A method and system for identifying tectonic mineralization models based on multi-scale aeromagnetic data processing

Abstract

This invention discloses a method and system for identifying tectonic mineralization models based on multi-scale aeromagnetic data processing, relating to the field of geological and mineral exploration technology. The method includes: collecting and preprocessing regional and local-scale raw aeromagnetic data; decomposing standardized aeromagnetic data into multiple components at different scales using the db4 wavelet basis, corresponding to geological body information at different depths; performing targeted interference removal on each scale component using adaptive threshold filtering; extracting magnetic anomaly information using anomaly enhancement technology; quantifying and extracting multi-scale tectonic information based on magnetic anomalies and known geological data, and performing apparent magnetic susceptibility imaging inversion; establishing a magnetic anomaly feature library for typical mineralization models and constructing a multi-scale tectonic-mineralization correlation model; constructing a discrimination model using a random forest algorithm, outputting the mineralization model type, mineralization potential level, and favorable target area, and verifying the results. This invention solves the problems of incomplete tectonic identification, inaccurate interference removal, and strong subjectivity in mineralization identification inherent in traditional methods.

Description

Technical Field

[0001] This invention relates to the field of geological and mineral exploration technology, and more specifically to a method and system for identifying tectonic mineralization models based on multi-scale aeromagnetic data processing. Background Technology

[0002] Airborne magnetic surveying, as an important tool for basic geological surveys and mineral resource exploration, boasts advantages such as low cost, high accuracy, wide coverage, and minimal impact from harsh geological environments. It has been widely applied in geodynamics research, geological structure analysis, and mineral resource exploration. The core of tectonic metallogenic models is to reveal the intrinsic relationship between tectonic evolution and mineralization, clarifying the spatial distribution and coupling relationships of ore-controlling elements such as ore-controlling structures, ore-bearing strata, and ore-forming rocks. The accuracy of this identification directly determines the efficiency and precision of mineral resource exploration.

[0003] Currently, the identification of tectonic mineralization models based on aeromagnetic data mainly relies on single-scale aeromagnetic data processing, which has the following technical shortcomings: First, traditional methods mostly use single-resolution aeromagnetic data for analysis, failing to take into account both the macroscopic tectonic background at the regional scale and the hidden mineralization anomalies at the local scale. This results in insufficient completeness and refinement of tectonic identification, making it difficult to capture the mineralization response characteristics of structures at different scales. Second, aeromagnetic data contains a large number of interference signals, and existing preprocessing methods have failed to accurately remove multi-scale interference, easily causing anomaly identification bias and thus affecting the accuracy of mineralization model identification. Third, the correlation analysis between tectonic information and mineralization elements lacks systematicity, relying heavily on qualitative judgments based on human experience, which is highly subjective and has failed to establish a quantitative correspondence between multi-scale tectonic features and mineralization models, resulting in multiple solutions and failing to meet the needs of precise mineral exploration. Fourth, existing technologies have not formed a complete integrated system of multi-scale data processing, tectonic identification, and mineralization identification, which is cumbersome, inefficient, and unable to achieve real-time linkage and updating of data processing and identification results.

[0004] As mineral exploration in my country deepens, shallow mineral resources are becoming increasingly depleted, making the exploration of concealed mineral resources increasingly urgent. Especially in areas with severe surface cover, conventional geological and geochemical exploration techniques are ineffective, making aeromagnetic data the core means of obtaining information on deep geological structures and mineralization.

[0005] Therefore, developing a method and system that can achieve precise processing of multi-scale aeromagnetic data, efficient extraction of structural information, and scientific identification of mineralization models, and solving the problems of multiple solutions, subjectivity, and inefficiency in existing technologies, is of great practical significance for improving the efficiency of mineral resource exploration and reducing exploration costs. Summary of the Invention

