Deep concealed ore body prediction method and system

Abstract

The invention relates to the technical field of mineral resource exploration, in particular to a deep concealed ore body prediction method and system. The method comprises the following steps: acquiring geological exploration data, geophysical data and lithologic physical property probability distribution data of a target area; constructing a three-dimensional geological model according to the geological exploration data; performing inversion calculation on the geophysical data by taking the three-dimensional geologic model and the lithologic physical property probability distribution data as constraints to obtain a three-dimensional geophysical model; fusing the three-dimensional geologic model and the three-dimensional geophysical model to obtain a three-dimensional geologic-geophysical model; and determining the mineralization probability of the deep concealed ore body according to the three-dimensional geology-geophysical model. The problem that in the prior art, deep hidden ore body prediction is not accurate enough is solved, the three-dimensional geology-geophysical model is gradually constructed by obtaining the geological exploration data, the geophysical data and the lithologic physical property probability distribution data of the target area to be explored, then ore body prediction is conducted, and the accuracy of ore body prediction is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mineral resource exploration, and particularly relates to a deep concealed ore body prediction method and system. BACKGROUND

[0002] The shallow covering area of grassland is an important potential area for mineral resource exploration, but the loose sediments and vegetation layer widely distributed on the surface form significant interference to traditional geological exploration methods. Neither surface geological mapping nor conventional geophysical surveying can effectively penetrate the shallow interference layer, resulting in great challenges in positioning the spatial location of deep concealed ore bodies, analyzing the ore-forming environment and depicting the ore-forming regularity.

[0003] Current prediction of deep concealed ore bodies in this area relies on a single technical path, or only uses geophysical inversion to infer underground physical property anomalies, or simply relies on geological reasoning to judge the ore-forming possibility. This single mode has obvious defects: geophysical inversion is prone to multiple solutions due to lack of geological background constraints, and the same physical property anomaly may correspond to multiple geological bodies; geological reasoning is difficult to accurately locate the range of deep ore bodies due to lack of quantitative physical property data support, and the information gap between the two leads to false anomaly misjudgment in the prediction process, resulting in low prediction accuracy and difficulty in meeting actual exploration needs.

[0004] More importantly, the existing technical system has the problem of paying more attention to unknown area detection and less attention to known data utilization: the measured data of the target area, including lithology combination, structure development characteristics, ore body and surrounding rock physical property differences, etc., are often only stored as independent data for archiving, and are not effectively converted into constraints for deep concealed ore body prediction, resulting in data resource waste and lack of reliable measured calibration basis in the prediction process.

[0005] Therefore, there is an urgent need for a more accurate deep concealed ore body prediction method. SUMMARY

[0006] In view of the defects in the prior art, the present application provides a deep concealed ore body prediction method and system, which solves the problem of inaccurate deep concealed ore body prediction in the prior art.

[0007] In order to achieve the above object, one aspect of the present application provides a deep concealed ore body prediction method, the method comprising: obtaining geological exploration data, geophysical data, lithology and physical property probability distribution data of a target area; constructing a three-dimensional geological model according to the geological exploration data; performing inversion calculation on the geophysical data to obtain a three-dimensional geophysical model by taking the three-dimensional geological model and the lithology and physical property probability distribution data as constraints; fusing the three-dimensional geological model and the three-dimensional geophysical model to obtain a three-dimensional geology-geophysics model; and determining a mineralization probability of a deep concealed ore body according to the three-dimensional geology-geophysics model.

[0008] The present application establishes a spatial framework by integrating exploration data, constructing a three-dimensional geological model from geological exploration data, combining lithology and physical property probability distribution data to constrain geophysical inversion, improving the accuracy of the three-dimensional geophysical model, forming a three-dimensional geology-geophysics model by fusing the two types of models, realizing accurate association of geological structure and physical property characteristics, and finally determining the mineralization probability based on the fused model, which can break through the limitations of single data, reduce false abnormal misjudgment, and improve the accuracy of deep concealed ore body prediction.

[0009] Optionally, the inversion calculation on the geophysical data to obtain a three-dimensional geophysical model by taking the three-dimensional geological model and the lithology and physical property probability distribution data as constraints comprises: extracting geological constraint conditions by using the three-dimensional geological model; calculating physical property probability constraint conditions according to the lithology and physical property probability distribution data; calculating a target function value based on the geological constraint conditions, the physical property probability constraint conditions and the geophysical data; and performing iteration to obtain a three-dimensional geophysical model by taking minimizing the target function value as the target.

[0010] The present application extracts geological constraints such as strata and structures from the three-dimensional geological model to clearly define the spatial boundaries, and then determines the reasonable interval of physical property parameters according to the lithology and physical property probability distribution data to provide double accurate constraints for inversion; the target function value is calculated based on the geophysical data, and iteration is performed by taking minimizing the value as the target, which can effectively avoid the multi-solution problem of unconstrained inversion, reduce false abnormalities caused by lack of geological background or physical property rules, and improve the rationality and scientificity of the three-dimensional geophysical model.

[0011] Optionally, the determination of the mineralization probability of a deep concealed ore body according to the three-dimensional geology-geophysics model comprises: determining an initial mineralization probability of a deep concealed ore body according to the three-dimensional geology-geophysics model; obtaining measured mineralization characteristics of known ore bodies in the target area; and correcting the initial mineralization probability by using the measured mineralization characteristics to obtain the mineralization probability.

[0012] The application lays the calculation foundation of the initial mineralization probability by the three-dimensional geological-geophysical model, guarantees the scientificity of the preliminary judgment, introduces the measured mineralization characteristics of the known ore body in the target area as a calibration basis to modify the initial probability, effectively makes up for the possible deviation of relying on model calculation only, and through the double links of initial calculation and measured modification, not only the advantages of the integrated model in comprehensively representing the deep geological and physical characteristics are exerted, but also the known data is used to avoid false abnormal misjudgment, so that the final mineralization probability is more in line with the actual geological conditions, and the accuracy and reliability of the deep concealed ore body prediction are greatly improved.

[0013] Optionally, the determining the initial mineralization probability of the deep concealed ore body according to the three-dimensional geological-geophysical model comprises: obtaining original geochemical exploration data and original remote sensing alteration data of the target area; preprocessing the original geochemical exploration data and the original remote sensing alteration data to obtain geochemical anomaly data and remote sensing alteration data; screening mineralization variables according to the three-dimensional geological-geophysical model, the geochemical anomaly data and the remote sensing alteration data; and calculating the initial mineralization probability of the deep concealed ore body by using the mineralization variables and the three-dimensional geological-geophysical model.

