenic belt scale, regional geochemical scanning and medium-precision geophysical exploration methods are used, combined with qualitative classification criteria and semi-quantitative interval criteria for regional optimization. At the ore field and ore deposit / ore body scale, detailed mineralogical analysis, fine geophysical exploration, and drilling verification methods are used, combined with continuous numerical quantification thresholds for precise target area positioning. This progressive mechanism ensures the matching of exploration input with information precision, avoiding excessive input at the large-scale stage and insufficient information at the small-scale stage.

[0058] Implementation of hierarchical aggregation: Taking mineral exploration prediction in the Eastern Tianshan metallogenic belt as an example, the specific implementation process of hierarchical aggregation is as follows: (1) Indicator element level combination: The detection results of three exploration indicators (SWIR alteration interpretation, induced polarization anomaly, and soil geochemical anomaly) under the indicator element of alteration mineralization stage (potassium silicate alteration zone) are weighted and fused to generate the comprehensive index of [potassium silicate alteration intensity].

[0059] The specific calculation process for each level of combination is as follows: The specific process of combining indicator elements: Taking the potassic silicate alteration zone in the alteration and mineralization stage as an example, this indicator element contains three exploration markers (SWIR alteration interpretation, induced polarization anomaly, and soil geochemical anomaly), and the detection results of each marker are spatially distributed raster layers. The combination method is as follows: First, spatial alignment and normalization are performed on each layer (standardizing each indicator value to the 0-1 range). Then, a weighted superposition method is used to calculate the comprehensive index. The weight is determined based on the information content or evidence weight of each exploration marker. The calculation formula is W_total = Σ(w_i × V_i), where w_i is the weight of the i-th exploration marker, and V_i is the normalized detection value of the i-th exploration marker.

[0060] The specific process of combining process-level elements involves cross-element fusion of the combined results of all indicator elements associated with the same process. Taking the alteration and mineralization stage as an example, the comprehensive indicators of various indicator elements, such as potassic silicate alteration zones, sericitized alteration zones, and argillaceous alteration zones, are superimposed twice. The fusion method employs either the comprehensive information quantity method or the evidence weight method, determining the fusion weight by calculating the information contribution of each indicator element to the known deposit.

[0061] The specific process of the key process-level combination involves integrating the comprehensive evaluations of the constituent processes of the five key mineralization processes. For example, it integrates five comprehensive indicators: source region formation evaluation, magma migration and emplacement evaluation, fluid exsolution and migration evaluation, alteration and mineralization evaluation, and post-mineralization preservation evaluation. The weights of each key process in this combination can be adaptively determined using the weight of evidence method based on the known spatial distribution of the deposit, or they can be manually set by geological experts based on the theoretical importance of the mineralization model.

[0062] The specific process of prediction-level combination is as follows: The comprehensive evaluation results of key processes are globally fused with additional spatial constraint information (such as tectonic intersection density maps, magma center distance fields, etc.). A random forest model or deep learning model is used for nonlinear combination, outputting the final probability map of the prospecting target area. The training samples of the model are known deposit / mineral occurrence locations (positive samples) and randomly sampled points in the background area (negative samples). The input features are the comprehensive index values ​​of the aforementioned combinations at each level.

[0063] (1) Composition process level combination: The combination results of each indicator element (potassic silicate alteration zone, sericite alteration zone, argillaceous alteration zone, etc.) in the alteration and mineralization stage are fused across elements to generate the comprehensive index of [alteration and mineralization system integrity].

[0064] (2) Key process level combination: The comprehensive indicators of the five key mineralization processes are integrated to generate a comprehensive evaluation of the whole process of the development degree of the mineralization system.

[0065] (3) Predictive-level combination: The evaluation of the development level of the mineralization system is combined with the spatial evidence layer, and a deep learning model is used for global fusion to generate the final mineral exploration target area probability map.