[0006] In view of this, the present invention provides a method and system for identifying tectonic mineralization modes based on multi-scale aeromagnetic data processing. Through layered processing of multi-scale aeromagnetic data, precise interference removal, quantitative extraction of structural information, and intelligent identification of mineralization modes, it achieves accurate and efficient identification of tectonic mineralization modes, solving the technical problems of incomplete structural identification, difficulty in removing abnormal interference, and strong subjectivity in mineralization identification in existing methods. To achieve the above objectives, the present invention adopts the following technical solution: A method for identifying structural mineralization models based on multi-scale aeromagnetic data processing includes: Raw aeromagnetic data at different scales were collected from the target area. The collected raw aeromagnetic data were preprocessed, and the standardized aeromagnetic data were decomposed into multiple scales using wavelet transform algorithm. The aeromagnetic components of different scales obtained by decomposition are subjected to interference removal using an adaptive threshold filtering algorithm, and the aeromagnetic components of each scale after interference removal are subjected to anomaly enhancement processing to extract magnetic anomaly information at different scales. Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, multi-scale structural information is quantitatively extracted. The apparent magnetic susceptibility imaging inversion algorithm is used to invert and calculate the magnetic anomalies at each scale, quantitatively characterize the magnetic features of the geological bodies, and establish the correspondence between structures and magnetic parameters. Collect metallogenic geological data of known ore deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale tectonic information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale tectonic-metallogenic correlation model, and quantitatively characterize the correlation between tectonic parameters and metallogenic probability. Based on the constructed multi-scale tectonic-metallogenic correlation model, combined with the random forest machine learning algorithm, a tectonic-metallogenic mode discrimination model is constructed. The extracted multi-scale tectonic information, magnetic parameters and anomalous features are used as input parameters and input into the discrimination model. The output is the tectonic-metallogenic mode type, metallogenic potential level and favorable metallogenic target area of ​​the target area.

[0007] Optionally, the raw aeromagnetic data includes: regional-scale aeromagnetic data and local-scale aeromagnetic data, wherein the regional-scale aeromagnetic data is used for macroscopic structural background analysis, and the local-scale aeromagnetic data is used for the identification of concealed mineralization anomalies.

[0008] Optionally, the preprocessing of the collected aeromagnetic raw data includes: sequentially performing data format standardization, diurnal variation correction, altitude correction, gradient correction, and trend leveling to remove instrument errors, environmental interference, and terrain influences, thereby obtaining standardized aeromagnetic data.

[0009] Optionally, the wavelet transform algorithm is used to perform multi-scale decomposition of the standardized aeromagnetic data, including: decomposing geological body information corresponding to different depths into components of different frequencies, wherein the low-frequency components correspond to deep macroscopic structures, and the high-frequency components correspond to shallow local structures and mineralization anomalies.

[0010] Optionally, the use of an adaptive threshold filtering algorithm for interference removal includes: for low-frequency components, focusing on removing regional magnetic field background interference while retaining magnetic anomalies caused by deep structures; for high-frequency components, focusing on removing surface human interference and shallow non-mineralized rock layer interference while retaining mineralized bodies and magnetic anomalies caused by local structures.

[0011] Optionally, the anomalous enhancement processing of the aeromagnetic components at each scale after removing interference includes: using the vertical second derivative and the horizontal first derivative combined with the tilt angle method to highlight the boundary features and amplitude differences of the magnetic anomaly, and extracting magnetic anomaly information at different scales, including anomaly amplitude, anomaly range, anomaly morphology and anomaly gradient.

[0012] Optionally, the quantitative extraction of multi-scale structural information includes: for the regional scale, extracting macroscopic structural information of regional fault zones, basement uplifts, and large intrusive intrusive bodies to determine the trend, extension length, scale, and spatial distribution of macroscopic structures; for the local scale, extracting local structural information of secondary faults, concealed intrusive bodies, and mineralization alteration zones to determine the occurrence, scale, and correlation with mineralization of local structures.

[0013] Optionally, the metallogenic geological data of known deposits in the target area include metallogenic type, ore-controlling structures, ore-bearing strata, ore-forming rock bodies and mineralization anomaly characteristics, and the typical metallogenic models include sedimentary metamorphic type, magmatic type, contact metasomatic-hydrothermal type, and volcanic type.

[0014] Optionally, it also includes verifying the discrimination results by combining known geological verification data of the target area. If the discrimination accuracy is lower than the preset threshold, the process returns to the previous step to adjust the interference stripping parameters and anomaly extraction threshold, and reprocesses and discriminates until the accuracy requirements are met.

[0015] Optionally, a metallogenic mode discrimination system based on multi-scale aeromagnetic data processing includes: an acquisition module for acquiring raw aeromagnetic data of different scales in the target area, preprocessing the acquired raw aeromagnetic data, and using wavelet transform algorithm to decompose the standardized aeromagnetic data into multiple scales; Feature processing module: Used to remove interference from the aeromagnetic components of different scales obtained by decomposition using an adaptive threshold filtering algorithm, and to perform anomaly enhancement processing on the aeromagnetic components of each scale after interference removal to extract magnetic anomaly information at different scales. Extraction module: Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, it quantitatively extracts multi-scale structural information, uses the apparent magnetic susceptibility imaging inversion algorithm to perform inversion calculations on magnetic anomalies at each scale, quantitatively characterizes the magnetic features of geological bodies, and establishes the correspondence between structures and magnetic parameters; The comparison module is used to collect metallogenic geological data of known mineral deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale structural information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale structural-metallogenic correlation model, and quantitatively characterize the correlation between structural parameters and metallogenic probability. The discrimination module is used to construct a tectonic mineralization pattern discrimination model based on the constructed multi-scale tectonic-mineralization correlation model and combined with the random forest machine learning algorithm. The extracted multi-scale tectonic information, magnetic parameters and anomaly features are used as input parameters to the discrimination model, and the output is the tectonic mineralization pattern type, mineralization potential level and favorable mineralization target area of ​​the target area.

