Ore prospecting target prediction method and system based on altered mineral analysis

Abstract

The invention discloses an altered mineral analysis-based prospecting target prediction method and system, and relates to the technical field of prospecting target prediction. An altered mineral analysis-based prospecting target prediction system comprises a data acquisition module, a clustering analysis module, an alteration combination discrimination module, a spatial modeling module, a space-time coupling module and a metallogenic evaluation module. According to the method, altered minerals and symbiotic combinations of the altered minerals are subjected to layered clustering treatment by introducing mineral thermodynamic phase diagram constraints, multi-stage superposed alteration information in a complex structure area is effectively analyzed, and a mineral symbiotic combination structure model with cause difference expression ability is established; by constructing cause period labels and forming a time sequence decoupling model, systematic distinguishing of alteration bodies formed under the mineralization effect of different times is achieved, and a time sequence basis is provided for identification of the multi-stage mineralization process.

Description

Technical Field

[0001] The present invention relates to the technical field of prospecting target area prediction, and in particular to a prospecting target area prediction method and system based on altered mineral analysis. Background Art

[0002] Altered minerals are important transformation products of geological bodies produced by the action of ore-forming fluids. Their types, compositional relationships, and spatial distribution are crucial for revealing mineralization processes and identifying prospecting targets. Existing techniques for identifying altered minerals rely primarily on spectral feature extraction from remote sensing imagery, alteration zone delineation in geological mapping, or element enrichment indices from geochemical anomalies. These methods are significantly limited in areas with frequent tectonic activity and overlapping mineralization processes.

[0003] In terms of alteration assemblage modeling, existing technologies mostly use methods based on rock thin sections, qualitative descriptions, or single spectral inversion, which are difficult to systematically reflect the thermodynamic stability and symbiotic combination characteristics between altered minerals, and lack quantitative expression methods for the genetic background of mineral assemblages. In terms of temporal discrimination, existing studies have failed to establish period identification rules associated with thermodynamic conditions, and there are obvious difficulties in identifying and staging multi-stage alteration bodies. Among mineralization prediction methods, conventional models usually rely on element anomaly thresholds, artificial weight settings, or simple statistical classification methods. They are unable to integrate uncertainty information in geological evolution and lack a unified cause-driven discrimination logic. In terms of spatial expression, traditional methods mostly use two-dimensional anomaly maps or local profile interpolation to display the distribution morphology of alteration bodies. They do not introduce the occurrence information in structural measurement data as a modeling constraint, and their three-dimensional spatial modeling capabilities are limited.

[0004] Existing methods are unable to distinguish multi-period superimposed information, lack the ability to model spatiotemporal coupling, rely on empirical judgment for mineralization identification, and are difficult to quantify the credibility of target area predictions. These methods make it difficult to meet the actual needs of deep mineral exploration and refined identification and prediction of mineralization systems in complex tectonic areas. Summary of the Invention

[0005] A method for predicting prospecting target areas based on altered mineral analysis, comprising:

[0006] Collect multi-source geological data of the target area, including structural measurement data, remote sensing image data, geochemical data, altered minerals and their paragenesis data;

[0007] Based on the constraints of mineral thermodynamic phase diagrams, hierarchical cluster analysis is performed on altered minerals and their paragenetic assemblage data, as well as mineral spectral features extracted from remote sensing image data. This generates a mineral paragenetic assemblage feature map, assigns genetic period labels, and constructs a temporal decoupling model.

[0008] The temporal decoupling model is used as a priori knowledge base and input into a Bayesian network-based probability discrimination model for mineralization-alteration combinations together with elemental anomaly indicators in geochemical data. The matching probability between the target alteration combination and the typical mineralization-alteration pattern is calculated, and the identification results of alteration bodies with matching probabilities higher than a preset threshold are output.

[0009] The spatial distribution trend characteristics of the alteration body identification results are extracted, and the input data set is constructed using the fault zone strike information extracted from the structural measurement data as spatial constraints. An implicit function field representing the alteration body boundary is generated through an implicit geological modeling algorithm, and a three-dimensional distribution boundary model of the alteration body is extracted.

[0010] Establish a spatiotemporal weight matrix, couple the genetic period label, matching probability, and the three-dimensional distribution boundary model of the alteration body, and generate the spatiotemporal evolution probability field of the alteration body, which serves as the basic data input for mineralization confidence calculation and target area prediction;

[0011] Based on the random forest algorithm, the mineralization confidence of the target area is calculated by integrating geological, geophysical, thermodynamic, remote sensing and geochemical indicators, and the target area prediction results and classification scores are output.