[0066] The specific construction process of each model in the synthesis operator is as follows: (1) Construction of the evidence weight model: Using known mineral deposit / mineral occurrence locations as training data, positive weight W+ and negative weight W- are calculated for the binarized layer of each exploration marker, W+ = ln[P(B|D) / P(B|D̄)], where P(B|D) is the probability of the marker appearing in the known mineral deposit area, and P(B|D̄) is the probability of the marker appearing in the non-mineral area. The contrast C = W+ - W- is used to evaluate the mineralization differentiation power of the marker. The larger the C value, the stronger the differentiation power of the marker. The final comprehensive evaluation value is the algebraic sum of W+ or W- for each marker.

[0067] (2) Construction of the comprehensive information content model: For each exploration marker's multi-level classification layer (not a binarized layer), calculate the information content value I(x_i) = ln[N(D∩x_i) / N(x_i)] - ln[N(D) / N(S)] for each level, where N(D∩x_i) is the number of known mineral deposits in the i-th level region, N(x_i) is the area (number of pixels) of the i-th level region, N(D) is the total number of known mineral deposits in the entire region, and N(S) is the total area of ​​the entire region. The information content values ​​of each marker are summed to obtain the comprehensive information content evaluation map.

[0068] (3) Construction of random forest model: The comprehensive index value of each level combination is used as the input feature matrix, and the binary label of the known ore deposit / ore point is used as the target variable to construct a random forest classification model. The model parameters include the number of decision trees (e.g., 500 trees), the maximum depth of each tree (e.g., 10 layers), and the number of candidate features in each split (e.g., the square root of the total number of features). The model training uses cross-validation to evaluate the generalization performance, and the output is the mineralization probability value of each pixel.

[0069] (4) Construction of deep learning model: multi-layer fully connected neural network or convolutional neural network is adopted. The input is multi-channel grid data composed of comprehensive index values ​​of each level combination. The network structure includes an input layer, several hidden layers (each layer contains batch normalization, ReLU activation and Dropout regularization) and an output layer (Sigmoid activation outputs the mineralization probability). Training adopts the binary cross-entropy loss function and Adam optimizer, and overfitting is prevented by early stopping strategy.

[0070] S6. Rule Constraints and Validation: Load all transformation results from S1 to S5 into the rule engine, and then perform compliance checks on each rule in the application domain knowledge rule set. The specific verification process includes the following: (1) Scale function conversion rule verification: Check each indicator element one by one whether the functional role assigned to it at its scale level conforms to the preset scale function conversion rule. For example, if it is found that the structural element is assigned the function of passage rather than the function of ore-containing space at the ore field scale, it is marked as a violation and a correction is prompted.

[0071] (2) Inspection of time-series constraints: Check whether the time relationship between each component process and its products meets the time-series logic constraints. For example, the formation age of altered minerals should not be earlier than the magma emplacement age; the time of post-mineralization alteration events should not be earlier than the time of mineralization events.

[0072] (3) Verification of mineral assemblage succession rules: Check whether the identification and zoning of alteration mineral assemblage conform to the known mineral co-occurrence sequence and hydrothermal evolution direction. For example, potassic silicate alteration should be located in the inner zone of the alteration system in space, sericite alteration and argillaceization should transition outward in sequence, and cyanite alteration should be located in the outermost zone.

[0073] (4) Verification of spatial geometric constraint rules: Check whether the spatial distribution pattern of geological entities such as mineralized bodies and alteration zones is reasonable. For example, alteration zones should be distributed in concentric rings or irregular cylindrical shapes around the magma center. If a significant deviation from the expected spatial pattern is found, it is necessary to make corrections in conjunction with post-mineralization tectonic modification information.

[0074] The test results are divided into three categories: pass (the conversion result conforms to the rule constraints), warning (the conversion result deviates from the rule partially but within a reasonable range), and violation (the conversion result obviously violates the rule constraints). For violations, the system automatically suggests possible correction directions, which are then reviewed and finalized by geological experts.