[0014] The application extracts the geochemical anomaly and remote sensing alteration data reflecting the mineralization potential by introducing and preprocessing the original geochemical exploration and remote sensing alteration data, screens the mineralization variables in combination with the three-dimensional geological-geophysical model, realizes the cooperation of multi-source data, avoids the limitation of single model information, and finally calculates the initial probability relying on the mineralization variables and the integrated model, which not only integrates the multi-dimensional mineralization clues of geology, physical property, geochemical exploration and remote sensing, but also reduces the interference information through data preprocessing and variable screening, and improves the accuracy of the initial mineralization probability calculation.

[0015] Optionally, the preprocessing the original geochemical exploration data and the original remote sensing alteration data to obtain geochemical anomaly data and remote sensing alteration data comprises: denoising and enhancing the original geochemical exploration data to obtain the geochemical anomaly data; and screening the original remote sensing alteration data to obtain the remote sensing alteration data.

[0016] The application can remove interference information and enhance geochemical anomalies related to mineralization by denoising the original geochemical exploration data, ensures that the obtained geochemical anomaly data is more accurate, and can exclude non-mineralization related redundant alteration information and retain effective alteration data by screening the original remote sensing alteration data, thereby improving the scientificity of the geochemical anomaly data and the remote sensing alteration data.

[0017] Optionally, the denoising and enhancing the original geochemical exploration data to obtain the geochemical anomaly data comprises: extracting active state content and total state content in the original geochemical exploration data; calculating the ratio of the active state content to the total state content; and delineating the geochemical anomaly data based on the spatial distribution of the ratio.

[0018] The present application can accurately represent the potential of ore-forming elements participating in mineralization by extracting the active content reflecting the element activity and the total content reflecting the total reserves in the original geochemical data, and calculating the ratio of the two, and then delineating the anomaly according to the spatial distribution of the ratio, which not only avoids the problem that single content data is easily disturbed, but also highlights the areas with high activity and high ore-forming potential, effectively eliminates false anomalies unrelated to mineralization, and improves the reliability of geochemical anomaly data.

[0019] Optionally, the screening of the original remote sensing alteration data to obtain remote sensing alteration data comprises: extracting bedrock alteration mineral information in the original remote sensing alteration data by using a spectral mixture decomposition technology; extracting a potential ore-forming target area by using the three-dimensional geological-geophysical model; and performing spatial superposition analysis on the bedrock alteration mineral information and the potential ore-forming target area, and extracting the bedrock alteration mineral information in the potential ore-forming target area to obtain remote sensing alteration data.

[0020] The present application can accurately extract bedrock alteration mineral information in the original remote sensing alteration data by using a spectral mixture decomposition technology, exclude non-bedrock interference, and then rely on a three-dimensional geological-geophysical model to lock a potential ore-forming target area, and determine the favorable spatial range of mineralization. Finally, through spatial superposition analysis, only the bedrock alteration information in the target area is retained, which not only ensures the geological correlation of the alteration data, but also focuses on the mineralization potential area, effectively eliminates redundant alteration information unrelated to mineralization, and improves the reliability of the remote sensing alteration data.

[0021] Optionally, the correction of the initial ore-forming probability to obtain an ore-forming probability by using the measured ore-forming characteristics comprises: setting a confidence level for the measured ore-forming characteristics; extracting simulated ore-forming characteristics in the three-dimensional geological-geophysical model; calculating the similarity of the simulated ore-forming characteristics and the measured ore-forming characteristics; and correcting the initial ore-forming probability based on the similarity and the confidence level to obtain an ore-forming probability.

[0022] The present application sets a confidence level for the measured ore-forming characteristics to determine the data reliability benchmark, then extracts simulated ore-forming characteristics from the fusion model to establish a comparison basis with the measured characteristics, quantifies the matching degree of the two by calculating the similarity, and finally corrects the initial probability by combining the similarity and the confidence level. This process not only activates the calibration value of the measured data of known ore bodies, but also reduces the model calculation deviation through multi-dimensional comparison, effectively avoids false anomaly misjudgment relying only on model calculation, and improves the accuracy of the ore-forming probability.

[0023] Optionally, the calculating the similarity of the simulated ore-forming feature and the measured ore-forming feature comprises: constructing an ore-non-ore classification model based on the measured ore-forming feature and the confidence; determining a sub-feature weight of the measured ore-forming feature according to the ore-non-ore classification model; calculating a sub-feature similarity according to the simulated ore-forming feature and the measured ore-forming feature; and obtaining the similarity by weighted summation of the sub-feature similarity based on the sub-feature weight.

[0024] The present application can make the model fit the actual ore-forming regularity by training the classification model with the measured ore-forming feature and the confidence, determine the sub-feature weight by the model, highlight the influence of the key sub-feature on ore-forming, then calculate the sub-feature similarity and combine the weight for weighted summation, which can not only avoid the interference of single sub-feature deviation, but also strengthen the role of the core sub-feature, thereby improving the accuracy of similarity calculation.

[0025] Another aspect of the present application also provides a deep concealed ore body prediction system, comprising: a processor, an input device, an output device and a memory, which are connected with each other, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to invoke the program instructions to execute the deep concealed ore body prediction method of any one of the previous aspect of the present application.

[0026] The deep concealed ore body prediction system of the present application has compact structure, stable performance, high integration and simple constitution, can stably execute the deep concealed ore body prediction method of the previous aspect of the present application, and further improves the overall applicability and practical application ability of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 A deep concealed ore body prediction method flow chart of an embodiment of the present application; Figure 2 A deep concealed ore body prediction system structure schematic diagram of an embodiment of the present application. DETAILED DESCRIPTION

[0028] The specific embodiments of the present application will be described in detail below, and it should be noted that the embodiments described herein are only used for illustration and do not limit the present application. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present application. However, it is obvious to those skilled in the art that the present application does not have to be implemented with these specific details. In other examples, in order to avoid obscuring the present application, well-known circuits, software or methods are not specifically described.

[0029] Throughout this specification, references can be made to "one embodiment", "an embodiment", "one example" or "an example" meaning that a particular feature, structure, or characteristic described in connection with the embodiment or example is included in at least one embodiment of the application. The appearances of the phrases "in one embodiment", "in an embodiment", "one example" or "an example" in various places in the specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics can be combined in any suitable

[0030] Please refer to Figure 1 In order to solve the problems in the prior art, in an optional embodiment, a deep concealed ore body prediction method as shown in the figure comprises the following steps: Step S1, obtaining geological exploration data, geophysical data, lithology and physical property probability distribution data of the target area.

