Multi-source fusion calibration method for hydrological parameters in complex coalfield structures

Abstract

This application discloses a multi-source fusion calibration method for hydrological parameters in complex geological coalfields. The invention establishes a physical mapping model between structural geometric parameters and hydrological response parameters by conducting in-situ pressure tests, fracture network seepage simulations, and mechanical tests within each water-controlling unit, generating a set of quantitative indicators for structural water control effects. Compared to the subjective methods in existing technologies that rely on expert experience scoring, semi-quantitative grading, or simple geometric statistics to establish quantitative indicators, this invention uses the actual water conductivity of the rock mass, fracture seepage patterns, and stress-fracture coupling responses as physical bases. This provides in-situ physical calibration support for the quantitative indicators, eliminates human uncertainty, and solves the problem of inaccurate water control level judgment due to differences in observation scale and stress state in the same structural location, thereby improving the objectivity and reliability of hydrological parameter characterization in complex structural areas.

Description

Technical Field

[0001] This application relates to the field of hydrological technology, and in particular to a multi-source fusion calibration method for hydrological parameters in complex coalfield structures. Background Technology

[0002] In geologically complex areas, the widespread development of faults, folds, and structural composites leads to fragmented spatial structures in coal-bearing aquifers and spatial variability in hydrogeological parameters, placing higher demands on the accuracy of hydrogeological parameter detection. In recent years, hydrogeological parameter calibration methods integrating multi-source detection data such as transient electromagnetic data, borehole television imaging, and dynamic hydrogeological monitoring have become important technical means for evaluating the water-bearing capacity and inverting permeability coefficients of aquifers in complex geological coalfields. Through multi-source information complementarity and cross-validation, the reliability of hydrogeological models is improved.

[0003] However, existing multi-source fusion calibration methods for hydrological parameters in complex structural coalfields still have significant shortcomings in quantifying the structural water control effect and allocating fusion weights. The quantitative indicators used in existing technologies, such as fracture network connectivity index, structural hydraulic conductivity classification coefficient, and fault activation probability, are mostly based on expert experience scoring, semi-quantitative classification, or simple geometric statistics, lacking the support of objective physical evidence such as in-situ pressure water tests, fracture seepage physical simulations, and structural rock mechanics tests. In the process of fusion weight allocation, it fails to effectively identify the differences in hydrological attributes caused by variations in observation scale, infill material properties, and stress states within the same structural region. This results in artificial uncertainties in the basic probability allocation and weight coefficients in the DS evidence theory or entropy weight method, leading to weight imbalances in structurally complex regions. Consequently, the fusion calibration results deviate significantly from the actual hydrogeological conditions in key structural regions such as fault intersections and fold axes, making it difficult to meet the engineering requirements for precise water hazard control in complex structural coalfields. Summary of the Invention

[0004] The main objective of this application is to provide a multi-source fusion calibration method for hydrological parameters in complex coalfield structures, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this application provides the following technical solution: a multi-source fusion calibration method for hydrological parameters in complex coalfield structures, the specific steps of which are as follows: S1. Detailed structural analysis: A three-dimensional structural model is constructed based on coalfield exploration data, and water-controlling units are divided according to structural mechanical properties and fracture characteristics; S2. Water control effect calibration: In-situ pressure water tests and fissure seepage simulations are carried out in each water control unit to establish a physical mapping between structural parameters and hydrological response; S3. Data decoupling and adaptation: Establish an error correction model for multi-source detection within the water control unit, convert borehole-scale parameters to unit-scale parameters, and achieve multi-source data decoupling; S4. Dynamic weight optimization: Construct a confidence function with physical calibration results as constraints, establish an adaptive weight mechanism based on response accuracy, and adjust the data source weight in the construction of composite parts. S5. Closed-loop iterative correction: Input the fused and calibrated hydrological parameters into the numerical simulation model for verification. When the error exceeds the threshold, it is fed back to step S2 and step S4 for iterative correction.

