Geological data-based mine hydrogeological intelligent exploration method, system, device and medium

Abstract

This application provides a method, system, equipment, and medium for intelligent hydrogeological exploration of mining areas based on geological data. The method includes: identifying the mineralization stages of acquired core samples from mineral deposit exploration to generate a fracture structure paleofluid channel archive; performing multi-index correlation evaluation on the fracture structure paleofluid channel archive and acquired fracture structure characteristics to generate fracture structure activation potential data; obtaining corresponding supplementary exploration datasets based on the fracture structure activation potential data; performing multi-source data collaborative inversion on the fracture structure paleofluid channel archive, supplementary exploration datasets, and acquired hydrogeological drilling data using a hierarchical Bayesian network to generate a three-dimensional hydrogeological structure model of the mining area; and dynamically superimposing the mining-induced stress field onto the three-dimensional hydrogeological structure model of the mining area to predict the activation probability of each fracture structure and the spatial location of water inrush channels at different mining stages, thereby generating a survey report on hidden disaster-causing factors. This method can improve the quality of exploration engineering in complex mining areas.

Description

Technical Field

[0001] This invention belongs to the field of mineral hydrogeological exploration technology, and in particular relates to an intelligent hydrogeological exploration method, system, equipment and medium for mining areas based on geological data. Background Technology

[0002] Accurate identification of the hydrogeological conditions of a mining area is fundamental to mineral resource evaluation, mining design, and water control and safety management. For a long time, the traditional exploration technology system, primarily based on surface hydrogeological mapping, geophysical exploration, hydrogeological drilling, pumping tests, water sampling, and groundwater dynamic monitoring, has provided basic hydrogeological data for mine construction. However, with the extension of mineral resource development to deeper levels and the large-scale development of complex mining areas (such as bedrock fracture deposits in arid regions and solid-liquid coexisting salt deposits), the limitations of traditional exploration methods are becoming increasingly apparent.

[0003] Traditional hydrogeological exploration relies primarily on manual data interpretation, profile analysis, parameter calculation, and results compilation. The determination of aquifer spatial structure, tectonic water-conducting characteristics, water-bearing zoning, and yield prediction largely depends on expert experience and simplified analytical models. This approach results in highly subjective and inconsistent exploration outcomes, and the simplified models struggle to characterize the inherent heterogeneity and anisotropy of aquifers. More importantly, while traditional methods have accumulated massive amounts of multi-source geological data (including surveying, geophysical exploration, drilling, testing, and monitoring), their utilization is often phased and isolated, making it difficult to meet the needs of high-quality exploration projects that require the identification of hidden hazardous factors potentially activated by mining disturbances. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, system, equipment and medium for intelligent hydrogeological exploration of mining areas based on geological data to address the above-mentioned technical problems. This method can improve the quality and efficiency of exploration engineering operations in complex mining areas and provide a reliable data foundation for the design of mining and water control.

[0005] Firstly, this application provides an intelligent hydrogeological exploration method for mining areas based on geological data, including:

[0006] The mineralization phases of the obtained mineral deposit exploration core samples were identified, and multi-phase fluid activity records of each fault structure were extracted to generate paleofluid channel archives of the fault structures.

[0007] A multi-index correlation assessment was conducted on the paleofluid channel archives of fracture structures and the acquired fracture structure characteristics, and the activation potential index of each fracture structure was calculated to generate fracture structure activation potential data.

[0008] Based on the activation potential data of fault structures, the mining area is divided into geological zones with different activation potential levels, and corresponding supplementary exploration datasets are obtained according to the activation potential level of each geological zone.

[0009] A three-dimensional hydrogeological structure model of the mining area is generated by multi-source collaborative inversion of the paleofluid channel archives of the fault structure, supplementary exploration datasets and acquired hydrogeological drilling data through hierarchical Bayesian network.

[0010] Dynamic superposition of mining-induced stress field on a three-dimensional hydrogeological structure model of the mining area is performed to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors.

[0011] In one embodiment, the mineralization phases of the obtained mineral deposit exploration core samples are identified, multi-phase fluid activity records of each fault structure are extracted, and paleofluid channel archives of the fault structures are generated, including:

[0012] Fluid inclusions were prepared based on the obtained core samples from mineral deposit exploration, and fluid inclusion combinations of different phases were identified to generate inclusion phase classification data.

[0013] Based on the inclusion phase classification data, microthermography is performed on the inclusions of the corresponding phases to obtain the homogenization temperature and freezing point temperature of the inclusions of each phase. The salinity of the inclusions is calculated using the empirical formula of freezing point temperature and salinity, generating a set of thermodynamic parameters of the inclusions that includes homogenization temperature, salinity and phase classification.

[0014] Based on the thermodynamic parameter set of inclusions, laser Raman spectroscopy analysis was performed on representative inclusions to obtain gas phase composition information of inclusions and generate a gas phase composition dataset of inclusions.

[0015] Spatial positioning was performed on the inclusion phase classification data, inclusion thermodynamic parameter set, and inclusion gas phase composition dataset. The thermodynamic parameters of inclusions in each phase were correlated with the associated fracture structures to generate a fracture structure paleofluid channel archive containing the intensity, phase, and filling mineralization characteristics of ore-forming fluid activity in each fracture structure.

[0016] In one embodiment, a multi-index correlation assessment is performed on the paleofluid channel archives of fracture structures and the acquired fracture structure features, and the activation potential index of each fracture structure is calculated to generate fracture structure activation potential data, including:

[0017] Based on the paleofluid channel archives of the fault structures, the average homogenization temperature, average salinity, and gas phase composition of the inclusions during the mineralization period of each fault structure were extracted, and the paleofluid activity intensity index of each fault structure was calculated to generate a paleofluid activity intensity dataset.

[0018] Based on the paleofluid channel archives of fracture structures, the density of inclusions, the width of vein filling and the degree of mineralization cementation after the mineralization period of each fracture structure were extracted, and the mineralization filling degree index of each fracture structure was calculated to generate a mineralization filling degree dataset.

[0019] Mechanical parameters are extracted from the obtained geometric and kinematic features of the fracture structure and modern geostress field data. The attitude, displacement, friction coefficient and the angle between the direction of the maximum principal stress and the fracture strike of each fracture structure are obtained, and a set of mechanical parameters of the fracture structure is generated.

[0020] Four-dimensional index fusion was performed on the paleofluid activity intensity dataset, mineralization filling degree dataset, and fracture structure mechanical parameter set. The activation potential index of each fracture structure was calculated through the activation potential assessment model to generate fracture structure activation potential data.

[0021] In one embodiment, a three-dimensional hydrogeological structure model of the mining area is generated by multi-source data collaborative inversion using a hierarchical Bayesian network on the paleofluid channel archives of the fault structure, supplementary exploration datasets, and acquired hydrogeological drilling data, including:

[0022] Coordinate transformation was performed on the paleofluid channel archives of fracture structures, supplementary exploration datasets, and acquired hydrogeological drilling data. The discrete inclusion data in the paleofluid channel archives of fracture structures were converted into a paleofluid attribute field with spatial continuity through spatial interpolation, resulting in a multi-source data field with unified coordinates.

[0023] Using paleofluid property fields as prior constraints, geophysical data in the supplementary exploration dataset as the main inversion data, and hydrogeological drilling data as calibration and verification data, a hierarchical Bayesian network is used to construct the probabilistic dependencies between different data sources and generate a collaborative inversion parameter set.

[0024] A three-dimensional geological structure model of the mining area is generated by iterative inversion of the parameter set and the model response is matched with the observation data until the fitting error is less than the preset accuracy threshold.

[0025] In one embodiment, a dynamic superposition of the mining-induced stress field is performed on a three-dimensional hydrogeological structure model of the mining area to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, in order to generate a survey report on hidden disaster-causing factors, including:

[0026] Based on the rock mechanics parameter library and mining scheme of the mining area, the mining-induced stress field numerical simulation is carried out on the three-dimensional hydrogeological structure model of the mining area. The shear stress increment and normal stress change at the location of the fracture structure in each mining stage are calculated, and the mining-induced stress field time series dataset is generated.

[0027] Coupled analysis was performed on the time series dataset of mining-induced stress field and the data on the activation potential of fracture structures. The shear stress increment of each mining stage was compared with the shear strength of the fracture structure. The activation probability of each fracture structure at different mining stages was calculated using the activation probability prediction formula, and time series data of fracture activation probability was generated.

[0028] Hydraulic connectivity analysis was performed on the time series data of fracture activation probability. For fracture structures whose activation probability exceeded the preset warning threshold, the hydraulic gradient of surface water, seepage path length and equivalent permeability coefficient were calculated based on the three-dimensional hydrogeological structure model of the mining area to obtain the location data of water inrush channels.

[0029] Correlation of prevention and control measures is carried out with time series data on fault activation probability, data on water inrush channel location, and risk level classification standards. The location of monitoring points is determined based on the activation probability of each fault structure and the development degree of water inrush channels, and a survey report on hidden disaster-causing factors is generated.