[0012] As a preferred technical solution of the present invention, the hierarchical cluster analysis includes:

[0013] The altered minerals are preliminarily grouped based on the stable domain characteristics of the thermodynamic phase diagram of the altered minerals and their paragenetic combination data; the mineral spectral characteristics extracted from the remote sensing image data are fused with the altered mineral grouping results through a preset weight coefficient; the fused data are input into the clustering model for hierarchical clustering processing; based on the grouping results output by the clustering model, the typical mineral assemblages and their paragenetic relationships in each group are extracted to construct the corresponding mineral paragenetic combination characteristic map.

[0014] As a preferred technical solution of the present invention, the fusion of the mineral spectral features extracted from the remote sensing image data with the altered mineral grouping results by using a preset weight coefficient includes:

[0015] Multiple characteristic reflection bands and spectral morphological parameters of altered minerals are extracted from remote sensing image data. The spectral characteristics are vectorized and encoded and combined with the altered minerals and their paragenetic combination data to form an input dataset. The input dataset is used as the input of the clustering model to perform hierarchical clustering processing.

[0016] As a preferred technical solution of the present invention, the construction of the causal period label includes:

[0017] Based on the characteristic map of mineral paragenesis and combined with the stable domain interval of the thermodynamic phase diagram of each group of altered minerals, a period discrimination rule is established; and according to the said rule, a unique genetic period label is assigned to each group of altered mineral combinations to characterize the alteration time attribute corresponding to the combination.

[0018] As a preferred technical solution of the present invention, the probability discrimination model of the mineralization alteration combination is a Bayesian network model constructed through typical mineralization alteration combination samples. The model uses the genetic period label generated in the time series decoupling model as the prior input node, and jointly constructs the input structure with the element anomaly indicators in the geochemical data; the matching probability between the target alteration combination and the typical mineralization alteration pattern is used as the output node, and the discrimination results are screened based on a preset probability threshold in the output stage, and the alteration body identification results with a matching probability higher than the threshold are output.

[0019] As a preferred technical solution of the present invention, the extraction of the three-dimensional distribution boundary model of the alteration body includes:

[0020] The spatial position and distribution trend characteristics of the alteration body identification results are obtained, and the fault zone attitude information extracted from the structural measurement data is used as a spatial constraint condition to construct a modeling data set including a spatial point set, structural control lines and distribution trend vectors. Based on the modeling data set, an implicit geological modeling algorithm is used to generate an implicit function field, and the three-dimensional distribution boundary model of the alteration body is obtained by extracting the function isosurface.

[0021] As a preferred technical solution of the present invention, generating the spatiotemporal evolution probability field of the alteration body includes:

[0022] The genetic period labels, the matching probability of the mineralization-alteration combination, and the three-dimensional distribution boundary model of the alteration body are used as input variables to construct a spatiotemporal weight matrix, in which the temporal compatibility weight is assigned according to the chronological order of the genetic period labels, with the weight of the early-stage alteration combination decreasing; the spatial continuity weight is calculated based on the spatial density gradient in the three-dimensional distribution boundary model of the alteration body, with the weight of the high-density area increasing; and the mineralization matching probability weight is directly based on the matching probability value output by the Bayesian network.

[0023] A gridded interpolation algorithm is used to perform three-dimensional spatial fitting on the spatiotemporal weight matrix. Within the spatial range defined by the three-dimensional distribution boundary model of the alteration body, a three-dimensional probability distribution model containing time attributes, spatial position and mineralization probability values ​​is generated as the spatiotemporal evolution probability field of the alteration body.

[0024] As a preferred technical solution of the present invention, the acquisition of the mineralization confidence includes:

[0025] The time distribution factor, spatial density index and mineralization matching probability extracted from the spatiotemporal evolution probability field of the alteration body are used as core feature parameters, and a feature vector set is constructed in combination with remote sensing image data, geochemical data and geological structure information. A random forest model is trained based on the feature vector set and applied to each grid cell in the target area. The mineralization confidence value of each cell is output as the probability scoring basis for target area prediction.

[0026] As a preferred technical solution of the present invention, the output target area prediction result includes:

[0027] Obtain the spatial distribution results of mineralization confidence, classify multiple grid cells in the target area according to the confidence threshold setting rules, and generate multiple candidate prospecting target cells; spatially aggregate adjacent cells of the same category confidence level, and output a prospecting target prediction layer with clear boundary range and level labels.