[0006] Preferably, in step S1, the specific method for constructing the fine-grained analysis is as follows: S1.1 Structural geometric modeling: Based on coalfield drilling, 3D seismic and mining geological data, fault attitude, fold morphology and structural intersection relationship are extracted to construct a 3D structural geological model, quantitatively characterize the degree of stratigraphic fragmentation and the spatial distribution law of structural elements, establish the structural framework of the study area, and provide geometric basis for the division of water-controlling units. S1.2 Water Control Unit Division: Based on the tectonic mechanical properties, fracture development degree and stratigraphic fracture characteristics, combined with the fault water conductivity, fold axis fracture enrichment degree and tectonic composite water-holding conditions in the three-dimensional tectonic model, the water control level of the study area is evaluated, and the fault water-conducting zone, fold axis fracture enrichment zone, tectonic composite water-holding zone and intact bedrock water-retaining zone are divided to form a spatial framework of water control units.

[0007] Preferably, in step S2, the specific method is as follows: S2.1 In-situ multi-parameter testing: In-situ pressure water tests are carried out in each water control unit to obtain the permeability coefficient and hydraulic conductivity of the rock mass. Simultaneously, triaxial mechanical tests and acoustic emission monitoring of the tectonic rock mass are conducted to analyze the evolution law of fracture aperture and the characteristics of abrupt change in hydraulic conductivity under different stress states. Physical response parameters of tectonic water control effect are obtained and water control level thresholds are calibrated. S2.2 Fracture Network Simulation: Based on core scanning imaging and 3D reconstruction, a physical model of the fracture network is established, seepage simulation is carried out, and the quantitative relationship between fracture aperture, connectivity and permeability is obtained. By combining the results of in-situ testing and simulation, a physical mapping model between structural geometric parameters and hydrological response parameters is established, and a set of quantitative indicators of structural water control effect is generated.

[0008] Preferably, in step S2.1, the in-situ multi-parameter test is performed in the following manner: S2.11 In-situ pressure water test: Select representative test sections according to structural parts within each water control unit, conduct segmented pressure water tests, determine the permeability coefficient and hydraulic conductivity of the rock mass, analyze the characteristics of the pressure-flow dynamic response curve, extract the true hydraulic conductivity parameters of the rock mass in different structural parts, and integrate them to form a basic hydrological response dataset for the water control unit. S2.12 Mechanical Coupling Monitoring: Simultaneously conduct triaxial mechanical tests on the tectonic rock mass, apply gradient confining pressure and axial pressure loads, use acoustic emission to capture the evolution sequence of fracture initiation, propagation and penetration in real time, analyze the evolution law of fracture aperture and the characteristics of abrupt changes in water conductivity under different stress states, calibrate the threshold of water control level, and establish the coupling discrimination relationship between stress and fracture water conductivity.

[0009] Preferably, in step S2.2, the specific method for simulating the fracture network is as follows; S2.21, Fracture Network Reconstruction: Based on core computer tomography and 3D reconstruction technology, the aperture, trace length, density and spatial connectivity characteristics of fractures inside the core are extracted, a discrete fracture network physical model is established, and the topological structure, geometric connectivity characteristics and seepage channel development degree of the fracture network are quantitatively characterized. S2.22, Seepage Law Mapping: Numerical seepage simulation is carried out based on a discrete fracture network model to analyze the quantitative response laws of fracture aperture, connectivity and equivalent permeability coefficient. By combining the results of in-situ pressure water test and seepage simulation, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, and a set of quantitative indicators of structural water control effect is generated.

[0010] Preferably, in step S3, the specific method for data decoupling and adaptation is as follows: S3.1 Multi-source error decoupling: Analyze the physical field response mechanism and applicable conditions of transient electromagnetic, borehole television and hydrological monitoring in structurally complex areas, identify the sources of systematic errors and detection blind spots of each detection method in different water control units, establish a regional differentiated error correction model, decouple and map the geophysical volume response data to the corresponding water control unit, and eliminate the multi-source detection system deviation. S3.2 Scale Adaptation and Conversion: Based on the aforementioned physical mapping model, the borehole-scale fracture network parameters are extended and converted to the structural unit volume scale to establish a cross-scale bridging relationship between borehole point measurements and geophysical volume response, thereby achieving accurate adaptation and spatial consistency registration of multi-source heterogeneous data within the structural spatial framework.

[0011] Preferably, in step S4, the specific method for dynamic weight optimization is as follows: S4.1 Credibility Calibration: Using the results of in-situ pressure water test as hard constraints and the data of fracture seepage simulation and rock mechanics test as soft constraints, a multi-source data credibility evaluation function is constructed. Based on the physical response accuracy and data quality of each detection method in different water control units, the confidence level of each data source is calibrated, a credibility benchmark system is established, and calibration parameters are generated. S4.2 Dynamic Weight Optimization: An adaptive weight allocation mechanism is established. In the structural composite parts, the contribution weights of each data source are dynamically adjusted and redistributed based on the differences in structural stress state and crack filling characteristics. This achieves adaptive optimization and imbalance correction of the fusion weights of key structural parts, forming a multi-source optimized fusion weight configuration scheme and outputting the fusion weights.