[0030] Secondly, this application also provides a data-based intelligent hydrogeological exploration system for mining areas, used to implement the data-based intelligent hydrogeological exploration method for mining areas as described above. The system includes:

[0031] The paleofluid archive construction module is used to identify the mineralization phases of the obtained mineral deposit exploration core samples, extract multi-phase fluid activity records of each fault structure, and generate paleofluid channel archives of the fault structures.

[0032] The activation potential assessment module is used to perform multi-index correlation assessment between the paleofluid channel archives of fracture structures and the acquired fracture structure features, and to calculate the activation potential index of each fracture structure to generate fracture structure activation potential data.

[0033] The zoning exploration module is used to divide the mining area into geological zones with different activation potential levels based on fault structure activation potential data, and to obtain corresponding supplementary exploration datasets according to the activation potential level of each geological zone.

[0034] The multi-source data modeling module is used to perform multi-source data collaborative inversion on the fracture structure paleofluid channel archive, supplementary exploration dataset and acquired hydrogeological drilling data through a hierarchical Bayesian network, and generate a three-dimensional hydrogeological structure model of the mining area.

[0035] The exploration prediction report module is used to dynamically overlay the mining-induced stress field on the three-dimensional hydrogeological structure model of the mining area, predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors.

[0036] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the intelligent hydrogeological exploration method for mining areas based on geological data as described above.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent hydrogeological exploration method for mining areas based on geological data as described above.

[0038] The aforementioned intelligent hydrogeological exploration method, system, computer equipment, and storage media for mining areas based on geological data identify the mineralization stages of core samples from ore deposit exploration and extract multi-stage fluid activity records from fault structures to generate paleofluid channel archives. These archives are then correlated with fault structure characteristics using multiple indicators to assess their activation potential and calculate an activation potential index, forming fault structure activation potential data. Based on this activation potential data, different levels of geological zones are defined, and corresponding supplementary exploration datasets are obtained. A hierarchical Bayesian network is then used to analyze the paleofluid channel archives, supplementary exploration datasets, and hydrogeological drilling data. Multi-source data collaborative inversion is used to construct a three-dimensional hydrogeological structure model of the mining area. Then, the model is dynamically superimposed with the stress field of mining to predict the activation probability of fracture structures and the spatial location of water inrush channels at different mining stages, and to form a survey report of hidden disaster-causing factors. This method can effectively overcome the shortcomings of traditional exploration methods, such as reliance on manual experience, isolated and scattered data processing, difficulty in accurately depicting the heterogeneity of aquifers and identifying hidden disaster-causing factors under mining disturbance. It significantly improves the objectivity, consistency and accuracy of hydrogeological exploration in complex mining areas, improves the quality and efficiency of exploration operations, and provides reliable data support for the design of mineral deposit mining and water control. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A flowchart illustrating an intelligent hydrogeological exploration method for mining areas based on geological data, provided as an embodiment of this application. Figure 1 ;

[0041] Figure 2 A flowchart illustrating an intelligent hydrogeological exploration method for mining areas based on geological data, provided as an embodiment of this application. Figure 2 ;

[0042] Figure 3 This application provides a schematic diagram of the structure of an intelligent hydrogeological exploration system for mining areas based on geological data, according to one embodiment. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0044] First, a brief introduction to the terms used in the embodiments of this application will be given.

[0045] Metallogenic phase identification refers to the technical process of dividing and determining the different metallogenic stages and sequences of metallogenic events experienced during the formation of a mineral deposit through comprehensive analysis of rock, ore, structure, and mineralization characteristics during geological exploration. Its purpose is to clarify the temporal relationship of mineralization, the stages of tectonic evolution, and the laws governing fluid activity, providing fundamental geological evidence for analyzing tectonic water-conducting characteristics, aquifer formation mechanisms, and water-bearing conditions of the deposit.

[0046] Multi-phase fluid activity records refer to a collection of geological evidence preserved in rock fissures, mineral alteration, inclusions, and tectonic surfaces, reflecting the multi-stage fluid migration, filling, replacement, and alteration processes throughout geological history. This record reflects the properties, sources, pathways, and intensity of fluid activity at different times, serving as an important indicator for determining tectonic conductivity, aquifer evolution, and paleohydrogeological conditions.

[0047] Hierarchical Bayesian networks are hierarchical intelligent analysis models based on probabilistic reasoning. They can learn and reason about the relationships between variables under conditions of multi-source, heterogeneous, and uncertain information. They are suitable for the fusion, inversion, and prediction of multi-source geological data and can effectively improve the stability and reliability of interpretation results under complex geological conditions.

[0048] Based on the above definitions, the implementation environment of the intelligent hydrogeological exploration method for mining areas based on geological data provided in this application embodiment is described. Indicatively, this implementation environment includes: multimodal sensors, a terminal, and a processor. The processor, multimodal sensors, and terminal are connected via network signals. The multimodal sensors include, but are not limited to, core spectral sensors, structural fracture sensors, fluid geochemical sensors, groundwater level sensors, water pressure sensors, geostress sensors, borehole imaging sensors, and / or hydrogeological parameter sensors. The processor can be a central processing unit, a multi-core processor, or an artificial intelligence chip, etc., and is not limited here.

[0049] Based on the above definitions and implementation environment, the application scenarios of the embodiments of this application are described. The intelligent hydrogeological exploration method for mining areas based on geological data provided in the embodiments of this application can be applied to scenarios including but not limited to the following:

[0050] In complex mining areas with intense tectonic activity and well-developed fault systems, traditional hydrogeological exploration methods struggle to accurately identify the activation risk of ancient fluid channels that were mineralized and sealed after the mineralization period, under the influence of mining disturbances. Taking the Karchar fluorite mine in Ruoqiang County, Xinjiang, as an example, two main faults pass through the north and south sides of the ore body, respectively. Fluid inclusion studies from the mineralization period indicate significant traces of ancient fluid activity at depth. However, after the mineralization period, the fault zones were filled and cemented by minerals such as calcite and quartz. Conventional geophysical methods struggle to distinguish between dense, intact rock masses and sealed ancient channels, leading to disputes regarding the classification of hydrogeological conditions in the mining area. This technical solution systematically extracts ancient fluid channel archives from fault structures, quantitatively assesses the activation potential index of each fault structure, and classifies risk levels accordingly for differentiated, detailed exploration. This approach effectively identifies these "dormant-activated" hidden disaster-causing factors. The dynamic risk prediction report output by the solution can directly guide mines in formulating phased water prevention and control measures and monitoring point layout plans, meeting the technical requirements of "comprehensive and systematic investigation and scientific and accurate measurement" in the "Specification for Survey of Hidden Disaster-Causing Factors in Mines", and improving the accuracy and effectiveness of mine major risk prevention and control.

[0051] As mineral resource development extends to deeper areas, the hydrogeological conditions in mining areas become increasingly complex. Traditional methods for predicting water inflow based on steady-flow analysis and empirical formulas are insufficient to accurately characterize the heterogeneity and anisotropy of deep aquifer systems, often resulting in significant discrepancies between predicted and measured values. During deep ore body mining, as the goaf expands and the mining-induced stress field evolves, previously closed fault structures may gradually become activated, transforming into new water-conducting channels. This leads to a nonlinear increase in water inflow, posing a serious threat to mine drainage capacity and safe production. This technical solution constructs a multi-scale hierarchical Bayesian network to collaboratively invert the paleofluid activity history reflected in microscopic inclusion data, the spatial structure of faults revealed by macroscopic geophysical data, and local hydrogeological parameters obtained from drilling data, establishing a high-precision three-dimensional hydrogeological structure model. Based on this, numerical simulation of mining-induced stress field is superimposed to predict the activation probability of each fault structure and the evolution trend of water inrush channels at different mining stages. This achieves a technological leap from static geological models to dynamic risk prediction throughout the entire life cycle, providing a scientific basis for developing phased drainage system optimization plans and emergency prevention and control plans for deep mines.

[0052] Large mining areas often accumulate decades of geological exploration results, encompassing massive amounts of multi-source data from regional geological surveys, mineral exploration, hydrogeological exploration, geophysical surveys, drilling projects, pumping tests, and long-term dynamic monitoring. This data spans decades in timescale and ranges from microscopic inclusions to macroscopic structural units in spatial scale. Physically, it involves various parameters such as temperature, pressure, resistivity, wave velocity, and permeability. Traditional exploration methods often process this data in stages and in isolation, failing to create a data-driven, dynamically optimized exploration loop. This technical solution establishes a unified data fusion framework, using fault structure paleofluid channel archives as prior constraints, differentiated supplementary exploration data as the main inversion data, and historical hydrogeological drilling data as calibration and verification data. Through a hierarchical Bayesian network, it achieves collaborative inversion of data across scales and physical dimensions. This solution can fully leverage the value of existing geological data in mining areas, improve the accuracy and reliability of three-dimensional hydrogeological models while supplementing a small amount of targeted exploration work, and achieve an optimal balance between exploration input and result quality. It is applicable to typical scenarios such as continuous exploration of large-scale mining resources, re-evaluation of hydrogeological conditions during the production period, and exploration of deep and peripheral areas of mining areas.