[0016] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method and system for identifying tectonic mineralization modes based on multi-scale aeromagnetic data processing, which has the following beneficial effects: This invention enables hierarchical processing and precise application of multi-scale aeromagnetic data. By decomposing different frequency components through wavelet transform, it corresponds to deep macroscopic structures and shallow local structures respectively, taking into account both the macroscopic mineralization background and the identification of local mineralization anomalies. This solves the problems of incomplete and insufficient refinement of structural identification caused by traditional single-scale processing, and improves the comprehensiveness and accuracy of structural information extraction.

[0017] An adaptive threshold filtering algorithm is used to remove interference at multiple scales. The interference characteristics of aeromagnetic components at different scales are processed in a targeted manner, effectively removing background interference, human interference and non-mineralization interference. Combined with anomaly enhancement technology, the accuracy of magnetic anomaly identification is significantly improved, the deviation of anomaly identification is reduced, and the ambiguity in mineralization identification is reduced.

[0018] A multi-scale tectonic-metallogenic correlation model was constructed, and machine learning algorithms were combined to achieve intelligent identification of tectonic metallogenic patterns. The correlation between tectonic parameters and metallogenic probability was quantitatively characterized, replacing traditional manual experience judgment, reducing subjectivity, and improving the scientificity and accuracy of metallogenic pattern identification.

[0019] By combining apparent magnetic susceptibility imaging inversion technology, a three-dimensional magnetic parameter model can be generated, which can intuitively characterize the deep distribution features of geological bodies, further reveal the spatial coupling relationship between tectonics and mineralization, and provide accurate technical support for the delineation of favorable mineralization target areas. This will help improve the efficiency of mineral resource exploration and reduce exploration costs. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of a structural mineralization mode discrimination method based on multi-scale aeromagnetic data processing provided by the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] This invention discloses a method for identifying structural mineralization modes based on multi-scale aeromagnetic data processing, such as... Figure 1 As shown, it includes: Raw aeromagnetic data at different scales were collected from the target area. The collected raw aeromagnetic data were preprocessed, and the standardized aeromagnetic data were decomposed into multiple scales using wavelet transform algorithm. The aeromagnetic components of different scales obtained by decomposition are subjected to interference removal using an adaptive threshold filtering algorithm, and the aeromagnetic components of each scale after interference removal are subjected to anomaly enhancement processing to extract magnetic anomaly information at different scales. Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, multi-scale structural information is quantitatively extracted. The apparent magnetic susceptibility imaging inversion algorithm is used to invert and calculate the magnetic anomalies at each scale, quantitatively characterize the magnetic features of the geological bodies, and establish the correspondence between structures and magnetic parameters. Collect metallogenic geological data of known ore deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale tectonic information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale tectonic-metallogenic correlation model, and quantitatively characterize the correlation between tectonic parameters and metallogenic probability. Based on the constructed multi-scale tectonic-metallogenic correlation model, combined with the random forest machine learning algorithm, a tectonic-metallogenic mode discrimination model is constructed. The extracted multi-scale tectonic information, magnetic parameters and anomalous features are used as input parameters and input into the discrimination model. The output is the tectonic-metallogenic mode type, metallogenic potential level and favorable metallogenic target area of ​​the target area.

[0024] In a specific implementation, a method for identifying structural mineralization modes based on multi-scale aeromagnetic data processing includes the following steps: Step 1: Multi-scale aeromagnetic data acquisition and preprocessing Raw aeromagnetic data at different scales were collected for the target area. This raw aeromagnetic data included regional-scale aeromagnetic data (scale < 1:50,000) and local-scale aeromagnetic data (scale ≥ 1:50,000). Regional-scale aeromagnetic data was used for macroscopic structural background analysis, while local-scale aeromagnetic data was used for identifying concealed mineralization anomalies. The collected raw aeromagnetic data underwent preprocessing, including data format standardization, diurnal variation correction, height correction, gradient correction, and trend leveling. This process removed instrument errors, environmental interference, and topographic effects, resulting in standardized aeromagnetic data. A wavelet transform algorithm was then used to decompose the standardized aeromagnetic data into multi-scale components, corresponding to geological body information at different depths. Low-frequency components corresponded to deep macroscopic structures, while high-frequency components corresponded to shallow local structures and mineralization anomalies.