[0012] Preferably, in step S4.1, the confidence level is determined in the following specific way: S4.11, Constraint Benchmark Construction: Using the in-situ pressure water test permeability coefficient and hydraulic conductivity as hard constraints, and the equivalent permeability coefficient of fracture seepage simulation and the fracture hydraulic conductivity characteristics of rock mechanics test as soft constraints, a credibility evaluation function is constructed, the water control unit type and the tectonic stress state correction coefficient are coupled, the response accuracy and data quality differentiation law of each detection method are analyzed, and evaluation indicators and physical constraint benchmarks are established. S4.12 Confidence Calibration: Based on the response accuracy, data quality and spatial coverage of each detection method in different water control units, and combined with the differences in tectonic stress state and fracture filling characteristics, the confidence levels of each data source of transient electromagnetic, borehole television and hydrological monitoring are graded and calibrated, multi-scale quantitative calibration parameters are generated, a confidence benchmark system is established and calibration results are formed.

[0013] Preferably, in step S4.2, the specific method for dynamic weight optimization is as follows: S4.21 Weight Initialization: Based on the confidence calibration results and the physical constraint benchmark, an adaptive weight allocation mechanism is established. According to the confidence level and data quality differences of each detection method in different water control units, the initial contribution weight of each data source is calculated to form the initial fusion weight configuration. S4.22 Dynamic Imbalance Correction: In structural composite parts, combining the differences in structural stress state and crack filling characteristics, the sudden change law of confidence of each data source is analyzed, the initial contribution weight is dynamically adjusted and redistributed, the weight imbalance of key structural parts is corrected, and adaptive optimized fusion weights are generated and output.

[0014] Preferably, in step S5, the specific method for closed-loop iterative correction is as follows: S5.1 Numerical simulation verification: Input the fused and calibrated hydrological parameters into the groundwater flow numerical simulation model, perform multi-objective fitting verification with measured water level, inflow and water quality data, establish an uncertainty evaluation system, calculate the confidence interval and sensitivity coefficient of hydrological parameters of each water control unit, analyze the source of verification deviation and extract deviation feature information; S5.2 Closed-loop iterative correction: If the verification error exceeds the preset threshold, the deviation information is fed back to step S2 to correct the physical mapping model parameters, and then fed back to step S4 to adjust the weight allocation strategy, forming a closed-loop iterative correction process of physical calibration, fusion optimization and verification feedback, so as to realize the continuous dynamic calibration of hydrological parameters.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention establishes a physical mapping model between structural geometric parameters and hydrological response parameters by conducting in-situ pressure water tests, fracture network seepage simulations, and mechanical tests within each water-controlling unit, generating a set of quantitative indicators for structural water control effects. Compared to the subjective methods in existing technologies that rely on expert experience scoring, semi-quantitative grading, or simple geometric statistics to establish quantitative indicators, this invention uses the actual water conductivity of the rock mass, fracture seepage laws, and stress-fracture coupling response as physical basis, enabling the quantitative indicators to obtain in-situ physical calibration support, eliminating human uncertainty, and solving the problem of inaccurate water control level judgment due to differences in observation scale and stress state in the same structural part, thereby improving the objectivity and reliability of hydrological parameter characterization in complex structural areas.