[0028] A prospecting target area prediction system based on altered mineral analysis, comprising:

[0029] Data acquisition module, used to collect multi-source geological data of the target area;

[0030] Cluster analysis module, used to perform hierarchical cluster analysis on altered minerals and their paragenetic combination data, as well as mineral spectral features extracted from remote sensing image data, and to construct a time series decoupling model;

[0031] The alteration combination discrimination module calculates the matching probability between the target alteration combination and the typical metallogenic alteration pattern based on the metallogenic alteration combination probability discrimination model, and outputs the alteration body identification results with matching probability higher than the preset threshold;

[0032] The spatial modeling module is used to generate an implicit function field representing the boundary of the erosion body through an implicit geological modeling algorithm, and to extract a three-dimensional distribution boundary model of the erosion body;

[0033] The spatiotemporal coupling module is used to couple the genesis period labels, matching probabilities and the three-dimensional distribution boundary model of the erosion volume to generate the spatiotemporal evolution probability field of the erosion volume;

[0034] The mineralization assessment module is used to calculate the mineralization confidence of the target area and output the target area prediction results and classification scores.

[0035] The present invention has the following advantages:

[0036] The present invention introduces mineral thermodynamic phase diagram constraints to perform hierarchical clustering on altered minerals and their paragenetic assemblages, effectively analyzing the alteration information of multiple superimposed periods in complex tectonic areas and establishing a mineral paragenetic assembly structural model with the ability to express genetic differences; by constructing genetic period labels and forming a temporal decoupling model, it is possible to systematically distinguish altered bodies formed under mineralization in different eras, providing a temporal basis for the identification of multi-stage mineralization processes.

[0037] The present invention establishes a probability discrimination mechanism for mineralization and alteration combinations by inputting the genetic period labels and geochemical anomaly indicators into the Bayesian network model, which significantly improves the scientificity and accuracy of the identification of mineralization directionality of the alteration body; by extracting the spatial distribution trend characteristics of the alteration body, combining structural information such as the fault zone occurrence as spatial constraints, and using the implicit geological modeling algorithm to realize the reconstruction of the three-dimensional boundary of the alteration body, the continuity and structural rationality of the spatial modeling of the alteration body are improved.

[0038] The present invention constructs a spatiotemporal weight matrix that includes time series, space and mineralization probability to generate a spatiotemporal evolution probability field of the alteration body, providing quantitative data support for subsequent target area prediction in a dynamic evolution context; by integrating multi-source data features such as remote sensing, geochemistry, and geological structure, the mineralization confidence evaluation is calculated, thereby enhancing the discriminative ability and credibility expression ability of prospecting results. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only schematic diagrams of the present invention. Those skilled in the art can also derive other drawings based on the provided drawings without inventive effort.

[0040] Figure 1 A schematic diagram of the structure of a prospecting target area prediction system based on altered mineral analysis adopted in an embodiment of the present invention;

[0041] Figure 2 This is a distribution diagram of altered mineral compositions used in the embodiments of the present invention;

[0042] Figure 3 This is a diagram of the mineralization-related alteration zoning pattern used in the embodiments of the present invention. DETAILED DESCRIPTION

[0043] To make the objectives, technical solutions, and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] Example 1, a method for predicting prospecting target areas based on altered mineral analysis, combined with Figure 2 and Figure 3 , including the following steps:

[0045] Step S1: Collect multi-source geological data of the target area, including structural measurement data, remote sensing image data, geochemical data, altered minerals and their paragenetic combination data, to build a data foundation for subsequent alteration identification and mineralization prediction.

[0046] Structural measurement data refers to the regional fault structure parameters extracted through field geological surveys or existing geological data, including fault strike, dip, inclination, displacement characteristics and fault control line distribution information, which are used as spatial constraint input in the subsequent implicit geological modeling process.

[0047] Remote sensing image data is medium-to-high-resolution multispectral remote sensing data covering the target area (including visible light, short-wave infrared, and thermal infrared bands), which is used to extract the spectral reflectance characteristics of altered minerals, such as characteristic absorption valleys, reflection peak positions, and spectral slopes.

[0048] Geochemical data include multi-element test data of rock, soil, water or sediment samples obtained through field sampling, and the element types include at least Cu, Pb, Zn, Mo, As, Au, Ag, Fe, Si, Al, K, Mg, etc.; they are used for subsequent identification of abnormal element enrichment areas and participate in the probabilistic discrimination modeling of mineralization and alteration combinations.

[0049] Altered mineral data refers to the types of altered minerals and their quantitative content data determined based on field sampling or mineral spectral detection, including typical altered minerals such as chlorite, sericite, muscovite, kaolinite, hematite, and pyrite.

[0050] Paragenesis data refers to a data structure based on the co-occurrence of different alteration minerals in a single-point mineral sample. This data is obtained by statistically analyzing the combination patterns of alteration minerals at different sampling points and is subsequently used to construct characteristic maps of alteration mineral paragenesis. Paragenesis data includes not only the names of co-occurring minerals, but also their order of magnitude ratio (such as the percentage relationship between the main and accessory minerals) and the frequency of occurrence.