[0053] This is merely an illustrative example; the power construction progress collaborative management method provided in this application embodiment can also be applied to other application scenarios. It is only an example and does not limit the specific application scenarios.

[0054] In one exemplary embodiment, such as Figure 1 As shown, a method for collaborative management and control of power construction progress is provided. This embodiment illustrates the application of this method to a terminal in the aforementioned implementation environment. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and is implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps 101 to 105:

[0055] Step 101: Identify the mineralization phases of the obtained mineral deposit exploration core samples, extract multi-phase fluid activity records of each fault structure, and generate paleofluid channel archives of the fault structures.

[0056] Specifically, the core samples from mineral exploration are derived from hydrogeological and mineral exploration boreholes already conducted within the mining area. Through core profile observation, mineral alteration sequence identification, fluid inclusion thermometry, and geochemical element analysis, the temporal relationship between different tectonic movement stages and mineralization stages is determined. Simultaneously, traces of fluid activity, fracture filling characteristics, mineral precipitation sequences, and water-conducting space development information from different periods are systematically extracted along the fault tectonic zone. The above information is structured and integrated according to spatial coordinates, activity periods, and fluid properties to form a paleofluid channel archive of the fault structure containing spatiotemporal evolution characteristics and water-conducting channel records. This archive can objectively reconstruct the historical water-conducting patterns of the fault structure, providing a real and continuous geological basis for subsequent tectonic activation evaluation.

[0057] Step 102: Perform multi-index correlation evaluation on the paleofluid channel archives of fracture structures and the acquired fracture structure characteristics, and calculate the activation potential index of each fracture structure to generate fracture structure activation potential data.

[0058] For example, this method can select fracture scale, occurrence, filling degree, lithological contact relationship, paleofluid activity intensity, and fracture connectivity as correlation indicators. It can construct a fracture structure activation potential evaluation system by combining weighted assignment, membership function and comprehensive index calculation. The activation potential index of each fracture structure is obtained through index quantification, weight allocation and coupling calculation, forming fracture structure activation potential data with spatial location and numerical attributes, so as to realize the quantitative characterization of the possibility of fractures reopening and conducting water under mining conditions.

[0059] Step 103: Based on the fault structure activation potential data, the mining area is divided into geological zones with different activation potential levels, and corresponding supplementary exploration datasets are obtained according to the activation potential level of each geological zone.

[0060] Specifically, the mining area is divided into high-activation-potential, medium-activation-potential, and low-activation-potential zones according to the activation potential index range. Based on the differences in the zone levels, corresponding exploration methods and data acquisition schemes are deployed to obtain supplementary exploration datasets that match the risk level of each zone. These datasets include targeted geophysical data, hydrogeological observation data, borehole logging data, and water level monitoring data, making the supplementary exploration work more targeted and effectively improving data acquisition efficiency and exploration accuracy.

[0061] Step 104: A three-dimensional hydrogeological structure model of the mining area is generated by multi-source collaborative inversion of the fracture structure paleofluid channel archive, supplementary exploration dataset and acquired hydrogeological drilling data through a hierarchical Bayesian network.

[0062] Specifically, the hydrogeological drilling data includes borehole depth, stratigraphic sequence, aquifer location, water level depth, and hydrogeological parameters. First, the multi-source heterogeneous data undergoes spatial coordinate unification, format normalization, and uncertainty quantification. Using a hierarchical Bayesian network as the reasoning framework, the paleofluid channel archive is used as a priori geological constraint, and supplementary exploration data and drilling data are used as observation variables. Through probability transfer, parameter iteration, and model optimization, the spatial distribution of aquifers, fracture water conductivity, spatial distribution of hydrogeological parameters, and structural interface location are determined by inversion. This yields a three-dimensional hydrogeological structure model that can truly reflect the heterogeneous characteristics of the mining area. This model has the characteristics of high accuracy, strong constraints, and low uncertainty.

[0063] Step 105: Dynamically superimpose the mining-induced stress field on the three-dimensional hydrogeological structure model of the mining area to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors.

[0064] Specifically, the mining-induced stress field is obtained through numerical simulation based on the mining boundary, mining sequence, and advance speed of the mining area. The stress distribution results of different mining stages are spatially coupled with the three-dimensional hydrogeological structure model to analyze the control effect of stress loading and unloading on the stability of the fracture structure. The activation probability of each fracture at different mining stages is calculated using a probabilistic statistical model. Combined with spatial superposition analysis, the specific location and development range of potential water inrush channels are determined. The results of fracture activation evaluation, water inrush channel prediction, and comprehensive analysis of hydrogeological conditions are systematically integrated to form a standardized survey report of hidden disaster-causing factors, providing direct support for the deployment of mine water control projects and safety production decisions.

[0065] The aforementioned intelligent hydrogeological exploration method for mining areas based on geological data identifies the mineralization stages of core samples from ore deposit exploration and extracts multi-stage fluid activity records from fault structures to generate paleofluid channel archives. This archive is then correlated with fault structure characteristics using multiple indicators to assess its activation potential and calculate an activation potential index, forming fault structure activation potential data. Based on this activation potential data, different levels of geological zones are defined, and corresponding supplementary exploration datasets are obtained. Finally, a hierarchical Bayesian network is used to perform multi-source data collaboration on the paleofluid channel archives, supplementary exploration datasets, and hydrogeological drilling data. Inversion is used to construct a three-dimensional hydrogeological structure model of the mining area. Then, the model is dynamically superimposed with the stress field of mining to predict the activation probability of fracture structures and the spatial location of water inrush channels at different mining stages, and to form a survey report of hidden disaster-causing factors. This method can effectively overcome the shortcomings of traditional exploration methods, such as reliance on manual experience, isolated and scattered data processing, difficulty in accurately depicting the heterogeneity of aquifers and identifying hidden disaster-causing factors under mining disturbance. It significantly improves the objectivity, consistency and accuracy of hydrogeological exploration in complex mining areas, improves the quality and efficiency of exploration operations, and provides reliable data support for the design of mineral deposit mining and water control.

[0066] In one embodiment, such as Figure 2 As shown, step 101 involves identifying the mineralization phases of the obtained mineral deposit exploration core samples, extracting multi-phase fluid activity records from each fault structure, and generating paleofluid channel archives for the fault structures. This can include:

[0067] Step 201: Based on the obtained mineral deposit exploration core samples, fluid inclusions are prepared, fluid inclusion combinations of different phases are identified, and inclusion phase classification data is generated.

[0068] Fluid inclusions are the original mineral-forming fluid media captured and retained in lattice defects or fractures during mineral crystallization. They are direct carriers reflecting ancient fluid activity. This technical solution selects representative samples from fracture walls, between mineral grains, and in tectonic infills. Double-sided polished inclusion slides are prepared through standard procedures such as slicing, polishing, and cleaning. This provides compliant analytical samples for subsequent observation and testing, ensuring the complete preservation of inclusion information.

[0069] For example, this method can select core samples from existing exploration borehole core libraries in the mining area near the target fault structure, prioritizing core segments from fault zones with visible quartz veins, calcite veins, or alteration halos. The selected core samples are cut into thin blanks with a thickness of 5 to 8 mm, and then subjected to coarse grinding, fine grinding, and polishing processes to produce fluid inclusion sections with a thickness of 0.1 to 0.3 mm. The sections must be parallel on both sides and free of scratches. These sections are then observed under a polarizing microscope in both transmitted and reflected light illumination modes. Based on the phase composition of the inclusions, they are classified into types such as two-phase gas-liquid inclusions, three-phase inclusions containing CO2, pure gas-phase inclusions, and pure liquid-phase inclusions. Then, primary and secondary inclusions are distinguished based on their morphological characteristics, distribution characteristics, and tangential relationship with the host mineral crystal faces. Primary inclusions are distributed along the growth zonation of the host mineral or are randomly distributed in an isolated manner, reflecting information about fluid activity during the mineralization period. Secondary inclusions are linearly arranged along microfractures inside the mineral, reflecting the superposition and modification of tectonic and thermal events after the mineralization period. Furthermore, this method can identify different mineralization periods based on the grouping characteristics of homogenization temperature of inclusions: after homogenization temperature measurement of each inclusion, a homogenization temperature frequency distribution curve is plotted using the kernel density estimation method. If the curve shows multiple separate peak intervals, each peak interval corresponds to an independent fluid activity period. Based on this, inclusions are divided into multiple period combinations such as pre-mineralization period, main mineralization period, and post-mineralization period, generating inclusion period classification data. This data records the type, morphology, distribution characteristics, and spatial relationship with the host mineral of inclusions in each period.