[0025] Step 2: Multi-scale interference stripping and anomaly extraction

[0026] For the aeromagnetic components of different scales obtained from step 1, an adaptive threshold filtering algorithm is used to remove interference: for low-frequency components, the focus is on removing regional magnetic field background interference while retaining magnetic anomalies caused by deep structures; for high-frequency components, the focus is on removing surface human interference and shallow non-mineralized rock layer interference while retaining magnetic anomalies caused by mineralized bodies and local structures; after removing interference, the aeromagnetic components of each scale are subjected to anomaly enhancement processing, using the vertical second derivative, the horizontal first derivative combined with the tilt angle method to highlight the boundary features and amplitude differences of magnetic anomalies, and extract magnetic anomaly information at different scales, including anomaly amplitude, anomaly range, anomaly morphology and anomaly gradient.

[0027] Step 3: Quantitative Extraction of Multi-Scale Construction Information

[0028] Based on the magnetic anomaly information at different scales extracted in step 2, and combined with known geological data of the target area, multi-scale structural information is quantitatively extracted: For the regional scale (low-frequency component), macroscopic structural information such as regional fault zones, basement uplifts, and large intrusive bodies is extracted to determine the strike, extension length, scale, and spatial distribution of macroscopic structures; for the local scale (high-frequency component), local structural information such as secondary faults, concealed intrusive bodies, and mineralized alteration zones is extracted to determine the occurrence, scale, and correlation with mineralization of local structures; the apparent magnetic susceptibility imaging inversion algorithm is used to invert and calculate magnetic anomalies at each scale to obtain three-dimensional data volumes of apparent magnetic susceptibility at different depths, quantitatively characterize the magnetic features of geological bodies, and establish the correspondence between structures and magnetic parameters.

[0029] Step 4: Analysis of mineralization correlation characteristics

[0030] Collect metallogenic geological data of known deposits in the target area, including metallogenic type, ore-controlling structures, ore-bearing strata, ore-bearing intrusive bodies, and mineralization anomaly characteristics. Establish a magnetic anomaly feature library for typical metallogenic models, including sedimentary metamorphic, magmatic, contact metasomatic-hydrothermal, and volcanic models. Compare the multi-scale structural information and magnetic parameters extracted in step 3 with the magnetic anomaly feature library to analyze the correlation between structures at different scales and metallogenic elements. Clarify the hierarchical characteristics of ore-controlling structures (macro-controlling and local-controlling), the magnetic response law of ore-bearing intrusive bodies, and the spatial coupling relationship between mineralization anomalies and structures. Construct a multi-scale structural-metallogenic correlation model to quantitatively characterize the correlation between structural parameters (strike, scale, and depth) and metallogenic probability.

[0031] Step 5: Intelligent identification of mineralization patterns

[0032] Based on the multi-scale tectonic-metallogenic correlation model constructed in step 4, a tectonic-metallogenic pattern discrimination model is constructed by combining a machine learning algorithm (random forest algorithm). The multi-scale tectonic information, magnetic parameters, and anomaly features extracted in step 3 are used as input parameters and input into the discrimination model. Through model training and iterative optimization, the tectonic-metallogenic pattern type, metallogenic potential level, and favorable metallogenic target area of ​​the target area are output. The discrimination results are verified by combining known geological verification data of the target area. If the discrimination accuracy is lower than the preset threshold (≥85%), the process returns to step 2 to adjust the interference stripping parameters and anomaly extraction threshold, and reprocesses and discriminates until the accuracy requirement is met.

[0033] In a specific implementation, the multi-scale tectonic-mineralization correlation model constructed in step 4, combined with a machine learning algorithm (random forest algorithm), is used to construct a tectonic-mineralization pattern discrimination model, including: I. Basic Data Preparation for Model Building The input feature parameters are determined using the quantitative parameters output by the multi-scale tectonic-metallogenic correlation model as the core input, including: regional scale tectonic features (fault strike, extension length, basement uplift scale), local scale tectonic features (secondary fault occurrence, concealed rock mass scale, spatial relationship of mineralization alteration zones), magnetic parameters (apparent magnetic susceptibility value, three-dimensional distribution characteristics), and magnetic anomaly features (anomaly amplitude, range, morphology, gradient), forming a standardized feature parameter matrix.