[0051] The above multi-source geological data were obtained through the following methods:

[0052] The structural survey data mainly come from existing regional geological survey reports and field structural surveys;

[0053] Remote sensing image data is obtained through public satellite platforms or purchased from commercial data service agencies;

[0054] Geochemical data is obtained from field sampling and testing using methods such as XRF (X-ray fluorescence spectroscopy) and ICP-MS (inductively coupled plasma mass spectrometry);

[0055] Alteration mineral data were obtained by XRD (X-ray diffraction), SWIR (short-wave infrared spectrometer), and VNIR (visible-near infrared spectrometer);

[0056] The paragenesis data are statistically analyzed based on the above-mentioned mineral identification samples and processed into a database.

[0057] Step S2: Based on the constraints of the mineral thermodynamic phase diagram, a hierarchical cluster analysis is performed on the altered mineral and its paragenetic combination data, as well as the mineral spectral features extracted from the remote sensing image data, to generate a mineral paragenetic combination feature map, assign genetic period labels, and construct a temporal decoupling model;

[0058] Mineral thermodynamic phase diagrams are used to express the stability boundaries of minerals under different temperature and pressure conditions. In this example, phase diagram data generated by a rock thermodynamics database (such as THERMOCALC) was used to analyze the stable occurrence ranges of altered minerals in the target area under different geological conditions, serving as genetic constraints in subsequent cluster analysis.

[0059] The hierarchical cluster analysis comprises:

[0060] The altered minerals are preliminarily grouped based on the stability domain characteristics of the thermodynamic phase diagram of the altered minerals and their paragenetic combination data; this process screens the minerals in all sampling points for thermodynamic stability and divides them into several initial mineral groups based on their stable overlapping areas on the PT (pressure-temperature) diagram.

[0061] The mineral spectral features extracted from remote sensing image data are fused with the altered mineral grouping results through a preset weight coefficient; the preset weight coefficient is determined based on the spectral-mineral consistency in the sample data. The typical setting method is to introduce a multimodal weighting factor into the distance function of the clustering model to fuse the contribution intensity of remote sensing features and ground mineral distribution.

[0062] The fused data is then fed into a clustering model for hierarchical clustering. Clustering algorithms such as hierarchical clustering, density peak clustering, or spectral clustering can be used. The core of these models is to classify alteration assemblages layer by layer based on the similarities and stability of minerals. The purpose of this stratification is to distinguish between primary mineral assemblages, associated assemblages, and boundary assemblages, thereby improving the ability to identify weak mineralization alteration signals.

[0063] Based on the grouping results output by the clustering model, the typical mineral assemblages and their symbiotic relationships in each group are extracted, and the corresponding mineral symbiotic assembly characteristic map is constructed. This map is a structured description of the mineral type, abundance ratio, and spatial co-occurrence frequency within each assembly, and is the key input for the subsequent construction of the genetic period model.

[0064] In the process of implementing hierarchical clustering in typical sample areas, Figure 2 As shown in Figure 1, different types of alteration minerals and their associated assemblages exhibit distinct spatial distribution characteristics and zoning patterns. This map forms a key input for subsequent genetic stage identification and temporal decoupling modeling.

[0065] Multiple characteristic reflectance bands and spectral morphological parameters of altered minerals are extracted from remote sensing image data. Typical bands include 2200nm (muscovite), 2330nm (chlorite), 1900nm (hydroxyl water molecules), etc.; spectral morphological parameters include slope, envelope angle, symmetry factor, etc.

[0066] After vectorizing these spectral features (e.g., PCA dimensionality reduction and normalization), they are combined with data on altered minerals and their associated assemblages to form an input dataset. This input dataset is then used as input for the hierarchical clustering model. This fused dataset represents multiple source metrics in a unified feature space, enabling the model to simultaneously leverage both mineral assemblage structure and spectral expression differences for clustering.

[0067] The construction of the genesis period label includes:

[0068] Based on the characteristic map of mineral paragenesis and combined with the stable domain interval of the thermodynamic phase diagram of each group of altered minerals, the period discrimination rules are established; the discrimination rules are generated by counting the intersection of the stable zones of each combination in the PT space, the overlap with the mineral combinations of known mineralization ages, and the stable evolution path (for example, the evolutionary sequence from early sericite to late kaolinite).

[0069] According to the above rules, each altered mineral assemblage is assigned a unique genetic phase label to characterize the temporal nature of the alteration associated with that assemblage. These labels are coded using discrete time periods (e.g., phase numbers T1, T2, T3) and, in conjunction with regional chronological context and known tectonic events during the mineralization phase, serve as temporal prior inputs for subsequent probabilistic discriminant models.

[0070] The constructed temporal decoupling model is a ternary expression system of alteration combination-thermodynamic interval-time attribute, which can distinguish alteration combinations with different genetic backgrounds under the background of multi-period superposition, and provide a clear basis for time dimension stratification for subsequent Bayesian discrimination.