[0070] Step 202: Based on the inclusion phase division data, perform microthermography on the inclusions of the corresponding phases to obtain the homogenization temperature and freezing point temperature of the inclusions of each phase, and calculate the salinity of the inclusions using the empirical formula of freezing point temperature and salinity, thereby generating a set of thermodynamic parameters of the inclusions that includes homogenization temperature, salinity and phase division.

[0071] Specifically, this method involves placing the prepared inclusion sheet in the sample chamber of a hot-cold stage for microthermometry. During the freezing phase, the sample is cooled to below -180°C at a rate of 20°C to 30°C per minute to ensure complete freezing of the fluid within the inclusion. Then, the temperature is slowly increased at a rate of 0.5°C to 1°C per minute, continuously observing phase changes within the inclusion. The temperature is recorded as the freezing point when the last ice crystal disappears and the gas-liquid phases are completely homogenized into a liquid phase. During the heating phase, the temperature is increased at a rate of 5°C to 10°C per minute, then reduced to a rate of 0.5°C to 1°C per minute as it approaches the homogenization temperature. The temperature is recorded as the homogenization temperature when the gas-liquid phases within the inclusion are completely homogenized into a single phase. For the same inclusion assembly, this method measures the homogenization temperature and freezing point temperature of 10 to 20 inclusions, and the arithmetic mean is taken as the representative thermodynamic parameter for that batch of inclusions. After obtaining the freezing point temperature, this method calculates the salinity of the inclusions using an empirical formula relating freezing point temperature to salinity. For the NaCl-H2O system, the salinity is calculated using the following formula:

[0072]

[0073] in, Salinity of inclusions, expressed as a mass percentage of NaCl. The absolute value of the freezing point temperature is expressed in degrees Celsius. The coefficient 1.78 is the linear term coefficient, indicating that salinity increases by approximately 1.78 percentage points for every 1 degree Celsius decrease in freezing point temperature, reflecting the dominant linear relationship between salinity and freezing point temperature. The coefficients 0.0442 and 0.000557 are the quadratic and cubic term coefficients, respectively, used to correct for nonlinear deviations between freezing point depression and salinity under high salinity conditions. By introducing higher-order terms, this formula can provide more accurate calculation results when the freezing point temperature is low and the corresponding salinity is high. Its applicable salinity range is typically 0 to 26 wt% NaCl equivalent. This formula is based on experimental data of freezing point depression in the NaCl-H₂O system and has good applicability within the freezing point temperature range of 0°C to -30°C. For inclusions containing multiple components of gas, such as CO2, this method uses the equation of state proposed by Brown and Lamb to calculate salinity. The salinity value is obtained by looking up the relationship curve between the melting temperature of CO2 cages and salinity. The calculated salinity is associated with and stored with the measured homogenization temperature and inclusion phase classification information to form a set of thermodynamic parameters of inclusions. This set of parameters reflects the temperature conditions and salinity characteristics of the ore-forming fluids in each phase, and is used to infer the evolution trajectory of the ore-forming fluids and the differences in the sources of fluids in different phases.

[0074] Step 203: Based on the thermodynamic parameter set of the inclusions, perform laser Raman spectroscopy analysis on representative inclusions to obtain gas phase composition information of the inclusions and generate a gas phase composition dataset of the inclusions.

[0075] Specifically, this method selects representative inclusions with uniform temperature distributions across different ranges from inclusions that have undergone microthermometric analysis as Raman spectroscopy analysis targets, ensuring that the analyzed inclusions cover all stages of mineralization. Thin sections of the inclusions are placed on the sample stage of a laser Raman spectrometer, using a laser source with wavelengths of 532 nm or 785 nm. The laser beam is focused onto the inclusion through a microscope objective, with a focused spot diameter of 1 to 2 micrometers, ensuring that the laser energy precisely targets the inclusion without damaging surrounding minerals. During data acquisition, this method operates within a wavenumber range of 100 to 4000 cm⁻¹. -1 The Raman scattering signal of the inclusions was collected internally, with an acquisition time set to 30 to 60 seconds. Multiple accumulations were used to improve the signal-to-noise ratio. The principle of Raman spectroscopy analysis is that different molecules have characteristic vibrational energy levels. When an incident laser undergoes inelastic scattering with a molecule, the frequency of the scattered light shifts accordingly to the molecular vibrational frequency, manifesting as characteristic peaks in the spectrum. The molecular type can be identified by the peak position. For example, at a wavenumber of 1285 cm⁻¹... -1 and 1388cm -1 The characteristic peak appearing at 2917 cm⁻¹ indicates the presence of CO₂, which is due to the symmetric stretching vibration and Fermi resonance of the CO₂ molecule; -1The characteristic peak appearing at indicates the presence of CH4, corresponding to the stretching vibration of the CH bond; at wavenumber 2328 cm⁻¹ -1 The characteristic peak appearing at this point indicates the presence of N2, corresponding to the stretching vibration of the N≡N bond. This method calculates the relative CO2 content using the following formula:

[0076]

[0077] in, This represents the relative content of CO2. and They are 1285cm respectively -1 and 1388cm -1 The peak area of ​​the characteristic peak is obtained by integrating the peak area after baseline correction. The total Raman signal is in the range of 100 to 4000 cm⁻¹ -1 The peak area integral value within the range represents the total Raman scattering intensity of the gaseous components of the inclusions. This formula, by using the ratio of the CO2 characteristic peak area to the total signal intensity as a relative content indicator, can eliminate the influence of instrument factors such as laser power and focusing state on quantitative analysis. Correlating the gaseous component data obtained from Raman spectroscopy analysis with the corresponding thermodynamic parameters of the inclusions generates a gaseous component dataset. This dataset is used to determine the redox environment and source characteristics of the ore-forming fluids. For example, a high CO2 content indicates that the fluid may originate from magma degassing or metamorphic decomposition of carbonate rocks, while the presence of CH4 indicates a reducing environment.

[0078] Step 204: Spatial positioning of inclusion phase classification data, inclusion thermodynamic parameter set, and inclusion gas phase composition dataset; correlation of thermodynamic parameters of inclusions in each phase with the associated fracture structures; generation of fracture structure paleofluid channel archives containing the intensity, phase, and filling mineralization characteristics of ore-forming fluid activity in each fracture structure.

[0079] Specifically, this method can establish a unified borehole database and fault structure spatial database for the mining area. Each borehole records the borehole coordinates, borehole depth trajectory, and core logging information. Each fault structure records its surface outcrop location, attitude, extension length, and intersection relationship with the borehole. For example, this method converts the sampling depth corresponding to each inclusion into three-dimensional spatial coordinates using the following formula:

[0080]

[0081]

[0082]

[0083] in, These are the three-dimensional coordinates of the borehole opening, in meters, obtained through measurements using the mine area control network. The depth of the pore at the location of the inclusion. This represents the borehole inclination angle, expressed in degrees. A positive value is used when drilling downwards. The borehole azimuth is given in degrees. In the formula... The projected length of the borehole trajectory on the horizontal plane, multiplied by and The displacement increments in the X and Y directions are obtained respectively; Let Z be the projected length of the borehole trajectory in the vertical direction. Since the Z-coordinate decreases as the borehole drills downwards, we use... Subtract this value. Further, this method determines the fracture structure associated with each inclusion through spatial analysis: calculating the shortest distance between the spatial coordinates of the inclusion and the spatial surface of each fracture structure; when this distance is less than a preset association threshold, the inclusion is assigned to the inclusion set of the corresponding fracture structure. For inclusions of different phases within the same fracture structure, this method calculates the development density of inclusions of each phase using the following formula:

[0084]

[0085] in, This refers to the density of inclusion development, expressed in units per cubic centimeter. The total number of inclusions in this phase within the fault structure is obtained by statistically analyzing the inclusions within the core segment controlled by the fault structure. The core volume controlled by the fault structure is calculated by multiplying the width of the fracture zone by the length of the sampled core. This method statistically analyzes the average homogenization temperature, average salinity, and gas phase composition of inclusions from different periods, and assesses the degree of mineralization infilling of the fault structure. Based on the development density of inclusions after the mineralization period, the vein filling width, and the degree of mineralization cementation, the mineralization infilling degree index is calculated using the following formula:

[0086]

[0087] in, It is a mineralization filling degree index, with a value ranging from 0 to 1. The larger the value, the higher the degree to which the fracture structure is filled and sealed by mineralized material. This represents the density of inclusions after the mineralization period, expressed in units of inclusions per cubic centimeter, reflecting the intensity of fluid activity after the mineralization period. This represents the maximum density of inclusions developed after the mineralization period in the mining area. It was obtained by statistically analyzing the density of inclusions developed after the mineralization period across all fault structures in the mining area and taking the maximum value. This value is used for... Perform normalization processing; The width of the vein filling is measured in meters and reflects the scale of the fracture zone being filled by the vein. It is obtained by statistical analysis of the cumulative thickness of the vein in the core logging. The maximum width of the vein filling in the mining area is used for... Perform normalization processing; The degree of mineralization cementation is represented by a value ranging from 0 to 1, where 1 represents complete cementation and no permeability, and 0 represents no cementation and good permeability. The degree of fracture development and cement density within the vein is assessed through core observation. This method generates a comprehensive record for each fault structure, including the intensity of fluid activity during the mineralization period, the number of activity periods, the characteristics of mineralization filling after the mineralization period, and the spatial distribution of inclusions. The collection of records for all fault structures constitutes a paleofluid channel archive for the fault structure. This archive links and binds microscopic inclusion information with the spatial location of the fault structure at the macroscopic scale.