[0031] In this embodiment, the geological exploration data is basic information reflecting the geological features of the target area on the ground and underground, which is mainly obtained through field investigation, drilling sampling and laboratory analysis. In the field work, the geologists will record the stratum distribution on the ground, the lithology of the rock outcrop and the structural features through geological mapping, and use GPS to locate the key geological points. For underground information, drill holes are drilled through drilling, and rock core samples at different depths are collected, and drill columnar graphs and logging data are recorded. These raw data will be combined with the experimental results of rock and mineral composition detection and fossil identification to form systematic exploration data, including not only intuitive drawings such as geological maps and profiles, but also quantitative data such as drill data tables and core analysis reports.

[0032] Geophysical data is information obtained by detecting the physical property differences of different geological bodies underground. The core is to indirectly detect underground structures by using physical methods, and to obtain it through geophysical prospecting technology. Geophysical prospecting technology includes seismic exploration, gravity exploration, magnetic exploration and electrical exploration. Among them, seismic exploration detects reflected signals by artificially exciting seismic waves and using geophones to receive them, and obtains seismic profiles reflecting stratum interfaces and structures. Gravity exploration uses gravimeters to measure changes in the surface gravity field to capture underground density anomalies. Magnetic exploration uses magnetometers to record small changes in the geomagnetic field to identify geological bodies containing magnetic minerals. Electrical exploration measures potential difference by injecting current through electrodes to invert underground resistivity distribution.

[0033] The lithology and physical property probability distribution data is statistical information recording the correlation between various types of rocks and physical parameters in the target area, and is presented in the form of correlation between lithology type, physical property interval and occurrence probability. Samples are taken from drill cores, and core physical parameters such as density, magnetic susceptibility and resistivity of different lithologies are tested in the laboratory. For example, the density of 100 granite samples is measured, and the proportion falling within the range of 2.6 g / cm 3 -2.7 g / cm 3 is counted. Meanwhile, the physical property records in the existing logging data and regional exploration reports are sorted out, and the parameter ranges of different lithologies are supplemented. Then, the data is analyzed by using nuclear density estimation or Bayesian algorithm to obtain the probability distribution characteristics of the physical parameters corresponding to each lithology.

[0034] Step S2, constructing a three-dimensional geological model according to the geological exploration data.

[0035] In this embodiment, the stratigraphic boundaries and structure lines on the geological map in the geological exploration data are calibrated by using GPS coordinates, the lithology records at different depths (such as sandstone at 500 meters underground and granite at 800 meters) in the drill hole data and the stratigraphic thickness data are sorted into a standardized table, and the stratigraphic contact relationship (such as parallel unconformity and angular unconformity) on the geological profile is clearly marked to ensure that the coordinate systems and depth references of all data are consistent. Then, the model framework is built based on these data, the interfaces of each stratum are stretched in the three-dimensional space according to the depth information of the drill hole and the stratigraphic relief of the profile, for example, the points of all sandstone outcrops on the surface are connected, and then the continuous sandstone top and bottom three-dimensional surfaces are formed by extending downward according to the sandstone thickness recorded by the drill hole. Then, the space enclosed by each stratigraphic interface is filled with lithology attributes, for example, the quartz sandstone is marked in the area between the sandstone interfaces, and the mineral composition, color and other characteristics are attached. Then, the structure elements are added, for example, the three-dimensional shape of the fault is drawn in the model according to the strike, dip and throw of the fault to simulate its cutting effect on the stratum, for example, the upthrown side of the fault is 200 meters higher than the downthrown side, which is reflected in the model. Finally, these elements are integrated by using professional three-dimensional modeling software to generate a three-dimensional model that can intuitively display the lithology distribution, stratigraphic relief and structure shape at different depths underground. The drill hole data not involved in the modeling is used for verification, for example, the lithology at 1000 meters underground in the model is checked to see whether it is consistent with the actual core of the new drill hole, and the slight errors are adjusted to make the model more consistent with the real geological conditions of the target area.

[0036] Step S3, performing inversion calculation on the geophysical data by using the three-dimensional geological model and the lithology and physical property probability distribution data as constraints to obtain a three-dimensional geophysical model.

[0037] The inverse calculation of the geophysical data with the three-dimensional geological model and the lithological physical property probability distribution data as constraints to obtain a three-dimensional geophysical model specifically includes the following sub-steps: In step S301, a geological constraint condition is extracted by using the three-dimensional geological model.

[0038] In this embodiment, the geological constraint condition includes a spatial range constraint of strata and lithology and a structural form and boundary constraint.

[0039] The spatial range constraint of strata and lithology is extracted from the three-dimensional geological model. The model has calibrated the top and bottom plate depths, lateral extension boundaries of each set of strata, and spatial distribution ranges of ore-bearing lithology and surrounding rock by drilling and geological mapping data. Based on this constraint, it can be determined that the physical property parameters of a certain depth interval during inversion should correspond to the characteristics of sandstone, and the physical properties of the ore-bearing lithology distribution area should preferentially match the parameter range of the granodiorite, thereby avoiding confusing the physical properties of different strata.

[0040] The structural form and boundary constraint is converted from the faults, folds and other structures described in the three-dimensional geological model. For example, the lateral width of the fault fracture zone needs to be controlled within 50m-80m, the resistivity of the region during inversion needs to meet the low-resistance characteristics of fault gouge, the dip angle of the strata at the fold wing needs to be consistent with the calibrated 30°-45° in the model, and the lateral variation of the physical property parameters needs to be adjusted synchronously with the strata occurrence, ensuring that the physical property distribution is consistent with the strata dip direction and preventing the inversion from obtaining a physical property distribution that contradicts the structural form.

[0041] In step S302, a physical property probability constraint condition is calculated according to the lithological physical property probability distribution data.

[0042] In this embodiment, the physical property probability constraint condition is calculated according to the lithological physical property probability distribution data. The core is to convert the statistical correlation of lithology, physical property and probability into quantifiable parameter limits during geophysical inversion.