[0016] 2. This invention uses in-situ pressure water test results as hard constraints and fracture seepage simulation and rock mechanics test data as soft constraints to construct a multi-source data reliability evaluation function and establish an adaptive weight allocation mechanism under physical constraints. It dynamically adjusts the contribution weights of each data source in structurally complex areas by considering differences in stress state and filling characteristics. Compared to existing technologies that rely on expert experience scoring to establish basic probability allocation, leading to arbitrary uncertainty in the weights of structurally complex areas, this invention integrates objective physical calibration results throughout the entire weight allocation process. This achieves adaptive redistribution and imbalance correction of the fusion weights for key structural areas, improving the objectivity and reliability of the fusion results of hydrological parameters in complex structural coalfields. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of the method described in this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0019] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] Example 1: Please refer to Figure 1 A multi-source fusion calibration method for hydrological parameters in complex coalfield structures, the specific steps of which are as follows: S1. Detailed structural analysis: A three-dimensional structural model is constructed based on coalfield exploration data, and water-controlling units are divided according to structural mechanical properties and fracture characteristics; S2. Water control effect calibration: In-situ pressure water tests and fissure seepage simulations are carried out in each water control unit to establish a physical mapping between structural parameters and hydrological response; S3. Data decoupling and adaptation: Establish an error correction model for multi-source detection within the water control unit, convert borehole-scale parameters to unit-scale parameters, and achieve multi-source data decoupling; S4. Dynamic weight optimization: Construct a confidence function with physical calibration results as constraints, establish an adaptive weight mechanism based on response accuracy, and adjust the data source weight in the construction of composite parts. S5. Closed-loop iterative correction: Input the fused and calibrated hydrological parameters into the numerical simulation model for verification. When the error exceeds the threshold, it is fed back to step S2 and step S4 for iterative correction.

[0022] In this embodiment: Step S1 involves multi-source fusion interpretation based on coalfield drilling, 3D seismic, and mining geological data to construct a 3D structural geological model that includes fault attitude, fold morphology, and structural intersection relationships. Based on structural mechanical properties, fracture development, and stratigraphic fragmentation characteristics, the model is divided into fault-conducting zones, fold axis fracture-rich zones, structurally composite perched zones, and intact bedrock aquitard zones, forming a spatially targeted framework for water-controlling units. This step transforms the structural framework of complex structural areas into water-controlling unit partitions, achieving precise matching between structural geometry and hydrological boundary conditions. This provides a targeted partitioning basis for subsequent in-situ physical testing, solving the problem of spatial disconnect between structural and hydrological models in existing technologies and improving the precision of the spatial representation of structural water-controlling effects.

[0023] Step S2 involves conducting in-situ pressure water tests within each water-controlling unit to obtain the rock mass permeability and hydraulic conductivity. Based on core scanning imaging and 3D reconstruction, a physical model of the fracture network is established, and seepage simulation is performed. Simultaneously, triaxial mechanical testing and acoustic emission monitoring of the tectonic rock mass are conducted to analyze the fracture evolution patterns and abrupt changes in hydraulic conductivity under different stress states. A physical mapping model between tectonic geometric parameters and hydrological response parameters is established, generating a quantitative index set of tectonic water-controlling effects. This step transforms the tectonic water-controlling effect from an empirical qualitative description to a physical quantitative characterization, establishing an objective calibration system based on in-situ and simulation coupling. This eliminates the subjective uncertainties of traditional expert-based scoring and semi-quantitative grading, providing a reliable physical constraint benchmark for weighted fusion.

[0024] Step S3 analyzes the physical field response mechanism and applicable boundaries of transient electromagnetic, borehole television, and hydrological monitoring in structurally complex areas. It identifies the sources of systematic errors and detection blind spots of each detection method in different water control units, establishes a zone-specific error correction model, and extends and transforms borehole-scale fracture network parameters to the structural unit volume scale based on the physical mapping model, establishing cross-scale bridging relationships. This step achieves decoupling and scale adaptation of the physical responses of multi-source heterogeneous data within the structural spatial framework, eliminates systematic deviations caused by structural shielding and lithological phase transitions of different detection methods, realizes spatial consistency registration between borehole point measurements and geophysical volumetric responses, and improves the basic quality of multi-source data fusion in key structural locations.

[0025] Step S4 uses in-situ pressure water test results as hard constraints and fracture seepage simulation and rock mechanics test data as soft constraints to construct a multi-source data reliability evaluation function. Based on the physical response accuracy of each detection method within different water control units, an adaptive weight allocation mechanism is established. In structurally complex locations, the contribution weights of each data source are dynamically adjusted in conjunction with the tectonic stress state and fracture filling characteristics. This step transforms the fusion weights from subjective experience-based assignment to adaptive optimization under physical constraints, realizing the redistribution and imbalance correction of fusion weights in key structural locations. It eliminates the artificial uncertainty in the basic probability allocation of traditional fusion algorithms, ensuring the objectivity and reliability of the hydrological parameter fusion results in structurally complex locations.