[0088] This embodiment organically combines petrographic observation of inclusions, microthermography, laser Raman spectroscopy analysis, and spatial relocation technology to extract thermodynamic parameters and gas phase composition information of inclusions in each phase. By transforming spatial coordinates, the microscopic inclusion data is spatially correlated with specific fault structures. Quantitative indicators such as inclusion development density and mineralization infill degree index are used to characterize the fluid activity history and self-sealing degree of each fault structure. This allows the fluid activity trajectory of each fault structure in geological history to be clearly presented, providing reliable basic data support for subsequent activation potential assessment.

[0089] In one embodiment, a multi-index correlation assessment is performed on the paleofluid channel archives of fracture structures and the acquired fracture structure features, and the activation potential index of each fracture structure is calculated to generate fracture structure activation potential data, including:

[0090] Based on the paleofluid channel archives of the fault structures, the average homogenization temperature, average salinity, and gas phase composition of the inclusions during the mineralization period of each fault structure were extracted, and the paleofluid activity intensity index of each fault structure was calculated to generate a paleofluid activity intensity dataset.

[0091] Specifically, this method retrieves mineralization-related inclusion data for each fault structure from the paleofluid channel archive, including a set of homogenization temperature measurements, a set of salinity measurements, and the relative CO2 content in the gas phase composition of that inclusion. For each fault structure, the method calculates the average homogenization temperature of the mineralization-related inclusions. and average salinity The relative CO2 content of mineralization inclusions in the fracture structure was extracted from the gas phase composition dataset. To eliminate the dimensional influence caused by differences in mineralization conditions between different mining areas, this method introduces the maximum homogenization temperature of inclusions during the mineralization period of the mining area. maximum salinity and the maximum CO2 content in the gas phase of the inclusions As a normalization benchmark, the above maximum value is obtained by statistically analyzing the mineralization period inclusion parameters of all fault structures in the mining area. For example, this method can calculate the paleofluid activity intensity index of each fault structure using the following formula:

[0092]

[0093] in, The paleofluid activity intensity index ranges from 0 to 1, with higher values ​​indicating greater activity of the fault structure as a fluid conduit during the mineralization period. This index integrates three dimensions: homogenization temperature, salinity, and CO2 content. Homogenization temperature reflects the thermal conditions of the ore-forming fluid; higher temperatures typically indicate the fluid originates from deeper layers or a stronger geothermal gradient. Salinity reflects the concentration characteristics of the ore-forming fluid; high salinity is often associated with magmatic hydrothermal fluids or deep-circulating brine. CO2 content reflects the volatile composition of the ore-forming fluid; high CO2 content usually indicates the fluid originates from magmatic degassing or mantle sources. Multiplying these three normalized indices provides a comprehensive characterization of the fault structure's fluid conduit potential during the mineralization period. This method calculates the corresponding paleofluid activity intensity index for each fault structure, and the calculation results for all fault structures constitute a paleofluid activity intensity dataset.

[0094] Based on the paleofluid channel archives of the fault structures, the density of inclusions, the width of vein filling and the degree of mineralization cementation after the mineralization period of each fault structure were extracted, and the mineralization filling degree index of each fault structure was calculated to generate a mineralization filling degree dataset.

[0095] Specifically, this method can retrieve post-mineralization inclusion data associated with each fault structure from the paleofluid channel archives of the fault structure. This includes the number of inclusions developed per unit volume of core for that period, the measured filling width of veins within the fault zone, and the degree of mineralization cementation determined through core observation and thin section identification. Among these, the inclusion development density... The total number of inclusions from the post-mineralization period within the core volume controlled by this fault structure was statistically analyzed. With core volume The ratio was calculated to obtain the pulse filling width. The degree of mineralization cementation was obtained by taking the average value after multi-point measurements of quartz veins, calcite veins, and other veins developed within the fault zone. For comprehensive evaluation, the value ranges from 0 to 1. 1 represents that the veins within the fault zone are completely cemented, pores and fractures are filled with material, and there is no permeability. 0 represents that there is no cementation within the fault zone or the cement has dissolved, and permeability is good. This indicator is determined by geologists during core logging based on the vein filling rate, cement type, and fracture opening degree. For normalization, this method statistically analyzes the maximum density of inclusions developed after the mineralization period in the mining area. and the maximum value of the pulse body filling width For example, the mineralization infill degree index of each fracture structure can be calculated using the following formula:

[0096]

[0097] in, The mineralization filling degree index ranges from 0 to 1. A higher value indicates a greater degree of mineralization filling and sealing of the fault structure, signifying stronger post-mineralization self-sealing and lower activity as a potential water inrush channel. This index integrates three indicators: post-mineralization inclusion density, vein filling width, and mineralization cementation degree. Post-mineralization inclusion density reflects the intensity of post-mineralization hydrothermal activity; denser inclusions indicate continued fluid activity after mineralization, potentially bringing new filling material. Vein filling width reflects the proportion of space occupied by filling material in the fault zone; a wider width indicates more complete filling. Mineralization cementation degree reflects the consolidation state of the filling material; a higher degree of cementation indicates that the filling material has hardened and sealed, making reactivation more difficult. Multiplying the first two terms by the complement of the third term provides a comprehensive characterization of the degree of self-sealing of the fault structure. This method can calculate the corresponding mineralization filling degree index for each fault structure, and the calculation results for all fault structures constitute a mineralization filling degree dataset.

[0098] Mechanical parameters are extracted from the obtained geometric and kinematic features of the fracture structure and modern geostress field data. The attitude, displacement, friction coefficient, and the angle between the direction of the maximum principal stress and the fracture strike of each fracture structure are obtained, and a set of mechanical parameters of the fracture structure is generated.

[0099] Specifically, this method can extract geometric characteristic data of each fault structure from the geological exploration report of the mining area, including the fault's strike, dip, dip angle, extension length, and displacement. The attitude parameter is used to subsequently calculate the angle between the direction of the maximum principal stress and the fault strike. This method also extracts kinematic characteristic data of each fault structure from the structural geological study report of the mining area, including the nature of the fault, displacement, and sliding direction. Combined with rock mechanics test results, the friction coefficient μ of each fault structure is determined. This coefficient is obtained through indoor direct shear tests on rock samples collected from the fault zone in the field. Different normal stresses are applied during the test, and the corresponding shear strength is measured. The friction coefficient is then fitted according to the Mohr-Coulomb criterion. Finally, this method extracts modern geostress field data from the geostress measurement report of the mining area. Typically, the magnitude and direction of the three principal stresses are measured using hydraulic fracturing or acoustic emission methods. The projection direction of the maximum principal stress direction onto the horizontal plane is determined, and then the angle θ between this direction and the strike of each fault structure is calculated, in degrees, ranging from 0° to 90°. This method stores the extracted mechanical parameters of each fracture structure in a structured manner to form a set of mechanical parameters of fracture structures. This set of parameters reflects the mechanical response characteristics of each fracture structure under the current stress field conditions.

[0100] Four-dimensional index fusion was performed on the paleofluid activity intensity dataset, mineralization filling degree dataset, and fracture structure mechanical parameter set. The activation potential index of each fracture structure was calculated through the activation potential assessment model to generate fracture structure activation potential data.