[0043] The basic statistical information of the physical property parameters corresponding to each lithology is extracted, and the core physical property parameters corresponding to each main lithology in the target area are selected from the lithology physical property probability distribution data, such as density, resistivity, and magnetic susceptibility, the high-probability value interval of each physical property parameter under each lithology is determined, usually the interval with a probability greater than 60% is selected, and the probability value in the interval is converted into a corresponding weight, that is, the higher the probability, the greater the weight, representing the higher priority of selecting the interval parameter during inversion, and the basic value requirement of the physical property parameter corresponding to different lithologies is clear. Then, the joint probability constraint of single lithology and multiple parameters is established, and the joint probability between multiple physical property parameters of the same lithology is calculated according to the probability distribution data, that is, the probability that multiple physical property parameters fall in the high-probability interval at the same time, and the parameter combination with a joint probability greater than 30% is defined as an effective joint interval (this threshold can be determined based on the lithology physical property statistical data of the target area or the ore-forming experience of similar deposits), which ensures that the combination of multiple physical property parameters in a lithology area during inversion will not appear a single parameter meeting the requirements but the parameters contradicting each other, and the rationality of the constraint is strengthened. Finally, the binding of the probability constraint and the spatial position is realized, and the calibrated lithology spatial distribution in the three-dimensional geological model is combined, that is, the lithology type corresponding to different spatial positions, the high-probability physical property interval, the parameter weight and the effective joint interval of the corresponding lithology are matched to the spatial position one by one, forming exclusive constraint conditions for lithology, spatial position and physical property parameters, ensuring that the selection of physical property parameters in each area during inversion meets the probability statistical law of the corresponding lithology, avoiding the selection of abnormal parameter values with extremely low probability, and reducing the inversion multi-solution.

[0044] In step S303, a target function value is calculated based on the geological constraint condition, the physical property probability constraint condition and the geophysical data.

[0045] In this embodiment, the core function of the target function is to quantify the fitting degree of the geophysical inversion result with the measured data, the geological rule and the physical property statistical law. When calculating, the error or coincidence degree of the three input items of the geological constraint condition, the physical property probability constraint condition and the geophysical data is integrated to obtain the total target function value.

[0046] First, the geophysical data fitting term is calculated, taking the measured gravity anomaly value, the measured resistivity value and other geophysical data as the benchmark, to calculate the deviation between the inversion simulated geophysical response value and the measured value. The smaller the deviation, the smaller the value of the fitting term, representing the higher the fitting degree of the inversion result and the measured data. Then, the prior model constraint term is calculated, comparing the spatial range constraint of the stratum and lithology, the constraint of the structural form and boundary, quantifying the deviation of the inversion physical property distribution from these constraints, and the smaller the deviation, the smaller the value of the fitting term. Then, the physical property probability constraint compliance term is calculated, according to the corresponding lithology high probability physical property interval and effective joint interval of each spatial position, to calculate the compliance degree of the inversion physical property value falling within the high probability interval, such as falling within the interval with a probability greater than 60% for high compliance, and falling within the interval with a probability less than 5% for low compliance, and converting the compliance degree into an error value. The lower the compliance degree, the larger the error value. Finally, the smoothing term is calculated, to avoid the inversion result from appearing similar to the abrupt fluctuation without geological significance, such as the abrupt fluctuation of the physical property value of adjacent grids. The smaller the difference between adjacent spatial grids, the smaller the value of the smoothing term, to ensure that the inversion physical property distribution meets the continuity characteristics of the underground geological body. Then, the geophysical data fitting term, the prior model constraint term, the physical property probability constraint term and the smoothing term are weighted and summed according to the preset weight. The preset weight can be adjusted according to the geological complexity of the target region. The final sum is the target function value. The smaller the value, the better the comprehensive effect of the current inversion result meeting the measured data, the geological law and the physical property statistical law.

[0047] The target function value satisfies the following formula: wherein, is the target function value of the inversion model parameter vector , is a geophysical forward operator, is geophysical anomaly data, is a data weight matrix, is a smoothing operator acting on the result of the inversion model parameter vector , is a geological body boundary indication matrix, is an initial physical property estimate value corresponding to a three-dimensional geological model, is the total number of grid cells obtained after discretization of the target region, is the physical property value of the i-th grid cell , is the probability density of the i-th grid cell under its corresponding lithology , , , is a weight parameter.

[0048] In the above formula, a geophysical data fitting term, a smoothing term, a prior model constraint term, a petrophysical probability constraint term, 、 、 To balance the contributions of the data fitting term, the smoothing term, the geological constraint term, and the petrophysical probability constraint term, the data quality, the geological complexity, and the inversion stability experience of the target area can be set empirically, or the cross-validation method can be used to determine.

[0049] Step S304: iteratively obtaining a three-dimensional geophysical model by minimizing the objective function value.

[0050] In this embodiment, the three-dimensional geophysical model is obtained by iteratively optimizing the model parameter vector to minimize the objective function value. Based on the initial petrophysical estimate or experience value of the three-dimensional geological model, the initial petrophysical parameters of each grid underground are determined, and then the gradient descent method is used to calculate the gradient or search direction of the objective function with respect to the model parameter vector. The petrophysical parameter values of each grid in the model parameter vector are adjusted by a predetermined step size. After each adjustment, the objective function value is recalculated to determine whether the convergence condition is met, such as the change in the objective function value being less than a threshold value or the number of iterations reaching an upper limit. If the convergence condition is not met, the parameter adjustment and function calculation process are repeated until the convergence condition is met. In each iteration, the parameter update of each step must ensure that the petrophysical parameters of each grid meet the requirements of geophysical data fitting, geological constraints, petrophysical probability constraints, and model smoothing. The model parameter vector at the final convergence defines the petrophysical distribution of each grid underground, and the three-dimensional geophysical model is obtained by combining the petrophysical distribution.

[0051] Step S4: fusing the three-dimensional geological model and the three-dimensional geophysical model to obtain a three-dimensional geological-geophysical model.

[0052] In this embodiment, by unifying the spatial coordinate system and carrying out multi-source data coupling analysis, the qualitative spatial framework of the stratigraphic interface, lithology distribution, and structural form in the three-dimensional geological model is associated and matched with the quantitative distribution of petrophysical parameters such as density and resistivity in the three-dimensional geophysical model. Among them, the geostatistical co-Kriging method or the machine learning fusion algorithm is used to constrain the lithology unit and the structural boundary of the geological model, and the petrophysical anomaly of the geophysical model is assigned and optimized according to the lithology-petrophysical probability relationship. At the same time, the spatial form of the concealed geological body in the geological model is refined by the petrophysical gradient characteristics of the geophysical model, and finally a three-dimensional geological-geophysical model that combines qualitative description of geological structure and quantitative characterization of geophysical petrophysical is generated.