[0071] Step S3: The temporal decoupling model is used as a priori knowledge base and input together with the element anomaly indicators in the geochemical data into the probability discrimination model of mineralization alteration combination based on the Bayesian network. The matching probability between the target alteration combination and the typical mineralization alteration pattern is calculated, and the identification result of the alteration body with a matching probability higher than the preset threshold is output;

[0072] The prior knowledge base refers to the temporal decoupling model constructed in step S2. Its core data includes the genetic period label of each mineral assemblage, the structural characteristics of the paragenetic assemblage, and the thermodynamic stability background. The genetic period label has dual temporal and structural representation capabilities, and as a priori input, it significantly improves the model's ability to classify multi-stage superimposed alteration.

[0073] The Bayesian network model is based on Figure 3 The typical alteration combination mineralization model shown is the basis of the a priori structure. The model is constructed through field mineralization samples and clarifies the probability relationship of various combinations in different time sequences and mineralization environments.

[0074] Elemental anomaly indicators in geochemical data include typical mineralization-related elements such as Cu, Pb, Zn, Au, As, Sb, Hg, and Mo. These are identified within geochemical survey lines or sampling points using background-outlier discrimination methods (e.g., the 3σ method). Each sampling point generates a geochemical anomaly vector, representing the geochemical characteristics of its mineralization potential.

[0075] The mineralization-alteration combination probability discrimination model is a Bayesian network model constructed using typical mineralization-alteration combination samples. The model uses the genetic period labels generated in the temporal decoupling model as prior input nodes, and jointly constructs the input structure with the element anomaly indicators in the geochemical data; the input structure is a set of node networks with conditional dependencies, in which the prior nodes are discrete (period T1 / T2), and the geochemical indicators are continuous or discretized inputs, representing the joint probability distribution structure between the data.

[0076] The Bayesian network is implemented in two stages: structure learning and parameter learning:

[0077] Structural learning: Based on a combination of typical mineralization samples (a priori tectonic belt mineralization points and their geochemical-mineral combination characteristics), a network topology is constructed, and the causal paths between nodes are determined using greedy search (K2 algorithm) or scoring function method;

[0078] Parameter learning: Use maximum likelihood estimation or Bayesian estimation methods to determine the conditional probability distribution of each node in different states;

[0079] Finally, the matching probability between the target alteration combination and the typical mineralization alteration pattern is used as the output node, and the output value is a continuous probability between [0,1], which is used to characterize the mineralization similarity of the current alteration combination.

[0080] During the output phase, the system filters the identification results based on a preset probability threshold (0.6), eliminating low-confidence samples lacking clear mineralization characteristics. Only alteration body identification results with a matching probability above the threshold are output, serving as the input for subsequent 3D modeling and spatiotemporal fusion. This result essentially constitutes a spatially discrete, probabilistically continuous distribution map of alteration-directed targets. The preset probability threshold is obtained through machine learning fitting or expert experience.

[0081] Step S4: Extract the spatial distribution trend characteristics of the alteration body identification results, use the fault zone strike information extracted from the structural measurement data as a spatial constraint condition, and construct an input data set; generate an implicit function field representing the alteration body boundary through an implicit geological modeling algorithm, and extract a three-dimensional distribution boundary model of the alteration body;

[0082] The spatial distribution trend characteristics of the erosion body identification results refer to extracting the distribution direction, distribution density gradient and change trend of the erosion bodies in the target area based on the spatial position data of the high matching probability erosion bodies output in step S3 by using kernel density estimation (KDE), directional gradient analysis (such as principal component axis fitting), and spatial trend surface regression methods.

[0083] Fault zone occurrence information, including parameters such as strike, dip, and inclination, is typically derived from fault structural measurements from field geological surveys or inverted through remote sensing interpretation, geophysical profiling, and DEM analysis. In this step, this geometric occurrence information is used to constrain the modeling boundaries, ensuring that the final model is more accurate than geological reality.

[0084] The process of constructing the input data set includes: using the spatial position points in the erosion body identification results as the initial spatial point set; extracting the distribution trend vector field to form a vector data set reflecting the main controlling direction of the erosion body; defining the fracture control line (extracted through structural measurement data) as the structural constraint boundary; and combining the above data to construct a ternary data structure as the input basis for three-dimensional modeling.

[0085] Implicit geological modeling algorithms use mathematical functions to continuously represent geological boundaries. Typical methods include RBF (Radial Basis Function), SDF (Signed Distance Function), or Level-Set methods. The core idea is to define an implicit function φ(x, y, z) in three-dimensional space, such that the contour surface where φ = 0 represents the boundary of the alteration volume.