[0075] In this embodiment, some of the domain knowledge rules applied are exemplified as follows: (1) Late alteration mineralization rule (mineral assemblage succession): Copper mineralization can occur on a large scale in the late chlorite-epidote-sericite alteration stage and should not be limited to the evaluation of the early potassic alteration stage. This rule constrains the extraction of indicator elements of the alteration mineralization stage in step S4 to avoid omitting the mineralization contribution of the late retrograde metamorphic facies zone.

[0076] (2) Rules for determining the absence of secondary enrichment zones (spatial geometric constraints): When there is a lack of secondary enrichment zones under extremely arid climate conditions, it should not be judged as having low mineralization potential. The preservation integrity of primary mineralization should be assessed in conjunction with low-temperature thermochronological data.

[0077] (3) Optimization rule for the wing of the orogenic bend (scale function conversion type): In the giant bend structure, the porphyry copper deposit clusters are systematically concentrated in the wing of the bend rather than the core closed end. The intersection of the magmatic arc overlaid on the wing and the boundary of the basement block is the optimal section. This rule constrains the extraction of structural indicator elements from the metallogenic belt to the ore field scale in S4.

[0078] (4) Altering zoning symmetry axis tilt correction rule (spatial geometric constraint type): Regional tectonic stress after mineralization can cause the porphyry system to tilt and rotate as a whole, causing the original vertical zoning pattern to undergo a systematic angular deflection. The high-grade core may shift to the deep flank. This rule constrains the spatial positioning criteria of the relevant exploration markers of alteration zoning in S5.

[0079] This invention can be used to construct a multi-scale exploration identification system guided by metallogenic systems. It includes an evolutionary chain establishment module for identifying several key metallogenic processes for a target metallogenic system and establishing an ordered evolutionary chain according to geological time sequence; a process decomposition module for decomposing each key metallogenic process into several constituent processes and establishing hierarchical relationships; a scale framework establishment module for establishing a hierarchical metallogenic unit framework of at least three levels; an indicator element extraction module for extracting indicator elements from the constituent processes according to scale function conversion rules within the multi-level metallogenic unit scale framework; an exploration identification determination module for determining the corresponding exploration identification, exploration method, and quantification threshold for each indicator element; and a rule constraint and verification module for constraining and verifying the conversion results using a set of domain knowledge rules.

[0080] The present invention can be used in an electronic device, including a processor and a memory, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to perform any one of the methods S1 to S5.

[0081] The present invention can be used in a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any one of the methods S1 to S5.

[0082] Although specific embodiments of the present invention have been described in detail with reference to the accompanying drawings, this should not be construed as limiting the scope of protection of the present invention. Various modifications and variations that can be made by those skilled in the art without inventive effort within the scope described in the claims are still within the scope of protection of the present invention.