[0034] The labeled data is defined using known typical mineralization modes in the target area as classification labels, including four types: sedimentary metamorphic, magmatic, contact metasomatic-hydrothermal, and volcanic. At the same time, mineralization potential level labels (high, medium, and low) and mineralization favorable target area labels are defined to construct a labeled training dataset.

[0035] II. Basic Setup of Random Forest Model

[0036] The total sample set is constructed based on the standardized feature parameter matrix and the labeled training dataset. The first dataset is then drawn using random sampling with replacement. The first dataset was trained using the random forest algorithm to obtain a mining pattern discrimination model; A second dataset is extracted using random sampling with replacement. This second dataset is then input into the mining pattern discrimination model for prediction, and the prediction results are output.

[0037] Furthermore, the first dataset is trained using the random forest algorithm to obtain a mining pattern discrimination model, specifically as follows: Multiple random sample sets are generated by randomly extracting feature information from the first dataset. A classification and regression tree is generated based on a classification and regression algorithm for multiple random sample sets, and the Gini coefficient of the multiple random sample sets is calculated. For multiple random sample sets where the current node of the classification and regression tree is located, if the Gini coefficient of the multiple random sample sets is less than the Gini coefficient threshold, the classification and regression algorithm returns to the sub-decision tree and stops recursion; or, if the Gini coefficient of the multiple random sample sets is not less than the Gini coefficient threshold, the Gini coefficient of each feature information in the random sample set where the current node is located is calculated. The feature information corresponding to the smallest Gini coefficient among the Gini coefficients of each feature information is taken as the optimal feature information. Based on the optimal feature information, the corresponding random sample set is divided into a first information set and a second information set, and the first information set and the second information set are respectively taken as the left child node and the right child node of the current node. Calculate the Gini coefficients of the first information set and the second information set. When the Gini coefficients of the first information set and the second information set are less than the Gini coefficient threshold, the classification and regression algorithm returns the sub-decision tree and stops recursion, generating multiple decision trees. A mining pattern discrimination model is obtained by learning from multiple decision trees using multiple classifiers.

[0038] Furthermore, the second dataset is input into the mining pattern discrimination model for prediction, and the prediction result is output as follows: The second dataset is divided into a training set and a test set according to a preset ratio; The feature information of the training set is randomly sampled multiple times to obtain a sampled sample set. After each random sampling, the sampled feature information is put back into the second dataset for the next random sampling. Multiple weight values ​​are generated for the sampled sample set based on multiple decision trees; Calculate the sum of multiple weight values, and divide the sum of multiple weight values ​​by the number of random samplings to obtain the prediction result.

[0039] Furthermore, it also includes: pre-setting matching rules for the feature information of the first dataset; correcting the weight values ​​output by the decision tree based on the matching rules; summing the corrected weight values ​​to obtain the weight correction value; predicting the test set based on the mining pattern discrimination model to obtain the average weight of the test set; and outputting the prediction result when the weight correction value is equal to the average weight.

[0040] Furthermore, multiple feature information is randomly extracted from the first dataset to obtain the random sample set, wherein the data format of the first dataset is a multi-row, multi-column data matrix.

[0041] Furthermore, the receiving operator feature curve is used to select from multiple output results of the mining pattern discrimination model so that the mining pattern discrimination model outputs a prediction result.

[0042] In a specific implementation, a metallogenic mode discrimination system based on multi-scale aeromagnetic data processing includes: Acquisition module: Used to acquire raw aeromagnetic data at different scales in the target area, preprocess the acquired raw aeromagnetic data, and use wavelet transform algorithm to decompose the standardized aeromagnetic data at multiple scales; Feature processing module: Used to remove interference from the aeromagnetic components of different scales obtained by decomposition using an adaptive threshold filtering algorithm, and to perform anomaly enhancement processing on the aeromagnetic components of each scale after interference removal to extract magnetic anomaly information at different scales. Extraction module: Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, it quantitatively extracts multi-scale structural information, uses the apparent magnetic susceptibility imaging inversion algorithm to perform inversion calculations on magnetic anomalies at each scale, quantitatively characterizes the magnetic features of geological bodies, and establishes the correspondence between structures and magnetic parameters; The comparison module is used to collect metallogenic geological data of known mineral deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale structural information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale structural-metallogenic correlation model, and quantitatively characterize the correlation between structural parameters and metallogenic probability. The discrimination module is used to construct a tectonic mineralization pattern discrimination model based on the constructed multi-scale tectonic-mineralization correlation model and combined with the random forest machine learning algorithm. The extracted multi-scale tectonic information, magnetic parameters and anomaly features are used as input parameters to the discrimination model, and the output is the tectonic mineralization pattern type, mineralization potential level and favorable mineralization target area of ​​the target area.