[0026] Step S5 inputs the fused and calibrated hydrological parameters into the groundwater flow numerical simulation model, performs multi-objective fitting verification with measured water level, inflow, and water quality data, establishes an uncertainty evaluation system, calculates the confidence interval and sensitivity coefficient of hydrological parameters for each water control unit, and when the verification error exceeds a preset threshold, feeds back the deviation information to the physical mapping model correction and weight allocation strategy adjustment stages. This step achieves accuracy verification and error source feedback of the fusion results, forming a closed-loop iterative correction process of physical calibration, fusion optimization, and verification feedback, realizing dynamic and continuous calibration of hydrological parameters, solving the problem of the lack of uncertainty transmission and feedback correction mechanisms in existing technologies, and ensuring the long-term reliability of hydrogeological models for complex geological coalfields.

[0027] Compared to existing technologies that rely on expert scoring, semi-quantitative grading, or simple geometric statistics for quantifying structural water control effects, and whose weight allocation lacks objective physical basis, leading to imbalanced weight allocation in structural composite parts and distorted fusion results, this method achieves precise mapping of structural geometry to hydrological boundaries by constructing a three-dimensional structural model and dividing it into water-controlling units. Through coupled calibration of in-situ pressure water tests, fracture network seepage simulation, and structural rock mechanics testing, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, providing in-situ physical basis for quantifying structural water control effects. By decoupling and cross-scale adaptation of multi-source data physical responses, systematic biases of different detection methods in structurally complex areas are eliminated. Through credibility evaluation under physical constraints and an adaptive weight allocation mechanism, objective calibration results are embedded into the fusion process to solve the problem of imbalanced weight allocation in structural composite parts. Finally, through numerical simulation verification and closed-loop iterative correction, dynamic and continuous optimization of hydrological parameters is achieved. The overall solution elevates the structural water control effect from empirical qualitative judgment to physical quantitative driving, optimizes the fusion weight allocation from subjective uncertainty to objective adaptive configuration, improves the calibration accuracy and reliability of hydrogeological parameters in key structural parts of complex coalfields, and provides technical support for precise prevention and control of mine water hazards.

[0028] Example 2: Please refer to Figure 1 In step S1, the specific method for constructing the refined analysis is as follows: S1.1 Structural geometric modeling: Based on coalfield drilling, 3D seismic and mining geological data, fault attitude, fold morphology and structural intersection relationship are extracted to construct a 3D structural geological model, quantitatively characterize the degree of stratigraphic fragmentation and the spatial distribution law of structural elements, establish the structural framework of the study area, and provide geometric basis for the division of water-controlling units. S1.2 Water Control Unit Division: Based on the tectonic mechanical properties, fracture development degree and stratigraphic fracture characteristics, combined with the fault water conductivity, fold axis fracture enrichment degree and tectonic composite water-holding conditions in the three-dimensional tectonic model, the water control level of the study area is evaluated, and the fault water-conducting zone, fold axis fracture enrichment zone, tectonic composite water-holding zone and intact bedrock water-retaining zone are divided to form a spatial framework of water control units.

[0029] In this embodiment: In step S1.1, lithological stratification and well inclination data from coalfield exploration boreholes, fault and fold wavefield response data from 3D seismic exploration, and structural attitude and stratigraphic contact relationship data from mining engineering geological profiles are collected. Coordinate system transformation and format standardization are performed on the above multi-source heterogeneous data to eliminate spatial reference differences. Based on the processed data, fault attitude, fold morphology, and structural intersection relationships are extracted to construct a 3D structural geological model. This model quantitatively characterizes the degree of stratigraphic fragmentation and the spatial distribution of structural elements, establishing a structural framework for the study area and providing a geometric basis for the division of water-controlling units. By integrating borehole point data, seismic volume response data, and mining profile line data into the same 3D spatial framework, precise positioning and quantitative characterization of structural elements are achieved, solving the problem of inconsistent spatial references between structural models and hydrological models in existing technologies, and improving the precision of the spatial characterization of structural water-controlling effects in complex structural areas.

[0030] In step S1.2, using a three-dimensional structural geological model as spatial constraint, and based on structural mechanical properties, fracture development degree, and stratigraphic fragmentation characteristics, combined with the model's fault hydraulic conductivity, fracture enrichment degree in fold axis regions, and structural perched conditions, a water control level evaluation criterion is established. Water control units are divided into four types: fault hydraulic zones, fold axis fracture enrichment zones, structural perched zones, and bedrock aquitard zones, forming a spatial framework for water control units. This division process transforms structural geometric parameters into hydrological boundary conditions, establishing a structural water control level evaluation system based on differences in mechanical properties and fracture development degree in different structural locations, achieving zonal characterization of aquifer spatial structure and hydrological attributes. This allows subsequent in-situ pressure water tests and fracture seepage simulations to be conducted in a targeted manner under differentiated structural backgrounds, avoiding the shortcomings of existing technologies that treat aquifers as continuous homogeneous media and ignore structural water control effects.