[0053] Step S5: determining the ore-forming probability of the deep concealed ore body according to the three-dimensional geological-geophysical model.

[0054] The determining of the ore-forming probability of the deep concealed ore body according to the three-dimensional geological-geophysical model specifically comprises the following steps: In step S501, an initial ore-forming probability of a deep concealed ore body is determined according to the three-dimensional geological-geophysical model.

[0055] The determining of the initial ore-forming probability of the deep concealed ore body according to the three-dimensional geological-geophysical model specifically comprises the following steps: In step S50101, original geochemical data and original remote sensing alteration data of a target area are obtained.

[0056] In this embodiment, the sampling range and the point density are determined in combination with the regional geological background. In the field, soil, rock, stream sediment and other samples are collected by using a special tool. Each sample is labeled with information such as sampling coordinates, sampling depth and medium type. After being packaged according to the specification, the samples are sent to the laboratory. In the laboratory, the content of target elements in the samples is detected by using instruments such as ICP-MS (inductively coupled plasma mass spectrometry) and atomic absorption spectrometry. The analysis accuracy is recorded at the same time. Finally, the original geochemical data including the sampling position, medium type, element content and analysis parameters are obtained.

[0057] According to the research requirements, a remote sensing data source is selected. The original remote sensing image file of the corresponding area is downloaded through a remote sensing data platform. Then, the original image is sequentially subjected to radiation calibration, atmospheric correction and geometric precision correction by using remote sensing processing software such as ENVI and ERDAS. Based on the characteristic spectrum of mineralization alteration minerals, the alteration information is extracted by using methods such as spectral angle matching and principal component analysis. Finally, the original remote sensing alteration data including the alteration anomaly raster data, the alteration mineral spectral curve and the geographic coordinates of the alteration area are obtained.

[0058] In step S50102, the original geochemical data and the original remote sensing alteration data are preprocessed to obtain geochemical anomaly data and remote sensing alteration data.

[0059] The preprocessing of the original geochemical data and the original remote sensing alteration data to obtain the geochemical anomaly data and the remote sensing alteration data specifically comprises the following steps: In step S5010201, the original geochemical data is denoised and enhanced to obtain the geochemical anomaly data.

[0060] The denoising and enhancement of the original geochemical data to obtain the geochemical anomaly data comprises: In step S501020101, the active state content and the total state content are extracted from the original geochemical data.

[0061] In the embodiment, the active content refers to the content of the ore-forming element in the medium such as soil, rock, stream sediment, and the like, in a form that can be migrated and involved in geochemical action. Common forms include ionic state, adsorbed state, weakly bound state, and the like. The element in such a form can be separated by a weak extractant without destroying the mineral lattice, and the content can reflect the recent geochemical activity and ore-forming potential of the target area. The total content refers to the total content of the ore-forming element in all forms in the medium, which is obtained by detecting the element after completely releasing all forms of the element by a strong digestion method. The content can reflect the total reserves of the element in the medium, and is a basic index for judging the enrichment degree of the element in the region.

[0062] First, the core ore-forming element is determined according to the ore-forming type of the target area. Then, in the structured storage system of the original geochemical exploration data, the active content data and the total content data of each ore-forming element are screened out by field identification matching (such as “element name + active concentration” and “element name + total content concentration”), and the supporting metadata information in the original data, including sampling coordinates, sample medium type, detection precision parameters, and analysis batch, needs to be associated to ensure that each set of active content and total content data can accurately correspond to the spatial position and sample attribute of the target area. Finally, the extracted data is subjected to validity verification to exclude blank values and abnormal values beyond the reasonable range, to form a structured data set corresponding one-to-one to the ore-forming element, active content, total content, spatial attribute, and detection information.

[0063] Step S501020102, calculating the ratio of the active content to the total content.

[0064] In the embodiment, the ratio of the active content to the total content is calculated, that is, the ratio of the two is obtained by dividing the active content by the total content. The core significance of the ratio is to reflect the active part of the ore-forming element that can be migrated and involved in ore-forming action, accounting for the proportion of the total reserves in the medium. The higher the ratio, the stronger the geochemical activity of the ore-forming element in the region, and the more likely the occurrence of mineralization enrichment or secondary leaching in the near future, which can assist in distinguishing primary geochemical anomalies from secondary mineralization anomalies.

[0065] Step S501020103, delineating the geochemical anomaly data based on the spatial distribution of the ratio.

[0066] In the embodiment, the ratio value is associated with the spatial coordinates of the corresponding sampling points to construct a spatial database of the ratio value-spatial position correspondence. Then, a GIS spatial analysis tool (such as Kriging interpolation, inverse distance weighted interpolation) is used to convert the ratio value data of the discrete sampling points into a continuous ratio value spatial distribution map. Subsequently, a statistical method (such as the 95% quantile method) is used to determine the ratio value anomaly threshold. The area with a ratio value higher than the threshold in the spatial distribution map is preliminarily identified as an anomaly candidate area. Further, in combination with the geological background of the target area, false anomalies caused by human pollution and non-mineralization geological action are removed, and the area with a continuous spatial distribution characteristic and superimposed with a mineralization favorable geological body is retained. Finally, the geochemical anomaly data capable of reflecting the high activity of the ore-forming element and indicating the mineralization potential are delineated, and the boundary range, spatial form and ratio value high value center of the anomaly area are determined.

[0067] Step S5010202, screening the original remote sensing alteration data to obtain remote sensing alteration data.

[0068] Specifically, the screening of the original remote sensing alteration data to obtain remote sensing alteration data includes the following sub-steps: Step S501020201, using a spectral mixture decomposition technology to extract bedrock alteration mineral information in the original remote sensing alteration data.