[0101] Specifically, this method uses the paleofluid activity intensity index of each fracture structure. Mineralization filling degree index coefficient of friction The sine value of the angle between the direction of the maximum principal stress and the fracture strike. and the sensitivity coefficient of mining disturbance The activation potential index is calculated using the following activation potential assessment model as input parameter:

[0102]

[0103] in, The activation potential index indicates that the higher the value, the greater the potential risk that the fault structure will be activated under mining disturbance and become a water inrush channel. For the weighting coefficients, satisfying The specific values ​​are determined based on the geological conditions of the mining area and engineering experience. Generally, the combination of paleofluid activity intensity and mineralization filling degree reflects the inherent properties of the fracture structure, the friction coefficient and stress angle term reflects the mechanical sensitivity of the fracture structure, and the mining disturbance sensitivity coefficient reflects the degree of disturbance of the fracture by mining activities. In the item, The proportion of the fracture structure that is not filled and sealed by mineralization is multiplied by the paleofluid activity intensity index to obtain the residual channel potential of the fracture structure after deducting the self-sealing effect. In the item, The coefficient of friction of the fracture structure reflects the magnitude of the frictional force that needs to be overcome for the fracture surface to slide. The value of the sine of the angle between the direction of the maximum principal stress and the fracture strike is given when the direction of the maximum principal stress is nearly parallel to the fracture strike. Approaching 0° When the shear stress is close to zero, it is difficult for the shear stress to generate an effective component on the fracture surface, resulting in a low activation probability. When the direction of the maximum principal stress is nearly perpendicular to the fracture strike, it approaches 90°. When the value is close to 1, the shear stress component is at its maximum, and the activation probability is relatively high. (Mining disturbance sensitivity coefficient) The activation potential index is determined based on the spatial distance and relative position of the fault to the initial mining area, as well as the mining thickness. Specifically, this method uses the minimum distance between the fault structure and the initial mining area, whether the fault is located on the roof or floor of the mining area, and the mining area thickness as inputs, and determines the activation potential index as follows: 0.8 when the distance between the fault and the initial mining area is less than 50 meters, 0.5 when the distance is between 50 and 200 meters, and 0.2 when the distance is greater than 200 meters; an additional coefficient of 1.2 when the fault is located on the immediate roof or floor of the mining area; and an additional coefficient of 1.1 when the mining area thickness is greater than 10 meters. This method calculates the corresponding activation potential index for each fault structure. The calculation results of all fracture structures constitute the data on the activation potential of fracture structures.

[0104] In one embodiment, a three-dimensional hydrogeological structure model of the mining area is generated by multi-source data collaborative inversion using a hierarchical Bayesian network on the paleofluid channel archives of the fault structure, supplementary exploration datasets, and acquired hydrogeological drilling data, including:

[0105] Coordinate transformation was performed on the paleofluid channel archives of fault structures, supplementary exploration datasets, and acquired hydrogeological drilling data. Then, the discrete inclusion data in the paleofluid channel archives of fault structures were converted into a paleofluid attribute field with spatial continuity through spatial interpolation, resulting in a multi-source data field with unified coordinates.

[0106] Specifically, this method performs coordinate transformation on the fault structure paleofluid channel archive, supplementary exploration dataset, and acquired hydrogeological drilling data. The fault structure paleofluid channel archive is a dataset containing information on fault paleofluid activity, formed in the early stage through core sample testing and spatial relocation. The supplementary exploration dataset consists of geophysical data and hydrological observation data collected specifically according to geological zoning levels. The hydrogeological drilling data comes from the logging and testing results of drilled boreholes in the mining area. This method uniformly transforms the three types of data into the CGCS2000 geodetic coordinate system to eliminate spatial reference differences. At the same time, it uses spatial interpolation methods such as Kriging interpolation and inverse distance weighted interpolation to convert the discretely distributed inclusion test data in the fault structure paleofluid channel archive into a continuously distributed paleofluid attribute field, resulting in a multi-source data field with a unified coordinate reference and spatial continuity, providing standardized data input for subsequent collaborative inversion.

[0107] For example, this method uses the Kriging interpolation method to spatially interpolate discretely distributed inclusion data in the paleofluid channel archive of fracture structures, and the paleofluid property estimate at unknown points can be calculated using the following formula:

[0108]

[0109] in, Points to be estimated The estimated paleofluid properties at the location, For known inclusion points The measured attribute values ​​at the location, The Kriging weights are determined by the semi-variogram model and the spatial distance between the point to be estimated and the known points.

[0110] Using paleofluid property fields as prior constraints, geophysical data from supplementary exploration datasets as the main inversion data, and hydrogeological drilling data as calibration and verification data, a hierarchical Bayesian network is used to construct the probabilistic dependencies between different data sources, generating a collaborative inversion parameter set.

[0111] For example, the method denotes the parameters to be inverted from the three-dimensional hydrogeological structure model of the mining area as follows: Including the width of each fracture zone ,inclination , extension Permeability anisotropy coefficient and the porosity distribution of each aquifer The Bayesian network of this method uses the spatial distribution of the paleofluid property field as the prior probability distribution of the model parameters. This prior distribution reflects the geological constraints imposed by the fluid activity record during the mineralization period on the current hydrogeological structure: for fault structures with a high paleofluid activity intensity index, the prior probability assigns them a larger fracture zone width and permeability coefficient; for fault structures with a high mineralization filling degree index, the prior probability assigns them a smaller permeability coefficient. This method uses geophysical data from the supplementary exploration dataset as the main inversion data, denoted as... Its relationship with the model parameters is expressed through the likelihood function. The likelihood function is constructed based on the residuals between geophysical data and the model's forward response. For example, for three-dimensional DC resistivity tomography data, the forward response is the apparent resistivity distribution calculated based on model parameters, and the likelihood function adopts an exponential form of the squared residuals. This method uses hydrogeological drilling data as calibration and validation data, denoted as... Its relationship with the model parameters is expressed through the likelihood function. The likelihood function is constructed based on the differences between measured values ​​and model predictions of formation interface depth, water level depth, and permeability coefficient revealed by borehole drilling. The probabilistic dependencies of the hierarchical Bayesian network are established using the following formula:

[0112]

[0113] in, Given geophysical and drilling data, the posterior probability distribution of model parameters is defined. This posterior distribution integrates constraint information from the three types of data. The method samples the posterior distribution using a Markov chain Monte Carlo sampling algorithm to generate a series of model parameter samples. The mean, standard deviation, and other statistics of the parameters are extracted from the samples to form a collaborative inversion parameter set containing geological constraint parameters, geophysical response parameters, and verification control parameters. This parameter set quantifies the uncertainty of different data sources and transmits it to the inversion process, realizing the organic integration of multi-source data.

[0114] A three-dimensional geological structure model of the mining area is generated by iterative inversion of the parameter set and the model response is matched with the observation data until the fitting error is less than the preset accuracy threshold.

[0115] Specifically, this method constructs a three-dimensional geological framework based on a collaborative inversion parameter set. It employs a modeling approach combining surface and volume models, constructing a three-dimensional geometric framework using stratigraphic and fault interface data. Then, aquifer parameters, hydraulic conductivity, and other attribute data from the collaborative inversion parameter set are assigned to the geological framework using kriging interpolation or co-kriging interpolation methods, forming an initial three-dimensional hydrogeological model. This method performs iterative inversion through a probabilistic inference mechanism using a hierarchical Bayesian network: in the... In the next iteration, based on the current model parameters Calculate the forward response of geophysical data and borehole data prediction values The fitting error between the model response and the observed data is calculated using the following formula:

[0116]

[0117] in, For the first The overall fitting error of the next iteration. This represents the total number of geophysical data points. For the first One geophysical observation value, For the current model parameters, the first Forward response values ​​of each geophysical observation point For the first Uncertainty of individual geophysical data; This represents the total number of borehole data points. For the first Measured values ​​from a borehole. This represents the predicted value of the i-th borehole under the current model parameters. For the first The method calculates the fitting error based on the uncertainty of individual borehole data. Compared with the preset accuracy threshold If a comparison is made, Then, according to the posterior probability distribution Adjust model parameters and generate new parameter samples. And proceed to the next iteration; if If the iteration fails, the iteration terminates, and the current model is used as the final output. Through the above iterative inversion process, the three-dimensional hydrogeological structure model of the mining area generated by this method can realistically depict the spatial distribution of strata, aquifers, and fault structures, as well as the heterogeneous distribution characteristics of hydrogeological parameters, providing a high-precision foundation for subsequent dynamic superposition of mining stress fields and prediction of hidden disaster-causing factors.

[0118] In one embodiment, a dynamic superposition of the mining-induced stress field is performed on a three-dimensional hydrogeological structure model of the mining area to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, in order to generate a survey report on hidden disaster-causing factors, including:

[0119] Based on the rock mechanics parameter library and mining scheme of the mining area, numerical simulation of mining-induced stress field is carried out on the three-dimensional hydrogeological structure model of the mining area. The shear stress increment and normal stress change at the location of the fracture structure in each mining stage are calculated, and a time series dataset of mining-induced stress field is generated.

[0120] Specifically, the rock mechanics parameter database for the mining area originates from previous geological exploration and rock mechanics tests. This includes physical and mechanical parameters such as density, elastic modulus, Poisson's ratio, internal friction angle, and cohesion obtained from laboratory tests on various lithological samples collected from borehole cores, as well as regional stress magnitude and direction data measured through on-site hydraulic fracturing or acoustic emission methods. The mining plan is extracted from mine design documents, including the excavation sequence of each mining section, stope dimensions, backfill parameters, pillar placement plan, and mining boundary expansion timeline. This method discretizes the three-dimensional hydrogeological structure model of the mining area into a finite element mesh, assigning corresponding rock mechanics parameters to each element. Then, it uses the finite element or finite difference method to numerically simulate the mining-induced stress field. During the simulation, this method gradually activates or deactivates corresponding mesh elements according to the time sequence set in the mining plan to simulate the formation and expansion of the goaf. At each time step, it calculates the stress distribution across the entire field and extracts the shear stress increment at the location of each fracture structure. and normal stress variation Among them, the shear stress increment reflects the accumulation of tangential stress on the fracture surface caused by mining activities, and the normal stress change reflects the change of normal stress on the fracture surface. This method generates a mining-induced stress field time series dataset by storing the shear stress increment and normal stress change of each mining stage in a time series. This dataset quantifies the mechanical disturbance process of mining activities on the fracture structure.