[0031] Compared to existing technologies where structural and hydrological models are constructed independently with disconnected spatial benchmarks, and aquifers are often simplified as continuous homogeneous media, this step achieves precise mapping of structural geometry to hydrological boundaries through multi-source geological data fusion interpretation and targeted water-controlling unit delineation. This process uses drilling lithology data, seismic wavefield data, and mining exposure data as inputs. After coordinate unification, structural element extraction, and 3D modeling, a structural framework including fault attitude, fold morphology, and structural intersection relationships is established. Then, based on structural mechanical properties, fracture development degree, and stratigraphic fragmentation characteristics, combined with water-controlling attributes such as fault conductivity and fold fracture enrichment, differentiated water-controlling units are delineated. This progressive processing flow allows subsequent steps such as in-situ pressure water testing and fracture seepage simulation to be carried out within a clearly defined structural zoning framework, ensuring precise correspondence between test locations and key structural components. This improves the targeting and spatial resolution of hydrogeological parameter detection in complex structural coalfields, laying a spatial foundation for quantitative calibration and multi-source fusion calibration of structural water-controlling effects.

[0032] Example 3: Please refer to Figure 1 In step S2, the specific method is as follows: S2.1 In-situ multi-parameter testing: In-situ pressure water tests are carried out in each water control unit to obtain the permeability coefficient and hydraulic conductivity of the rock mass. Simultaneously, triaxial mechanical tests and acoustic emission monitoring of the tectonic rock mass are conducted to analyze the evolution law of fracture aperture and the characteristics of abrupt change in hydraulic conductivity under different stress states. Physical response parameters of tectonic water control effect are obtained and water control level thresholds are calibrated. S2.2 Fracture Network Simulation: Based on core scanning imaging and 3D reconstruction, a physical model of the fracture network is established, seepage simulation is carried out, and the quantitative relationship between fracture aperture, connectivity and permeability is obtained. By combining the results of in-situ testing and simulation, a physical mapping model between structural geometric parameters and hydrological response parameters is established, and a set of quantitative indicators of structural water control effect is generated.

[0033] In this embodiment: In step S2.1, representative test sections are selected according to structural locations within each water control unit, and segmented pressure water tests are conducted to collect dynamic response data of pressure and flow. The permeability coefficient and hydraulic conductivity of the rock mass are calculated according to the pressure water test specifications. Simultaneously, standard samples of the structural rock mass are prepared, and triaxial mechanical tests are performed. Confining pressure and axial pressure are applied, and elastic wave signals during the initiation, propagation, and penetration of fractures are collected in real time using an acoustic emission system. Ringing counts and energy analysis are performed on the acoustic emission signals to locate the fracture evolution sequence. Combined with the stress-strain curves obtained from the triaxial mechanical tests, the evolution law of fracture aperture and the abrupt change characteristics of hydraulic conductivity under different stress states are analyzed. Based on the above test results, physical response parameters of the structural water control effect are obtained, and the water control level threshold is calibrated. This process establishes a quantitative relationship between the actual hydraulic conductivity of the rock mass and the stress-fracture coupling response, providing hard constraint data for the subsequent construction of a physical mapping model, and solving the problem of the lack of in-situ physical basis for the quantification of structural water control effects in existing technologies.

[0034] In step S2.2, representative rock cores from the water-controlling unit are collected and scanned to extract geometric parameters such as fracture aperture, trace length, and density, as well as spatial connectivity characteristics. A discrete fracture network physical model is established, and numerical seepage simulation is conducted to solve the fluid transport laws within the fracture network and analyze the quantitative response relationship between fracture aperture, connectivity, and equivalent permeability coefficient. The rock mass permeability and hydraulic conductivity measured by in-situ pressure water tests are used as model calibration constraints to invert and correct the fracture network model parameters. By combining the results of in-situ tests and simulations, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, generating a quantitative index set of structural water control effects. This process realizes the physical quantitative conversion of structural geometry into hydrological parameters, providing objective physical constraints for fusion weight allocation.