[0069] In the embodiment, the original remote sensing alteration data is preprocessed. The noise is removed through smoothing filtering, and the illumination difference is eliminated through spectral normalization to ensure that the spectral curve can truly reflect the ground object composition. Then, in combination with the known geological data of the target area and the standard spectral library, the pure end member is screened from the original data by using the pure pixel index and N-FINDR algorithm, including the bedrock end member, the target alteration mineral end member and the interference end member. Subsequently, the linear spectral mixture decomposition model is used to decompose the spectral reflectivity of each pixel in the original data into a weighted combination of each end member spectrum. The abundance value of the bedrock, alteration mineral and interference component in each pixel, i.e. the proportion of each component in the pixel, is calculated. Finally, the bedrock distribution area is focused, the alteration mineral abundance threshold (such as removing the low reliability area with an abundance less than 5%) is set, the non-bedrock interference area such as vegetation and water body is removed, and the spatial distribution information of each target alteration mineral in the bedrock area is extracted. The type, abundance high value area and spatial correlation with the bedrock of the alteration mineral are determined, such as being hosted in the bedrock fissure zone and the contact surface between the bedrock and the surrounding rock.

[0070] Step S501020202, using the three-dimensional geological-geophysical model to extract a potential ore-forming target area.

[0071] In this embodiment, based on the core ore-controlling elements of the target area screening based on the three-dimensional geological-geophysical model, the three-dimensional spatial distribution range of the ore-bearing stratum, the trend, tendency and deep extension form of the ore-controlling structure are extracted from the geological structure dimension, and the geological environment corresponding to the known ore points or mineralization points is labeled, from the geophysical physical property dimension, combined with the typical physical property characteristics of the target mineral, the physical property anomaly threshold is set, the physical property anomaly body meeting the threshold requirement is extracted from the three-dimensional geological-geophysical model, and the three-dimensional boundary, size and spatial superposition relationship with the geological body are determined. Subsequently, through three-dimensional spatial superposition analysis, the geological favorable area formed by the ore-bearing stratum combined with the ore-controlling structure is matched with the physical property anomaly body in space, and the area where the geological favorable area and the physical property anomaly body are highly superimposed is screened out, and further through comparison of the physical property difference between the anomaly body and the surrounding rock in the model, combined with the physical property background value of the stratum lithology, it is verified whether the anomaly is related to mineralization, and non-mineralization interference anomaly is excluded. Finally, according to the spatial coincidence ratio of geology and physical property anomaly, the amplitude of the physical property parameter deviating from the background value, the three-dimensional volume of the anomaly body and the similarity with the known mineralization, the candidate area is graded, the potential metallogenic target area is delineated, and the three-dimensional spatial coordinate range, core high value area position and main ore-controlling elements of each metallogenic target area are determined.

[0072] Step S501020203, spatially superimposing the bedrock alteration mineral information and the potential metallogenic target area, and extracting the bedrock alteration mineral information in the potential metallogenic target area to obtain remote sensing alteration data.

[0073] In this embodiment, the alteration mineral information and the potential metallogenic target area are unified to the same geographic coordinate system, ensuring accurate spatial position matching of the two, and then with the help of GIS spatial analysis tools, the bedrock alteration mineral information is cropped and screened with the spatial boundary of the potential metallogenic target area as the constraint range, and the bedrock alteration data completely or partially falling within the target area is extracted, including the alteration mineral species, the abundance value and the spatial distribution characteristics of the corresponding mineral at each position in the target area, while excluding irrelevant alteration information outside the target area, and verifying the spatial consistency of the extracted data, finally forming structured remote sensing alteration data corresponding one-to-one between the potential metallogenic target area, the bedrock alteration mineral type in the target area, the alteration intensity and the spatial coordinates.

[0074] Step S50103, screening metallogenic variables according to the three-dimensional geological-geophysical model, the geochemical anomaly data and the remote sensing alteration data.

[0075] In this embodiment, the three-dimensional geological-geophysical model is used as the core to extract the basic variables, and then the geochemical and remote sensing data are used for auxiliary verification and elimination of redundancy. First, the core ore-forming variables are extracted from the three-dimensional geological-geophysical model. The geological dimension focuses on the depth, thickness, lateral extension length of the ore-bearing stratum, and the spatial distance from the ore-controlling structure, representing the characteristics of the mineralization hosting carrier. At the same time, the dip angle of the fault zone, the width of the fracture zone, the depth and integrity of the fold axis of the ore-controlling structure are extracted, reflecting the effectiveness of the ore fluid migration channel. The geophysical dimension selects the average value of the density / resistivity of the physical property anomaly body, the difference value from the background value, the three-dimensional volume and the length of the strike, representing the response strength of the mineralization physical property, as well as the density gradient zone, the resistivity sudden change zone and other physical property gradient values, indicating the potential mineralization enrichment interface. Subsequently, the geochemical anomaly data and the remote sensing alteration data are used as auxiliary screening basis to verify the mineralization correlation of the core variables. For example, if a certain ore-controlling fault in the core variable is not covered by the geochemical anomaly area or there is no remote sensing alteration response in the corresponding area, it is evaluated whether it is a non-mineralization related redundant variable. If the physical property anomaly body in the core variable is highly overlapped with the geochemical anomaly and remote sensing alteration area, its effectiveness as an ore-forming variable is strengthened. Finally, combined with the ore-controlling regularity of the target mineral, the core variables verified by the auxiliary verification are integrated, and the non-mineralization related redundant parameters are eliminated, such as isolated physical property anomalies without geochemical / remote sensing support and non-mineralization section ore-bearing stratum parameters. Finally, a multi-dimensional ore-forming variable combination is formed, which is mainly based on the three-dimensional geological-geophysical model parameters and verified by geochemical anomaly data and remote sensing alteration data.

[0076] Step S50104, using the ore-forming variables and the three-dimensional geological-geophysical model to calculate the initial ore-forming probability of the deep concealed ore body.

[0077] In this embodiment, first, the weight of each ore-forming variable is determined according to the ore-controlling regularity of the target mineral (such as the correlation between stratum, structure, physical property and mineralization) by the analytic hierarchy process (AHP) and expert scoring method. Then, the three-dimensional geological-geophysical model is used as a spatial framework to divide the model into a plurality of three-dimensional grid units. The actual values of each ore-forming variable corresponding to each simulation unit are extracted and standardized. Then, the standardized variable values of each unit are multiplied by the corresponding weight and then accumulated to obtain the unit ore-forming potential index. Finally, through normalization processing, the potential index is converted into an initial ore-forming probability in the range of 0-1 (the higher the index, the greater the probability), which represents the possibility of the existence of a concealed ore body in each deep grid unit.

[0078] Step S502, obtaining the measured ore-forming characteristics of the known ore body in the target area.