[0121] A coupled analysis was performed on the time series dataset of mining-induced stress field and the data on the activation potential of fracture structures. The shear stress increment at each mining stage was compared with the shear strength of the fracture structure. The activation probability of each fracture structure at different mining stages was calculated using the activation probability prediction formula, and time series data of fracture activation probability were generated.

[0122] Specifically, the data on the activation potential of fracture structures are derived from the multi-index correlation assessment results of the paleofluid channel archives and fracture structure characteristics of fracture structures obtained in the aforementioned steps, including the paleofluid activity intensity index of each fracture structure. Mineralization filling degree index and the friction coefficient extracted from mechanical parameters and normal stress This method first considers the friction coefficient of the fracture structure. and normal stress Combined with the mineralization filling degree index The critical activation shear stress of the fracture structure is determined by the following formula:

[0123]

[0124] in, The critical activation shear stress, The friction coefficient of the fracture structure is denoted as . The initial normal stress on the fracture structure, This represents the change in normal stress caused by mining. The mineralization filling degree index is given by the formula, where the change in normal stress is the value of the index. Based on the time-series dataset of mining-induced stress fields, its sign determines the increase or decrease of normal stress on the fracture surface; mineralization filling degree index As a reduction factor, fracture structures with higher filling degree and denser cementation have lower critical activation shear stresses and are more prone to activation under smaller shear stress disturbances. Subsequently, this method calculates the activation probability for each mining stage using the following activation probability prediction formula:

[0125]

[0126] in, Mining stage The activation probability of fracture structures ranges from 0 to 1; The sensitivity coefficient is determined based on the geometric characteristics of the fracture structure and the engineering geological conditions of the mining area, and is usually taken as a value between 0.5 and 2.0. Mining stage The shear stress increment on the fracture structure is extracted from the mining-induced stress field time series dataset. The critical activation shear stress is defined by this formula, which uses a sigmoid function. When the shear stress increment is much smaller than the critical activation shear stress, the activation probability approaches 0; when the shear stress increment is much larger than the critical activation shear stress, the activation probability approaches 1. Near the critical value, the probability changes rapidly in an S-shape with the shear stress increment. This method calculates the activation probability for each mining stage of each fracture structure. All calculation results are organized by fracture number and mining stage to generate time-series data of fracture activation probabilities.

[0127] Hydraulic connectivity analysis was performed on the time series data of fracture activation probability. For fracture structures whose activation probability exceeded the preset warning threshold, the hydraulic gradient of surface water, seepage path length and equivalent permeability coefficient were calculated based on the three-dimensional hydrogeological structure model of the mining area to obtain the location data of water inrush channels.

[0128] For example, this method can set a preset warning threshold of 0.6, meaning that when the activation probability of a fault structure reaches 60% or more in a certain mining stage, it is classified as a high-risk fault and subjected to hydraulic connectivity analysis. For each high-risk fault structure, the method extracts the spatial distribution of the fault fracture zone, its spatial relationship with surface water, and the permeability distribution inside and outside the fault zone based on a three-dimensional hydrogeological structure model of the mining area. The hydraulic gradient is calculated using the following formula:

[0129]

[0130] in, For hydraulic gradient, The elevation of surface water levels is obtained from long-term hydrological observation data. The water level elevation of the activated section of the fault structure was extracted from the three-dimensional hydrogeological structure model of the mining area. The seepage path length, i.e., the shortest seepage distance between the surface water body and the activated section of the fault structure, is calculated using three-dimensional spatial distance. Simultaneously, the equivalent permeability coefficient is calculated. For a composite medium composed of fractured zones and intact bedrock, the equivalent permeability coefficient along the seepage path can be calculated using the following formula:

[0131]

[0132] in, The equivalent permeability coefficient, For the first The length of the seepage path, This is the permeability coefficient corresponding to this path segment. The method uses the hydraulic gradient to represent the total number of segments in the seepage path. Equivalent permeability coefficient and the length of the penetration path A comprehensive assessment is conducted. When the hydraulic gradient is large, the equivalent permeability coefficient is high, and the path length is short, it is determined that the activated fault structure can establish an effective hydraulic connection with the surface water body, forming a water inrush channel. The spatial coordinates and extension direction of the channel are recorded to generate water inrush channel location data.

[0133] Correlation of prevention and control measures is carried out with time series data on fault activation probability, data on water inrush channel location, and risk level classification standards. The location of monitoring points is determined based on the activation probability of each fault structure and the development degree of water inrush channels, and a survey report on hidden disaster-causing factors is generated.

[0134] For example, this method can classify each fault structure into three risk levels based on the time series data of fault activation probability. For instance, a activation probability greater than 0.8 is classified as high-risk, between 0.6 and 0.8 as medium-risk, between 0.4 and 0.6 as low-risk, and activation probabilities below 0.4 are not included in the risk control scope. For fault structures identified as having water inrush channels, the risk level is increased by one regardless of the activation probability. This method can also determine the location of monitoring points based on the risk level: for high-risk fault structures, no fewer than three water level and stress monitoring points are set at the intersection of the fault zone and the mining area, the midpoint of the line connecting the fault zone and the surface water body, and key locations on the hanging wall of the fault, with a monitoring frequency of once a day; for medium-risk fault structures, one to two monitoring points are set at the intersection of the fault zone and the mining area, with a monitoring frequency of twice a week; for low-risk fault structures, only one benchmark monitoring point is set near the fault zone, with a monitoring frequency of once a month. Furthermore, this method can match corresponding prevention and control measures according to the risk level: for high-risk fracture structures with water inrush channels, a comprehensive prevention and control measure combining grouting curtain reinforcement and drainage depressurization is adopted, with the grouting range covering the fracture zone and 20 meters on both sides, and the spacing of drainage holes arranged according to the radius of influence; for medium-risk fracture structures, local grouting reinforcement or the reservation of waterproof pillars are adopted; for low-risk fracture structures, daily inspection and periodic observation measures are adopted. Even further, this method summarizes all analysis results to form a hidden disaster-causing factor survey report, which includes the risk level classification of each fracture structure, the coordinates and monitoring frequency of monitoring points, prevention and control measures recommendations, and risk evolution trend diagrams at different mining stages. This report is presented in a combination of text and graphics, providing a scientific basis for mines to formulate phased safety production plans and emergency prevention and control plans.

[0135] In summary, the intelligent hydrogeological exploration method for mining areas based on geological data provided in this application achieves refined reconstruction and spatial positioning of fluid activity trajectories during geological history by constructing a paleofluid channel archive of fracture structures that integrates thermodynamic parameters of fluid inclusions during the mineralization period with the geometric, kinematic, and mechanical characteristics of fracture structures. Based on this, a multi-index activation potential assessment model is established, incorporating paleofluid activity intensity index, mineralization filling degree index, and modern stress field characteristics. This model quantitatively couples the inherent channel attributes of fracture structures with their later self-closure degree and mechanical response sensitivity, achieving a transformation from qualitative geological information to quantitative risk classification. Furthermore, supplementary exploration projects are deployed differentiatedly based on the activation potential zoning results. A hierarchical Bayesian network is used to perform cross-scale collaborative inversion of the micro-scale paleofluid attribute field, the macro-scale geophysical response data, and the point-like high-precision drilling verification data, effectively overcoming the ambiguity of geophysical inversion by using paleofluid activity traces as prior constraints. Finally, through dynamic superposition of mining-induced stress fields and activation probability prediction, the temporal prediction of fracture structure activation risk and water inrush channel location at different mining stages is achieved. This technical solution solves the technical challenges of identifying "dormant-activated" hidden disaster-causing factors in complex geological mining areas, insufficient cross-scale fusion of multi-source heterogeneous data, and the inability of static hydrogeological models to support dynamic risk prediction throughout the entire life cycle. It organically integrates the evolution law of metallogenic systems with the hydrogeological exploration technology system, realizing a technical leap from the current general survey of water-bearing conditions to the prediction of active risks throughout the entire life cycle. It provides reliable technical support for the accurate general survey of hidden disaster-causing factors in mines, the dynamic optimization of water control projects, and the intelligent management and control of safe production.

[0136] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0137] Based on the same inventive concept, this application also provides a data-based intelligent hydrogeological exploration system 10 for implementing the above-mentioned data-based intelligent hydrogeological exploration method for mining areas. The solution provided by this system is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more embodiments of the data-based intelligent hydrogeological exploration system 10 provided below can be found in the limitations of the data-based intelligent hydrogeological exploration method for mining areas described above, and will not be repeated here.