Claims

1. A method for constructing a multi-scale exploration identification system guided by metallogenic systems, characterized in that: The method includes the following steps: S1. Establish the metallogenic process evolution chain: For the target metallogenic system, identify the key metallogenic processes and establish an ordered evolution chain between each key metallogenic process according to the geological time sequence. The ordered evolution chain represents the sequence of complete dynamic processes of the metallogenic system from the formation of the source region to the preservation of the mineralization process. S2, Decomposition and Composition Process: Decompose the key mineralization process in S1 into composition processes, correspond each composition process to a specific geodynamic event and physicochemical action, and establish the direct subordinate relationship between the composition process and the key mineralization process. S3. Establish a multi-level metallogenic unit scale framework: Establish a multi-level hierarchical metallogenic unit framework, with strict hierarchical subordinate relationships between metallogenic units at each level. The metallogenic unit framework provides a multi-scale spatial reference system for subsequent transformations. S4: Extraction of indicator elements by scale coupling: Under the multi-level metallogenic unit scale framework, indicator elements are extracted from the composition process for each level of metallogenic unit; among them, the same geological element is assigned different functional roles at different scale levels according to the preset scale function conversion rules, and the indicator elements are associated with the composition process and metallogenic unit. The scale-function conversion rules in S4 include: a dual-function conversion rule for tectonic elements, where tectonic elements primarily function as conduits at the metallogenic domain, metallogenic province, and metallogenic belt scales, and transform into ore-hosting space functions at the ore field, deposit, and ore body scales; a triple-role conversion rule for stratigraphy, where stratigraphic elements function as source region modulation as basement at the metallogenic domain and metallogenic province scales, as situ modulation as ore-hosting space at the metallogenic belt and ore field scales, and as mineralization modulation as reaction medium at the deposit and ore body scales; and a progressive rule for age-constrained precision, where temporal resolution increases as the metallogenic unit scale decreases, progressing from epoch-level precision to instantaneous metallogenic event precision. S5: Determine exploration identifiers: For each indicator element, determine the corresponding exploration identifier. Each exploration identifier is associated with at least one exploration method and corresponding quantitative thresholds and discrimination criteria. S6: Rule Constraints and Verification: The conversion results from S1 to S5 are constrained and verified using a preset domain knowledge rule set. The domain knowledge rule set includes constraints on the functional roles of geological elements, the definition of spatiotemporal constraint ranges, and the evaluation of evidence strength. S61. Construct a domain knowledge rule set: This rule set consists of at least one of the following: scale function transformation rules, temporal constraint rules, mineral assemblage succession rules, and spatial geometric constraint rules. S62. Apply rule checks to the conversion results from S1 to S5 one by one: assign a functional role value to each indicator element and check whether it conforms to the scale function conversion rule; check whether the temporal relationship of the mineralization process meets the temporal constraint rule; check whether the alteration mineral assemblage conforms to the mineral assemblage succession rule; check whether the spatial distribution of the mineralization body meets the spatial geometric constraint rule; if a conversion result that violates the rules is found, mark it as abnormal and perform manual review or automatic correction.

2. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The specific steps for establishing the ore-forming process evolution chain in S1 are as follows: S11. Based on the geological background data and existing research results of the target metallogenic system, identify the key geodynamic processes and material evolution links directly related to mineralization. S12. Based on geochronological data and the chronological superposition of geological events, the key mineralization processes are sorted in order of geological time from early to late. S13. Using a directed linked list data structure, each key mineralization process is linked together into an ordered evolution chain. Each node in the chain represents a key mineralization process, and the directed edges between nodes represent the temporal progression relationship and the causal driving relationship. The ordered evolution chain covers the complete dynamic process sequence from the formation of the mineral source area, magma migration and emplacement, fluid exsolution and migration, alteration and mineralization to post-mineralization preservation and modification.

3. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The specific steps of S2 are as follows: S21. For each key mineralization process, based on the geodynamic sequence and physicochemical reaction mechanism, it is decomposed into several subprocesses that are distinguishable in time and space. Each subprocess is a component process. The decomposition of component processes follows the following principles: each component process should correspond to an independently identifiable geodynamic event or physicochemical action; a one-to-many tree-like hierarchical relationship is established between each component process and the key mineralization process to which it belongs, that is, a key mineralization process contains multiple component processes, and each component process belongs to only one key mineralization process.

4. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The multi-level metallogenic unit scale framework in S3 includes five levels: metallogenic domain, metallogenic province, metallogenic belt, ore field, ore deposit, and ore body. These levels have a strict hierarchical relationship from top to bottom. The typical area of ​​a metallogenic domain is on the order of millions of square kilometers, a metallogenic province is on the order of hundreds of thousands of square kilometers, a metallogenic belt is on the order of tens of thousands of square kilometers, an ore field is on the order of hundreds to thousands of square kilometers, and ore deposits and ore bodies are on the order of square kilometers to sub-square kilometers. The specific steps are as follows: S31. Based on the spatial distribution range of the target metallogenic system and the hierarchical nature of geological ore-controlling factors, determine the number of metallogenic unit levels, with at least three levels. S32. Metallogenic units are divided into levels from top to bottom. The division of each level of metallogenic unit is based on the geological criteria of structural framework, metallogenic geological background, consistency of ore source, area scale, and spatial nesting relationship. Strict hierarchical subordinate constraints are established between metallogenic units at each level. That is, the metallogenic unit at the next higher level completely contains the metallogenic unit at the next lower level in space, and the metallogenic unit at the next lower level must not cross the boundary of the metallogenic unit at the next higher level.

5. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The specific steps of S4 are as follows: S41. Under the multi-level metallogenic unit scale framework, for each level of metallogenic unit, traverse each component process determined in S2, determine the functional role that the geological elements involved in the component process should play at the current scale level according to the preset scale function conversion rules, and extract the corresponding indicator elements accordingly. S411, wherein an association is established between the indicator element and the composition process and the metallogenic unit: the source composition process and the scale level of each indicator element are recorded in the form of a triple, forming a structured association index of the indicator element, so that the indicator element can be traced back to its corresponding composition process and metallogenic unit. The triple includes the composition process, the metallogenic unit level, and the indicator element.

6. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The specific steps of S5 are as follows: S51. For each indicator element, based on its geological attributes and observability, determine the exploration identifier for detecting the indicator element under field and laboratory conditions; the determination of each exploration identifier must be associated with the following information simultaneously: at least one structured exploration method and corresponding quantification threshold and discrimination criteria. The quantification threshold is determined based on the statistical data and geochemical benchmark values ​​of known mineral deposits, and its accuracy increases as the scale of the metallogenic unit decreases, progressing step by step from existence criteria, qualitative classification criteria, semi-quantitative interval criteria to continuous numerical quantification thresholds. S511. The structured exploration method includes the definition of input data, operators, and output results. The operators include feature operators bound to a single exploration identifier and comprehensive operators that dynamically combine the detection results of exploration identifiers. The specific definition of the feature operators is as follows: for each exploration identifier, the input data, processing rules, and output format of the identifier are clearly defined, and the complete process from the original exploration data to the identifier detection result is solidified into a reproducible rule definition. The specific definition of the comprehensive operators is as follows: the input is a layer of detection results of multiple exploration identifiers, the processing rules are weighted superposition, evidence weight accumulation, and machine learning fusion algorithm, and the output is a comprehensive evaluation layer of combined identifiers. The comprehensive operators include evidence weight method, comprehensive information content method, random forest model, and deep learning model.

7. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 6, characterized in that: The detection results of the exploration markers in S511 are aggregated hierarchically. This hierarchical aggregation corresponds to the decomposition hierarchy of S1 to S5, including indicator element level combination, component process level combination, key process level combination, and prediction level combination. The specific steps are as follows: S511.1, The indicator element level combination performs weighted fusion of the exploration marker detection results under the same indicator element, and the weight is determined based on the information content and evidence weight of each marker; the composition process level combination uses the comprehensive information content method and evidence weight method to fuse the cross-indicator results of the same composition process to generate a composition process level comprehensive evaluation. The key process-level combination integrates the results of cross-component processes associated with the same key mineralization process to generate a key process-level comprehensive evaluation; the prediction-level combination uses a random forest model or a deep learning model to perform global nonlinear fusion of all key mineralization process-level results to output the final prospecting target area probability map.

8. The method for constructing a multi-scale exploration identification system guided by ore-forming systems as described in claim 1, characterized in that: The quantitative accuracy of the exploration identifier in S5 increases as the scale of the metallogenic unit decreases: at the scale from metallogenic domain to metallogenic province, the exploration identifier adopts the existence criterion or qualitative classification criterion; at the scale from metallogenic belt to ore field, the semi-quantitative classification criterion or quantitative interval criterion is adopted; at the scale of ore deposit / ore body, the continuous numerical quantification threshold is adopted.

Images