[0079] In the embodiment, based on the measured data of the known ore body in the target area, the measured unit body divided by the three-dimensional grid is taken as the basic unit, and the measured ore-forming characteristics of each measured unit body are extracted, including the lithological combination characteristics, the structural characteristics, and the geophysical field characteristics. The lithological combination characteristics include the ore-bearing lithology type, the wall rock type, the thickness and lateral extension length of the ore-bearing layer, and the contact relationship between the ore-bearing lithology and the wall rock, which are directly extracted from the drilling core logging report and the rock and mineral identification data. The thickness and extension length are directly read from the report record values. The structural characteristics include the ore-controlling structure type, the fracture zone strike, the dip angle, the width of the fracture zone, the fold axis depth, and the occurrence position of the ore body in the structure, which are directly extracted from the field geological mapping measured data and the drilling structure logging. The occurrence position of the ore body is directly marked on the structure distribution map. The geophysical field characteristics specifically include the measured density of the ore body, the resistivity of the ore body, the physical property difference between the ore body and the wall rock, and the abnormal spatial range of the physical property, which are directly read from the well logging data and the core physical property test report. The abnormal range is directly delineated from the ground geophysical profile.

[0080] In step S503, the initial ore-forming probability is corrected by using the measured ore-forming characteristics to obtain an ore-forming probability.

[0081] In step S503, the initial ore-forming probability is corrected by using the measured ore-forming characteristics to obtain an ore-forming probability. In step S50301, a confidence level is set for the measured ore-forming characteristics.

[0082] In the embodiment, the known ore body in the target area and the surrounding area are first discretized into three-dimensional grid units consistent with the three-dimensional geological-geophysical model, and then the units are respectively assigned values according to their attributes. The ore-forming attribute of the original ore-forming unit can be regarded as a certain ore-forming (confidence level 100%), and the non-ore unit can be regarded as a certain non-ore (confidence level 0%). However, in order to prevent excessive absolute correction due to the existence of exploration errors in the measured data and incomplete ore-forming characteristics, the measured ore-forming characteristics of the ore-forming unit are set to a confidence level of 95% due to high exploration accuracy and traceable data, and the measured ore-forming characteristics of the non-ore unit are set to a confidence level of 5%. The confidence level of each grid unit is bound to the corresponding measured ore-forming or non-ore characteristics of the unit.

[0083] In step S50302, simulated ore-forming characteristics are extracted in the three-dimensional geological-geophysical model.

[0084] In this embodiment, simulated mineralization features are extracted from simulated units divided by a three-dimensional mesh, including lithological assemblage features, structural features, and geophysical field features. Lithological assemblage features are extracted by considering the lithological distribution information of the simulated units within the three-dimensional geological-geophysical model, including the spatial range of simulated ore-bearing lithology, the simulated thickness of the ore-bearing layer, and the simulated type of surrounding rock. Structural features are extracted by considering the structural information of the simulated units within the three-dimensional geological-geophysical model, including the strike, dip angle, simulated width of the fracture zone, and simulated burial depth of the fold axis of the simulated ore-controlling structures, while simultaneously locating the spatial correlation between the potential ore-bearing areas and structures within the simulated units. Geophysical field features are extracted by considering the simulated physical property values ​​of the simulated units within the three-dimensional geological-geophysical model, calculating the differences in physical properties between the simulated ore body and the surrounding rock, and extracting the corresponding spatial range of simulated physical property anomalies.

[0085] Step S50303: Calculate the similarity between the simulated mineralization features and the measured mineralization features.

[0086] The calculation of the similarity between the simulated mineralization features and the measured mineralization features specifically includes the following sub-steps: Step S5030301: Construct a mineralization-non-mineralization classification model based on the measured mineralization characteristics and the confidence level.

[0087] In this embodiment, a training dataset is constructed using measured unit cells. First, the measured mineralization features of each measured unit cell are used as input features, and the binary classification confidence of the measured unit cell is used as the output label. The binary classification confidence of the measured unit cell is determined based on the confidence of the mineralization features of the measured unit cell, i.e., 95% for mineralization and 5% for non-mineralization. To improve training, mineralization can be set to 1 and non-mineralization to 0 to form a more optimized binary classification label. Then, normalization is performed to generate training samples, and a mineralization-non-mineralization classification model is obtained by training the training samples based on random forest.

[0088] Step S5030302: Determine the sub-feature weights of the measured mineralization features based on the mineralization-non-mineralization classification model.

[0089] In this embodiment, based on the trained mineralized-non-mineralized classification model, and relying on its built-in feature contribution evaluation mechanism, the influence of each input sub-feature on the mineralized and non-mineralized binary classification results is automatically extracted. The ratio of the influence of each sub-feature to the sum of the influence of all sub-features is calculated to obtain the weight of each sub-feature of the measured mineralized features.

[0090] The sub-features mentioned here refer to further subdivisions of the measured mineralization characteristics. Specifically, the measured mineralization characteristics include major categories such as lithological assemblage characteristics, structural characteristics, and geophysical field characteristics. The specific sub-features under each major category are called sub-features. For example, sub-features of lithological assemblage characteristics may include ore-bearing lithology type, host rock type, and lithological contact relationship; sub-features of structural characteristics may include fault zone dip angle, fracture zone width, and fold morphology; and sub-features of geophysical field characteristics may include magnetic anomaly intensity, gravity anomaly gradient, and resistivity distribution.

[0091] Step S5030303: Calculate the sub-feature similarity based on the simulated mineralization features and the measured mineralization features.

[0092] In this embodiment, it is first necessary to clarify that the measured mineralization features are the set of measured sub-features of all verified mineralized and non-mineralized units in the target area, that is, each measured unit corresponds to a complete set of sub-features. The simulated mineralization features are the set of simulated sub-features of all units to be evaluated in the three-dimensional geological-geophysical model, that is, each simulated unit also corresponds to a complete set of sub-features, and the dimensions of the sub-features are completely consistent with the measured ones. For each simulated unit cell, the similarity of its sub-features in the same dimension must be calculated with each measured unit cell. For numerical sub-features, the simulated value of the sub-feature in the simulated unit cell and the measured value of the sub-feature in the currently compared measured unit cell are first normalized to the 0-1 range. Then, the absolute difference between the simulated and measured normalized values ​​is subtracted from 1 to obtain the similarity of each sub-feature in that dimension. For type sub-features (such as ore-bearing lithology type or ore-controlling structure type), if the simulated type of the sub-feature in the simulated unit cell is completely consistent with the measured type of the sub-feature in the currently compared measured unit cell, the similarity is 1; otherwise, it is 0. For example, if the measured normalized value of the ore-bearing layer thickness is 0.8 and the simulated normalized value is 0.7, the similarity is 1. The closer the values ​​are, the closer the similarity is to 1.