[0086] After generating the implicit function field, the φ=0 isosurface is extracted by a numerical solution method (MarchingCubes voxel algorithm) to construct a three-dimensional alteration volume boundary model with geological continuity and structural constraints.

[0087] The extraction of the three-dimensional distribution boundary model of the erosion body includes:

[0088] The spatial position and distribution trend characteristics of the alteration body identification results are obtained, and the fault zone attitude information extracted from the structural measurement data is used as a spatial constraint condition to construct a modeling data set including a spatial point set, structural control lines and distribution trend vectors. Based on the modeling data set, an implicit geological modeling algorithm is used to generate an implicit function field, and the three-dimensional distribution boundary model of the alteration body is obtained by extracting the function isosurface.

[0089] The three-dimensional model of the alteration body output by this step not only visually expresses the spatial morphology of the mineralization-related alteration body, but also provides a spatial structural basis for the subsequent construction of the spatiotemporal evolution probability field.

[0090] Step S5: Establish a spatiotemporal weight matrix, couple the genetic period label, matching probability, and the three-dimensional distribution boundary model of the alteration body, and generate the spatiotemporal evolution probability field of the alteration body as the basic data input for the calculation of mineralization confidence and target area prediction;

[0091] The generating of the spatiotemporal evolution probability field of the erosion body comprises:

[0092] The genetic period label, the matching probability of the mineralization alteration combination and the three-dimensional distribution boundary model of the alteration body are used as input variables to construct the spatiotemporal weight matrix, where:

[0093] The temporal compatibility weight is used to reflect the correlation strength of various alteration combinations on mineralization in different genetic stages. According to the time sequence of the genetic period labels, the time weight function is set, such as using an exponential decreasing function. ,in The period number, To adjust the coefficient, the early alteration effect accounts for a smaller proportion in the weighting;

[0094] Spatial continuity weights are used to quantify the degree of clustering of erosion artifacts or the structural integrity of spatial structures. These weights are constructed based on the spatial density gradient extracted from the 3D distribution boundary model. Specifically, the modeling area is divided into a 3D voxel grid and the weights are estimated based on the number density of erosion artifacts per unit volume, with denser regions being given higher weights.

[0095] The metallogenic matching probability weight is directly derived from the matching probability result output by the Bayesian network, indicating the statistical similarity and credibility between the current target alteration assemblage and the typical metallogenic assemblage. A weight conversion function can be set on the original probability value, such as using a normalization function or logarithmic mapping, to improve the gradient response.

[0096] The spatiotemporal weight matrix is ​​a four-dimensional data structure, denoted as W(x, y, z, t), where x, y, and z represent the three-dimensional spatial coordinates and t represents the time dimension. Each cell records the combined weight value at a specific time and spatial location.

[0097] A gridded interpolation algorithm is used to perform three-dimensional spatial fitting on the spatiotemporal weight matrix. Within the spatial range defined by the three-dimensional distribution boundary model of the alteration body, a three-dimensional probability distribution model containing time attributes, spatial position and mineralization probability values ​​is generated as the spatiotemporal evolution probability field of the alteration body.

[0098] In the specific fitting process, commonly used spatial interpolation algorithms in geology, such as multivariate spline interpolation, inverse distance weighted interpolation (IDW) or Kriging interpolation, can be used to smoothly construct the mineralization probability distribution field and avoid local outliers interfering with the expression of the overall evolution trend.

[0099] Step S6: Based on the random forest algorithm, the mineralization confidence of the target area is calculated by integrating geological, geophysical, thermodynamic, remote sensing and geochemical indicators, and the target area prediction results and classification scores are output.

[0100] The acquisition of the mineralization confidence includes:

[0101] The time distribution factor, spatial density index and mineralization matching probability extracted from the spatiotemporal evolution probability field of the alteration body are used as core characteristic parameters, and the characteristic vector set is constructed by combining remote sensing image data, geochemical data and geological structure information.

[0102] Among them: the time distribution factor is derived from the overlapping density of genetic period labels within the target grid, reflecting the possibility that the location has experienced multiple alteration processes; the spatial density index is calculated from the voxel density generated by the three-dimensional distribution boundary model, and is used to characterize the enrichment of local alteration bodies; the mineralization matching probability is extracted from the probability value output by the Bayesian model; remote sensing image features are extracted from the principal component analysis results of alteration-sensitive bands such as NDVI, Al-OH, and Fe-OH; geochemical indicators include the abundance and anomaly level of target elements (such as Cu, Mo, Pb, Zn, As, etc.) in surface samples; geological structural characteristics include structural complexity evaluation parameters such as fracture intersection density and fold axis distribution intensity derived from structural measurement data;

[0103] The constructed feature vector set uses each spatial grid cell as the basic unit to form a training sample set. Data from typical mining areas with supervised information is used as training labels to train a random forest model. This model performs ensemble learning across multiple decision trees to improve classification stability and predictive generalization capabilities.