[0043] In a specific embodiment, a method for identifying tectonic metallogenic models based on multi-scale aeromagnetic data processing is provided. Taking a certain region as the research object, this region has abundant tectonic development and excellent metallogenic conditions, but the surface cover is severe and shallow geological outcrops are scarce. The specific steps are as follows: Step 1: Multi-scale aeromagnetic data acquisition and preprocessing Raw aeromagnetic data of the target area were collected, including regional-scale aeromagnetic data (scale 1:100,000) and local-scale aeromagnetic data (scale 1:50,000). The regional-scale aeromagnetic data covered the entire target area and its surroundings and was used for macroscopic structural background analysis. The local-scale aeromagnetic data focused on covering known mineral deposits and their surrounding areas, with the flight altitude controlled within 160m, and was used for the identification of concealed mineralization anomalies.

[0044] The collected aeromagnetic raw data underwent preprocessing: First, aeromagnetic data in different formats (GRD, dat format) were converted into a unified standard format to eliminate format differences; then, diurnal variation correction was performed using data from synchronously observed geomagnetic stations to remove interference from diurnal variations in the geomagnetic field; next, altitude correction was performed to eliminate the influence of terrain undulations on the aeromagnetic data based on flight altitude and terrain data; then, gradient correction was performed to correct gradient biases in the aeromagnetic data; finally, trend leveling was performed using a trend approximation method to remove interference from slow changes in the regional magnetic field background, resulting in standardized aeromagnetic data.

[0045] The standardized aeromagnetic data was decomposed into five scales using a wavelet transform algorithm (db4 wavelet basis). Scales 1-2 are high-frequency components, corresponding to shallow (0-500m) local structures and mineralization anomalies; scales 3-4 are mid-frequency components, corresponding to secondary structures in the middle (500-1500m); and scale 5 is a low-frequency component, corresponding to deep (>1500m) macroscopic structures and basement uplift.

[0046] Step 2: Multi-scale interference stripping and anomaly extraction

[0047] For the five aeromagnetic components obtained from the decomposition, an adaptive threshold filtering algorithm is used to remove interference: For scale 5 (low-frequency component), a lower filtering threshold is set to remove regional magnetic field background interference, while retaining magnetic anomalies caused by deep macroscopic structures; For scales 1-2 (high-frequency component), a higher filtering threshold is set to remove surface human interference (such as pipelines and buildings) and shallow non-mineralized rock layer interference, while retaining magnetic anomalies caused by mineralized bodies and local faults; For scales 3-4 (mid-frequency component), an intermediate threshold is set to remove mid-scale interference signals, while retaining magnetic anomalies caused by secondary structures.

[0048] Anomaly enhancement processing was performed on the aeromagnetic components at each scale after removing interference: the vertical second derivative was used to highlight the boundary features of the magnetic anomaly, and the horizontal first derivative (135° direction) was used to highlight the anomalous response of the NE-trending fault structure. The anomaly boundary identification effect was further optimized by combining the tilt angle method. Magnetic anomaly information at each scale was extracted, including anomaly amplitude, anomaly range, anomaly morphology and anomaly gradient. The amplitude range of the mineralization anomaly is 50-200 nT, and the gradient is greater than 10 nT / km.

[0049] Step 3: Quantitative Extraction of Multi-Scale Construction Information

[0050] Based on the known geological data (strata, structures, and ore deposits) of the target area, and using the extracted multi-scale magnetic anomaly information, structural information was quantitatively extracted: For scale 5 (low-frequency component), the distribution information of large fault zones, basement uplifts, and large intrusive bodies in the region was extracted, and their trend was determined to be northeast, with an extension length greater than 50 km, indicating a large scale and representing macroscopic ore-controlling structures in the region; For scales 3-4 (mid-frequency component), secondary fault zones were extracted, mainly trending northeast, with an extension length of 10-50 km, representing local ore-controlling structures; For scales 1-2 (high-frequency component), the distribution information of concealed rock masses and mineralized alteration zones was extracted. The apparent magnetic susceptibility of the concealed rock masses ranged from 500 to 1500 × 10^-6 SI, and the mineralized alteration zones corresponded to high magnetic anomaly areas, consistent with the locations of known ore deposits.