[0138] In one exemplary embodiment, such as Figure 3 As shown, a smart hydrogeological exploration system 10 for mining areas based on geological data is provided, including:

[0139] The paleofluid archive construction module 11 is used to identify the mineralization period of the obtained mineral deposit exploration core samples, extract multi-stage fluid activity records of each fault structure, and generate paleofluid channel archives of the fault structure.

[0140] The activation potential assessment module 12 is used to conduct multi-index correlation assessment of the paleofluid channel archives of fracture structures and the acquired fracture structure characteristics, and to calculate the activation potential index of each fracture structure to generate fracture structure activation potential data.

[0141] The zoning exploration module 13 is used to divide the mining area into geological zones with different activation potential levels based on the fault structure activation potential data, and to obtain the corresponding supplementary exploration dataset according to the activation potential level of each geological zone.

[0142] The multi-source data modeling module 14 is used to perform multi-source data collaborative inversion on the fracture structure paleofluid channel archive, supplementary exploration dataset and acquired hydrogeological drilling data through a hierarchical Bayesian network to generate a three-dimensional hydrogeological structure model of the mining area.

[0143] The exploration prediction report module 15 is used to dynamically superimpose the mining-induced stress field on the three-dimensional hydrogeological structure model of the mining area, predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors.

[0144] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the intelligent hydrogeological exploration method for mining areas based on geological data as described above.

[0145] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0146] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0147] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for intelligent exploration of a mine site hydrogeology based on geological data, characterized in that, The method includes: The mineralization phases of the obtained mineral deposit exploration core samples were identified, and multi-phase fluid activity records of each fault structure were extracted to generate paleofluid channel archives of the fault structures. The paleofluid channel archives of the fracture structures and the acquired fracture structure features are evaluated using a multi-index correlation, and the activation potential index of each fracture structure is calculated to generate fracture structure activation potential data. Based on the fault structure activation potential data, the mining area is divided into geological zones with different activation potential levels, and corresponding supplementary exploration datasets are obtained according to the activation potential level of each geological zone. A three-dimensional hydrogeological structure model of the mining area is generated by multi-source collaborative inversion of the fracture structure paleofluid channel archive, the supplementary exploration dataset, and the acquired hydrogeological drilling data through a hierarchical Bayesian network. The three-dimensional hydrogeological structure model of the mining area is dynamically superimposed with the stress field of mining to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors.

2. The method of claim 1, wherein, The process involves identifying the mineralization phases of the obtained mineral deposit core samples, extracting multi-phase fluid activity records from each fault structure, and generating paleofluid channel archives for the fault structures, including: Fluid inclusions were prepared based on the obtained core samples from mineral deposit exploration, and fluid inclusion combinations of different phases were identified to generate inclusion phase classification data. Based on the inclusion phase division data, microthermography is performed on the inclusions of the corresponding phases to obtain the homogenization temperature and freezing point temperature of each phase of the inclusions. The salinity of the inclusions is calculated using the empirical formula of freezing point temperature and salinity, generating a set of thermodynamic parameters of the inclusions that includes homogenization temperature, salinity and phase division. Based on the set of thermodynamic parameters of the inclusions, laser Raman spectroscopy analysis was performed on representative inclusions to obtain gas phase composition information of the inclusions and generate a dataset of gas phase composition of the inclusions. Spatial positioning is performed on the inclusion phase division data, the inclusion thermodynamic parameter set, and the inclusion gas phase composition dataset. The thermodynamic parameters of each phase of inclusions are associated with the associated fracture structures to generate a fracture structure paleofluid channel archive containing the intensity of ore-forming fluid activity, activity phase, and filling mineralization characteristics of each fracture structure.

3. The method of claim 1, wherein, The process involves a multi-index correlation evaluation of the paleofluid channel archives of the fracture structures and the acquired fracture structure features, and the calculation of the activation potential index for each fracture structure to generate fracture structure activation potential data, including: Based on the paleofluid channel archives of the fracture structures, the average homogeneous temperature, average salinity, and gas phase composition of the inclusions during the mineralization period of each fracture structure were extracted, and the paleofluid activity intensity index of each fracture structure was calculated to generate a paleofluid activity intensity dataset. Based on the paleofluid channel archives of the fracture structures, the density of inclusions, the width of vein filling and the degree of mineralization cementation after the mineralization period of each fracture structure are extracted, and the mineralization filling degree index of each fracture structure is calculated to generate a mineralization filling degree dataset. Mechanical parameters are extracted from the obtained geometric and kinematic features of the fracture structure and modern geostress field data. The attitude, displacement, friction coefficient and the angle between the direction of the maximum principal stress and the fracture strike of each fracture structure are obtained, and a set of mechanical parameters of the fracture structure is generated. The paleofluid activity intensity dataset, the mineralization filling degree dataset, and the fracture structure mechanical parameter set are fused into four-dimensional indices. The activation potential index of each fracture structure is calculated through the activation potential assessment model to generate fracture structure activation potential data.

4. The method of claim 1, wherein, The process involves multi-source data collaborative inversion using a hierarchical Bayesian network on the paleofluid channel archives of the fracture structure, the supplementary exploration dataset, and the acquired hydrogeological drilling data to generate a three-dimensional hydrogeological structure model of the mining area, including: The coordinate transformation is performed on the fracture structure paleofluid channel archive, the supplementary exploration dataset and the acquired hydrogeological drilling data. The discrete inclusion data in the fracture structure paleofluid channel archive is converted into a paleofluid attribute field with spatial continuity through spatial interpolation method, resulting in a multi-source data field with unified coordinates. Using the paleofluid property field as a priori constraint, the geophysical data in the supplementary exploration dataset as the main inversion data, and the hydrogeological drilling data as the calibration and verification data, a collaborative inversion parameter set is generated by constructing the probabilistic dependencies between different data sources through a hierarchical Bayesian network. The collaborative inversion parameter set is used to model the three-dimensional geological structure. The model is iterated until the fitting error between the model response and the observation data is less than a preset accuracy threshold, thereby generating a three-dimensional hydrogeological structure model of the mining area.

5. The method of claim 1, wherein, The process involves dynamically superimposing the mining-induced stress field onto a three-dimensional hydrogeological structure model of the mining area to predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, thereby generating a survey report on hidden disaster-causing factors, including: Based on the rock mechanics parameter library and mining scheme of the mining area, the mining-induced stress field numerical simulation is carried out on the three-dimensional hydrogeological structure model of the mining area. The shear stress increment and normal stress change at the location of the fracture structure in each mining stage are calculated, and a mining-induced stress field time series dataset is generated. The time series dataset of mining-induced stress field and the data on the activation potential of fracture structures are coupled and analyzed. The shear stress increment of each mining stage is compared with the shear strength of the fracture structure. The activation probability of each fracture structure at different mining stages is calculated by the activation probability prediction formula, and time series data of fracture activation probability is generated. Hydraulic connectivity analysis is performed on the time series data of the fracture activation probability. For fracture structures whose activation probability exceeds the preset warning threshold, the hydraulic gradient of the surface water body, the seepage path length and the equivalent permeability coefficient are calculated based on the three-dimensional hydrogeological structure model of the mining area to obtain the location data of the water inrush channel. The time series data of the activation probability of the fracture, the location data of the water inrush channel, and the risk level classification criteria are correlated with prevention and control measures. The location of monitoring points is determined according to the activation probability of each fracture structure and the development degree of the water inrush channel, and a survey report of the hidden disaster-causing factors is generated.

6. A geological data based mine site hydrogeological intelligent exploration system for implementing the method of any one of claims 1 to 5, characterized in that, The system includes: The paleofluid archive construction module is used to identify the mineralization phases of the obtained mineral deposit exploration core samples, extract multi-phase fluid activity records of each fault structure, and generate paleofluid channel archives of the fault structures. The activation potential assessment module is used to perform multi-index correlation assessment on the paleofluid channel archives of the fracture structures and the acquired fracture structure features, and to calculate the activation potential index of each fracture structure to generate fracture structure activation potential data. The zoning exploration module is used to divide the mining area into geological zones with different activation potential levels based on the fault structure activation potential data, and to obtain corresponding supplementary exploration datasets according to the activation potential level of each geological zone. The multi-source data modeling module is used to perform multi-source data collaborative inversion on the fracture structure paleofluid channel archive, the supplementary exploration dataset, and the acquired hydrogeological drilling data through a hierarchical Bayesian network to generate a three-dimensional hydrogeological structure model of the mining area. The exploration and prediction report module is used to dynamically superimpose the mining-induced stress field on the three-dimensional hydrogeological structure model of the mining area, predict the activation probability of each fault structure and the spatial location of water inrush channels at different mining stages, so as to generate a survey report on hidden disaster-causing factors. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-6 when the computer program is executed by the processor. When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images