[0104] After training, the model is applied to all grid cells in the area to be predicted, and their mineralization potential is classified and evaluated, and the mineralization confidence value of each cell is output. The numerical range is generally set to the interval of 0~1, indicating the size of the mineralization probability.

[0105] The output target area prediction result includes:

[0106] Obtain the spatial distribution results of mineralization confidence, classify multiple grid cells in the target area according to the confidence threshold setting rules, and generate multiple candidate prospecting target cells;

[0107] The confidence threshold can be preset according to different prospecting task requirements. Common grading standards can be set as follows: [0.8–1.0] for the first-level target area, [0.6–0.8] for the second-level target area, [0.4–0.6] for the third-level target area, etc. The specific threshold can be dynamically adjusted in the system;

[0108] Adjacent units of the same confidence level are spatially aggregated and merged using a connectivity analysis method based on a spatial adjacency graph, outputting a prospecting target prediction layer with clear boundary ranges and level labels.

[0109] This layer can be displayed graphically on the system terminal interface and can be exported to standardized spatial data file formats (such as GeoTIFF, Shapefile, GDB, etc.) for results submission, map making and prospecting decision support.

[0110] Example 2, a prospecting target area prediction system based on altered mineral analysis, see Figure 1 As shown, it includes the following modules:

[0111] Data acquisition module, used to collect multi-source geological data of the target area;

[0112] Cluster analysis module, used to perform hierarchical cluster analysis on altered minerals and their paragenetic combination data, as well as mineral spectral features extracted from remote sensing image data, and to construct a time series decoupling model;

[0113] The alteration combination discrimination module calculates the matching probability between the target alteration combination and the typical metallogenic alteration pattern based on the metallogenic alteration combination probability discrimination model, and outputs the alteration body identification results with matching probability higher than the preset threshold;

[0114] The spatial modeling module is used to generate an implicit function field representing the boundary of the erosion body through an implicit geological modeling algorithm, and to extract a three-dimensional distribution boundary model of the erosion body;

[0115] The spatiotemporal coupling module is used to couple the genesis period labels, matching probabilities and the three-dimensional distribution boundary model of the erosion volume to generate the spatiotemporal evolution probability field of the erosion volume;

[0116] The mineralization assessment module is used to calculate the mineralization confidence of the target area and output the target area prediction results and classification scores.

[0117] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for predicting prospecting target areas based on altered mineral analysis, characterized in that: include: Collect multi-source geological data of the target area, including structural measurement data, remote sensing image data, geochemical data, altered minerals and their paragenesis data; Based on the constraints of mineral thermodynamic phase diagrams, hierarchical cluster analysis is performed on altered minerals and their paragenetic assemblage data, as well as mineral spectral features extracted from remote sensing image data. This generates a mineral paragenetic assemblage feature map, assigns genetic period labels, and constructs a temporal decoupling model. The temporal decoupling model is used as a priori knowledge base and input into a Bayesian network-based probability discrimination model for mineralization-alteration combinations together with elemental anomaly indicators in geochemical data. The matching probability between the target alteration combination and the typical mineralization-alteration pattern is calculated, and the identification results of alteration bodies with matching probabilities higher than a preset threshold are output. Extract the spatial distribution trend characteristics of the alteration body identification results, use the fault zone strike information extracted from the structural measurement data as the spatial constraint condition, and construct the input data set; An implicit geological modeling algorithm is used to generate an implicit function field representing the boundary of the alteration volume, and a three-dimensional distribution boundary model of the alteration volume is extracted; Establish a spatiotemporal weight matrix, couple the genetic period label, matching probability, and the three-dimensional distribution boundary model of the alteration body, and generate the spatiotemporal evolution probability field of the alteration body, which serves as the basic data input for mineralization confidence calculation and target area prediction; Based on the random forest algorithm, the mineralization confidence of the target area is calculated by integrating geological, geophysical, thermodynamic, remote sensing and geochemical indicators, and the target area prediction results and classification scores are output.

2. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The hierarchical cluster analysis comprises: The altered minerals are preliminarily grouped based on the stable domain characteristics of the thermodynamic phase diagram of the altered minerals and their paragenetic combination data; the mineral spectral characteristics extracted from the remote sensing image data are fused with the altered mineral grouping results through a preset weight coefficient; the fused data are input into the clustering model for hierarchical clustering processing; based on the grouping results output by the clustering model, the typical mineral assemblages and their paragenetic relationships in each group are extracted to construct the corresponding mineral paragenetic combination characteristic map.