[0093] Step S5030304: The similarity is obtained by weighted summation of the sub-feature similarities based on the sub-feature weights.

[0094] In this embodiment, all simulated units are traversed to determine the most similar measured unit for each simulated unit, and their similarity is recorded. For a matching pair between a single simulated unit and a single measured unit, the weight of each sub-feature is multiplied by its corresponding sub-feature similarity set using the sub-feature weights and the set of sub-feature similarities of the matching pair. This yields a weighted contribution value for each sub-feature, and the weighted contribution values ​​of all sub-features are summed to obtain the overall similarity between the simulated unit and the current measured unit. Subsequently, keeping the simulated unit unchanged, the above weighted summation operation is repeated to calculate the overall similarity between the simulated unit and all measured units, forming a numerical set containing the overall similarity between the simulated unit and each measured unit.

[0095] Finally, the maximum value is selected from this set of values. This maximum value is the similarity value between the current simulated unit cell and all measured unit cells. At the same time, the attributes of the corresponding most similar measured unit cell are recorded.

[0096] The similarity satisfies the following formula: in, For simulation unit Similarity to the most similar measured unit cell This represents the total number of measured unit cells. The total number of sub-features Let be the weight of the t-th sub-feature. For simulation unit With measured unit In the Similarity in individual features.

[0097] Step S50304: Based on the similarity and the confidence level, the initial mineralization probability is corrected to obtain the mineralization probability.

[0098] The mineralization probability satisfies the following formula: in, For simulation unit The probability of mineralization, To be compatible with the simulation unit The similarity of the most similar measured unit cells To be compatible with the simulation unit Confidence level of the most similar measured unit cell For simulation unit The initial mineralization probability.

[0099] like Figure 2As shown, in another aspect, the present invention also provides a deep concealed ore body prediction system, including: a processor, an input device, an output device, and a memory, wherein the processor, the input device, the output device, and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the relevant steps of a relevant embodiment of the deep concealed ore body prediction method of the present invention.

[0100] This invention provides a deep, concealed ore body prediction system. The functional components can be integrated into a single processing unit, or each component can exist independently, or two or more components can be integrated into one unit. The integrated components can be implemented in hardware or software.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method of predicting a deep concealed ore body, characterized by, The method comprises: obtaining geological exploration data, geophysical data, lithology and physical property probability distribution data of a target area; constructing a three-dimensional geological model according to the geological exploration data; performing inversion calculation on the geophysical data to obtain a three-dimensional geophysical model by taking the three-dimensional geological model and the lithology and physical property probability distribution data as constraints; fusing the three-dimensional geological model and the three-dimensional geophysical model to obtain a three-dimensional geological-geophysical model; determining the ore-forming probability of a deep concealed ore body according to the three-dimensional geological-geophysical model.

2. The method for predicting a deep concealed ore body according to claim 1, characterized in that, The method comprises: extracting geological constraint conditions by using the three-dimensional geological model; calculating physical property probability constraint conditions according to the lithology and physical property probability distribution data; calculating a target function value based on the geological constraint conditions, the physical property probability constraint conditions and the geophysical data; performing iteration to obtain a three-dimensional geophysical model by taking minimizing the target function value as a target.

3. The method of claim 1, wherein, The method comprises: determining an initial ore-forming probability of a deep concealed ore body according to the three-dimensional geological-geophysical model; obtaining measured ore-forming characteristics of known ore bodies in a target area; correcting the initial ore-forming probability by using the measured ore-forming characteristics to obtain an ore-forming probability.

4. The method for predicting a deep concealed ore body according to claim 3, characterized in that, The method comprises: obtaining original geochemical exploration data and original remote sensing alteration data of a target area; performing preprocessing on the original geochemical exploration data and the original remote sensing alteration data to obtain geochemical anomaly data and remote sensing alteration data; screening ore-forming variables according to the three-dimensional geological-geophysical model, the geochemical anomaly data and the remote sensing alteration data; calculating an initial ore-forming probability of a deep concealed ore body by using the ore-forming variables and the three-dimensional geological-geophysical model.

5. The method for predicting a deep concealed ore body according to claim 4, characterized in that, The method comprises: performing denoising and enhancement on the original geochemical exploration data to obtain geochemical anomaly data; performing screening on the original remote sensing alteration data to obtain remote sensing alteration data.

6. The method for predicting a deep concealed ore body according to claim 5, characterized in that, The method comprises: extracting active state content and total state content in the original geochemical exploration data; calculating a ratio of the active state content to the total state content; based on the spatial distribution of the ratio, delineating the geochemical anomaly data.

7. The method of claim 5, wherein the step of identifying the deep concealed ore body is characterized by, The method comprises: extracting bedrock alteration mineral information in the original remote sensing alteration data by using spectral unmixing decomposition technology; extracting potential ore-forming target areas by using the three-dimensional geological-geophysical model; performing spatial superposition analysis on the bedrock alteration mineral information and the potential ore-forming target areas, and extracting the bedrock alteration mineral information in the potential ore-forming target areas to obtain remote sensing alteration data.

8. The method of claim 3, wherein, The utilization of the measured ore-forming characteristics to correct the initial ore-forming probability to obtain an ore-forming probability comprises: setting a confidence level for the measured ore-forming characteristics; extracting simulated ore-forming characteristics in the three-dimensional geology-geophysical model; calculating the similarity of the simulated ore-forming characteristics and the measured ore-forming characteristics; correcting the initial ore-forming probability based on the similarity and the confidence level to obtain an ore-forming probability.

9. The method of claim 8, wherein, The calculation of the similarity of the simulated ore-forming characteristics and the measured ore-forming characteristics comprises: constructing an ore-non-ore classification model based on the measured ore-forming characteristics and the confidence level; determining the sub-feature weight of the measured ore-forming characteristics according to the ore-non-ore classification model; calculating the sub-feature similarity according to the simulated ore-forming characteristics and the measured ore-forming characteristics; weighting and summing the sub-feature similarity based on the sub-feature weight to obtain the similarity.

10. A deep concealed ore body prediction system characterized by, comprises: a processor, an input device, an output device and a memory, which are connected to each other, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to invoke the program instructions to execute a deep concealed ore body prediction method according to any one of claims 1 to 9.

Images