[0051] The apparent magnetic susceptibility imaging inversion algorithm was used to invert and calculate magnetic anomalies at various scales, resulting in three-dimensional data volumes of apparent magnetic susceptibility at different depths. This clearly characterized the spatial distribution features of deep basement, intermediate secondary structures, and shallow mineralization bodies, and established the correspondence between structural strike, scale, and apparent magnetic susceptibility parameters.

[0052] Step 4: Analysis of mineralization correlation characteristics

[0053] Collect metallogenic geological data of known mineral deposits in the target area. The main metallogenic model in this area is the magmatic metallogenic model. The ore-controlling structure is the northeast-trending left-lateral shear fracture zone. The ore-forming rock mass is an intermediate-acidic intrusive rock mass. Mineralization anomalies are mainly distributed near the fault zone and the contact zone of the rock mass.

[0054] A magnetic anomaly feature library for the magmatic mineralization model was established, and its magnetic anomaly features were identified as follows: at the regional scale, there is a NE-trending linear high magnetic anomaly zone, corresponding to macroscopic ore-controlling faults; at the local scale, there are blocky and strip-shaped high magnetic anomalies, corresponding to concealed ore-forming rock bodies and mineralization alteration zones, with high apparent magnetic susceptibility (500-1500×10^-6SI) and large anomaly gradient (>10nT / km).

[0055] By comparing the extracted multi-scale structural information and magnetic parameters with the feature library, the analysis revealed that: the NE-trending macro-fault zone controls regional magmatic activity and ore body migration, and is the core of regional ore control; the secondary fault zone provides channels for the migration and enrichment of ore-forming fluids; the concealed ore-forming rock mass is the material basis for mineralization; and the mineralization anomaly is mainly distributed in the contact zone between the fault zone and the rock mass, forming a metallogenic correlation relationship of macro-fault controlling ore, secondary fault guiding ore, and rock mass bearing ore. A multi-scale structural-metallogenic correlation model was constructed, and the correlation degree between fault strike and mineralization probability was 0.82, and the correlation degree between apparent magnetic susceptibility of rock mass and mineralization intensity was 0.78.

[0056] Step 5: Intelligent identification of mineralization patterns

[0057] Based on the constructed multi-scale tectonic-metallogenic correlation model, a random forest algorithm was used to construct a tectonic metallogenic mode discrimination model. The structural strike, fault scale, apparent magnetic susceptibility of the rock mass, anomalous amplitude, and anomalous gradient were selected as input parameters, and the known metallogenic mode type was selected as output parameter. The model was trained (with 100 training samples, of which 80 were training samples and 20 were validation samples). The model parameters were iteratively optimized to achieve a model validation accuracy of 88%.

[0058] The multi-scale structural information, magnetic parameters and anomaly features extracted in step 3 are input into the discrimination model. The output shows that the structural mineralization mode type of the target area is magmatic mineralization mode, the mineralization potential level is high, and three favorable mineralization target areas are delineated, all located near the contact zone between the NE-trending fault zone and the concealed rock mass.

[0059] The accuracy rate of the identification was 89%, which is higher than the preset threshold (85%), and the identification results are reliable. For the delineated favorable mineralization target areas, it is recommended to carry out ground geophysical exploration profile verification and drilling verification to provide guidance for subsequent mineral exploration work.

[0060] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0061] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for identifying structural mineralization models based on multi-scale aeromagnetic data processing, characterized in that, include: Raw aeromagnetic data at different scales were collected from the target area. The collected raw aeromagnetic data were preprocessed, and the standardized aeromagnetic data were decomposed into multiple scales using wavelet transform algorithm. The aeromagnetic components of different scales obtained by decomposition are subjected to interference removal using an adaptive threshold filtering algorithm, and the aeromagnetic components of each scale after interference removal are subjected to anomaly enhancement processing to extract magnetic anomaly information at different scales. Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, multi-scale structural information is quantitatively extracted. The apparent magnetic susceptibility imaging inversion algorithm is used to invert and calculate the magnetic anomalies at each scale, quantitatively characterize the magnetic features of the geological bodies, and establish the correspondence between structures and magnetic parameters. Collect metallogenic geological data of known ore deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale tectonic information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale tectonic-metallogenic correlation model, and quantitatively characterize the correlation between tectonic parameters and metallogenic probability. Based on the constructed multi-scale tectonic-metallogenic correlation model, combined with the random forest machine learning algorithm, a tectonic-metallogenic mode discrimination model is constructed. The extracted multi-scale tectonic information, magnetic parameters and anomalous features are used as input parameters and input into the discrimination model. The output is the tectonic-metallogenic mode type, metallogenic potential level and favorable metallogenic target area of ​​the target area.