3. The method for predicting prospecting target areas based on altered mineral analysis according to claim 2, characterized in that: The fusion of the mineral spectral features extracted from the remote sensing image data with the altered mineral grouping results by using a preset weight coefficient includes: Multiple characteristic reflection bands and spectral morphological parameters of altered minerals are extracted from remote sensing image data. The spectral characteristics are vectorized and encoded and combined with the altered minerals and their paragenetic combination data to form an input dataset. The input dataset is used as the input of the clustering model to perform hierarchical clustering processing.

4. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The construction of the genesis period label includes: Based on the characteristic map of mineral paragenesis and combined with the stable domain interval of the thermodynamic phase diagram of each group of altered minerals, a period discrimination rule is established; and according to the said rule, a unique genetic period label is assigned to each group of altered mineral combinations to characterize the alteration time attribute corresponding to the combination.

5. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The probability discrimination model for mineralization-alteration combinations is a Bayesian network model constructed using typical mineralization-alteration combination samples. The model uses the genetic period labels generated in the temporal decoupling model as prior input nodes and combines the element anomaly indicators in the geochemical data to jointly construct the input structure. The matching probability between the target alteration combination and the typical mineralization alteration pattern is used as the output node, and the discrimination results are screened based on the preset probability threshold in the output stage, and the alteration body identification results with matching probability higher than the threshold are output.

6. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The extraction of the three-dimensional distribution boundary model of the erosion body includes: The spatial position and distribution trend characteristics of the alteration body identification results are obtained, and the fault zone attitude information extracted from the structural measurement data is used as a spatial constraint condition to construct a modeling data set including a spatial point set, structural control lines and distribution trend vectors. Based on the modeling data set, an implicit geological modeling algorithm is used to generate an implicit function field, and the three-dimensional distribution boundary model of the alteration body is obtained by extracting the function isosurface.

7. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The generating of the spatiotemporal evolution probability field of the erosion body comprises: The genetic period labels, the matching probability of the mineralization-alteration combination, and the three-dimensional distribution boundary model of the alteration body are used as input variables to construct a spatiotemporal weight matrix, in which the temporal compatibility weight is assigned according to the chronological order of the genetic period labels, with the weight of the early-stage alteration combination decreasing; the spatial continuity weight is calculated based on the spatial density gradient in the three-dimensional distribution boundary model of the alteration body, with the weight of the high-density area increasing; and the mineralization matching probability weight is directly based on the matching probability value output by the Bayesian network. A gridded interpolation algorithm is used to perform three-dimensional spatial fitting on the spatiotemporal weight matrix. Within the spatial range defined by the three-dimensional distribution boundary model of the alteration body, a three-dimensional probability distribution model containing time attributes, spatial position and mineralization probability values ​​is generated as the spatiotemporal evolution probability field of the alteration body.

8. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The acquisition of the mineralization confidence includes: The time distribution factor, spatial density index and mineralization matching probability extracted from the spatiotemporal evolution probability field of the alteration body are used as core feature parameters, and a feature vector set is constructed in combination with remote sensing image data, geochemical data and geological structure information. A random forest model is trained based on the feature vector set and applied to each grid cell in the target area. The mineralization confidence value of each cell is output as the probability scoring basis for target area prediction.

9. The method for predicting prospecting target areas based on altered mineral analysis according to claim 1, characterized in that: The output target area prediction result includes: Obtain the spatial distribution results of mineralization confidence, classify multiple grid cells in the target area according to the confidence threshold setting rules, and generate multiple candidate prospecting target cells; spatially aggregate adjacent cells of the same category confidence level, and output a prospecting target prediction layer with clear boundary range and level labels.

10. A prospecting target area prediction system based on altered mineral analysis, characterized in that: The system applies a method for predicting prospecting target areas based on altered mineral analysis as described in any one of claims 1 to 9, comprising: Data acquisition module, used to collect multi-source geological data of the target area; Cluster analysis module, used to perform hierarchical cluster analysis on altered minerals and their paragenetic combination data, as well as mineral spectral features extracted from remote sensing image data, and to construct a time series decoupling model; The alteration combination discrimination module calculates the matching probability between the target alteration combination and the typical metallogenic alteration pattern based on the metallogenic alteration combination probability discrimination model, and outputs the alteration body identification results with matching probability higher than the preset threshold; The spatial modeling module is used to generate an implicit function field representing the boundary of the erosion body through an implicit geological modeling algorithm, and to extract a three-dimensional distribution boundary model of the erosion body; The spatiotemporal coupling module is used to couple the genesis period labels, matching probabilities and the three-dimensional distribution boundary model of the erosion volume to generate the spatiotemporal evolution probability field of the erosion volume; The mineralization assessment module is used to calculate the mineralization confidence of the target area and output the target area prediction results and classification scores.

Images