2. The method for determining structural mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The raw aeromagnetic data includes regional-scale aeromagnetic data and local-scale aeromagnetic data, wherein the regional-scale aeromagnetic data is used for macroscopic structural background analysis, and the local-scale aeromagnetic data is used for the identification of concealed mineralization anomalies.

3. The method for determining structural mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The preprocessing of the collected aeromagnetic raw data includes: sequentially performing data format standardization, diurnal variation correction, altitude correction, gradient correction, and trend leveling to remove instrument errors, environmental interference, and terrain influences, thereby obtaining standardized aeromagnetic data.

4. The method for determining structural mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The wavelet transform algorithm is used to decompose the standardized aeromagnetic data into multi-scale components, including: decomposing geological body information corresponding to different depths into components of different frequencies, where low-frequency components correspond to deep macroscopic structures and high-frequency components correspond to shallow local structures and mineralization anomalies.

5. The method for determining structural mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The aforementioned adaptive threshold filtering algorithm for interference removal includes: for low-frequency components, focusing on removing regional magnetic field background interference while retaining magnetic anomalies caused by deep structures; for high-frequency components, focusing on removing surface human interference and shallow non-mineralized rock layer interference while retaining mineralized bodies and magnetic anomalies caused by local structures.

6. The method for determining tectonic mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The process of enhancing the aeromagnetic components at each scale after removing interference includes: using the vertical second derivative and the horizontal first derivative combined with the tilt angle method to highlight the boundary features and amplitude differences of the magnetic anomaly, and extracting magnetic anomaly information at different scales, including anomaly amplitude, anomaly range, anomaly morphology and anomaly gradient.

7. The method for determining tectonic mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The quantitative extraction of multi-scale structural information includes: at the regional scale, extracting macroscopic structural information of regional fault zones, basement uplifts, and large intrusive rock masses to determine the trend, extension length, scale, and spatial distribution of macroscopic structures; at the local scale, extracting local structural information of secondary faults, concealed rock masses, and mineralization alteration zones to determine the occurrence, scale, and correlation with mineralization of local structures.

8. The method for determining tectonic mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, The metallogenic geological data of known deposits in the target area include metallogenic type, ore-controlling structures, ore-bearing strata, ore-forming rock bodies and mineralization anomaly characteristics. The typical metallogenic models include sedimentary metamorphic type, magmatic type, contact metasomatic-hydrothermal type, and volcanic type.

9. The method for determining tectonic mineralization models based on multi-scale aeromagnetic data processing according to claim 1, characterized in that, It also includes verifying the discrimination results by combining known geological verification data of the target area. If the discrimination accuracy is lower than the preset threshold, the process returns to the previous step to adjust the interference stripping parameters and anomaly extraction threshold, and then reprocesses and discriminates until the accuracy requirements are met.

10. A structural mineralization mode discrimination system based on multi-scale aeromagnetic data processing, characterized in that, Includes: Acquisition module: used to acquire raw aeromagnetic data at different scales in the target area, preprocess the acquired raw aeromagnetic data, and use wavelet transform algorithm to decompose the standardized aeromagnetic data into multiple scales; Feature processing module: Used to remove interference from the aeromagnetic components of different scales obtained by decomposition using an adaptive threshold filtering algorithm, and to perform anomaly enhancement processing on the aeromagnetic components of each scale after interference removal to extract magnetic anomaly information at different scales. Extraction module: Based on the extracted magnetic anomaly information at different scales, combined with known geological data of the target area, it quantitatively extracts multi-scale structural information, uses the apparent magnetic susceptibility imaging inversion algorithm to perform inversion calculations on magnetic anomalies at each scale, quantitatively characterizes the magnetic features of geological bodies, and establishes the correspondence between structures and magnetic parameters; The comparison module is used to collect metallogenic geological data of known mineral deposits in the target area, establish a magnetic anomaly feature library of typical metallogenic models, compare the extracted multi-scale structural information and magnetic parameters with the magnetic anomaly feature library, construct a multi-scale structural-metallogenic correlation model, and quantitatively characterize the correlation between structural parameters and metallogenic probability. The discrimination module is used to construct a tectonic mineralization pattern discrimination model based on the constructed multi-scale tectonic-mineralization correlation model and combined with the random forest machine learning algorithm. The extracted multi-scale tectonic information, magnetic parameters and anomaly features are used as input parameters to the discrimination model, and the output is the tectonic mineralization pattern type, mineralization potential level and favorable mineralization target area of ​​the target area.

Images