[0035] Step S2, through steps S2.1 and S2.2, achieves a leap from empirical qualitative description to physical quantitative characterization of the structural water control effect. Existing technologies mostly rely on expert scoring, semi-quantitative grading, and simple geometric statistics to establish quantitative indicators of structural water control effects, lacking objective calibration through in-situ pressure water tests, fracture seepage simulations, and rock mechanics tests. When the same structural part exhibits different water control levels due to differences in observation scale, infill material properties, and stress states, existing methods struggle to accurately distinguish them. Step S2.1 conducts in-situ pressure water tests and triaxial mechanical tests within each water control unit to obtain the true water conductivity of the rock mass and stress-fracture coupling response characteristics. Step S2.2 establishes a fracture network model based on core scanning imaging and conducts seepage simulations to establish a quantitative relationship between geometric topology and seepage capacity. Combined with the above test results, a physical mapping model between structural geometric parameters and hydrological response parameters is established. This progressive processing flow elevates the construction of water control effects from empirical qualitative judgment to physical quantitative driving, enabling quantitative indicators to obtain in-situ physical basis, eliminating subjective uncertainty of experts, and providing a reliable physical constraint benchmark for subsequent multi-source fusion weight allocation.

[0036] Example 4: Please refer to Figure 1 In step S2.1, the specific method for in-situ multi-parameter testing is as follows: S2.11 In-situ pressure water test: Select representative test sections according to structural parts within each water control unit, conduct segmented pressure water tests, determine the permeability coefficient and hydraulic conductivity of the rock mass, analyze the characteristics of the pressure-flow dynamic response curve, extract the true hydraulic conductivity parameters of the rock mass in different structural parts, and integrate them to form a basic hydrological response dataset for the water control unit. S2.12 Mechanical Coupling Monitoring: Simultaneously conduct triaxial mechanical tests on the tectonic rock mass, apply gradient confining pressure and axial pressure loads, use acoustic emission to capture the evolution sequence of fracture initiation, propagation and penetration in real time, analyze the evolution law of fracture aperture and the characteristics of abrupt changes in water conductivity under different stress states, calibrate the threshold of water control level, and establish the coupling discrimination relationship between stress and fracture water conductivity.

[0037] In this embodiment: In step S2.1.1, within each water-controlling unit, test sections representing the structural attributes of the unit are selected according to the spatial distribution of fault water-conducting zones, fold axis fracture-rich zones, structural composite perched water zones, and intact bedrock water-resistant zones. A segmented pressurization method is used, injecting pressurized water into the test sections through pressurized water testing equipment, and collecting real-time dynamic response data of pressure and flow. The pressure-flow curves are analyzed according to the pressurized water testing specifications to calculate the rock mass permeability and hydraulic conductivity, extracting the true hydraulic conductivity parameters of the rock mass in different structural locations. The permeability, hydraulic conductivity, and pressure-flow response characteristics obtained from each test section are integrated to form a basic hydrological response dataset covering each water-controlling unit. This process directly obtains the in-situ hydraulic conductivity of the rock mass, avoiding parameter distortion caused by stress release in laboratory tests, and providing a true hydrological parameter benchmark for quantifying the structural water-controlling effect.

[0038] In step S2.1.2, while conducting in-situ pressure water tests, representative rock samples were collected from the water-controlling unit, and standard triaxial specimens were prepared. Gradient confining pressure and axial pressure loads were applied using a triaxial mechanical testing machine to simulate different burial depths and tectonic stress environments. Simultaneously, an acoustic emission sensor array was deployed to capture elastic wave signals during the initiation, propagation, and penetration of fractures in real time. Ringing counts and energy analysis were performed on the acoustic emission signals to determine the fracture evolution sequence. Combined with stress-strain curves, the evolution law of fracture aperture and the characteristics of abrupt changes in hydraulic conductivity under different stress states were analyzed. Based on the test results, water control level thresholds corresponding to different stress levels were calibrated, establishing a coupling discriminant relationship between stress state and fracture hydraulic conductivity. This process reveals the control mechanism of tectonic stress on fracture hydraulic conductivity, providing a physical basis for considering tectonic stress state in subsequent weight allocation.

[0039] Step S2.1, through steps S2.1.1 and S2.1.2, achieves the comprehensive acquisition of in-situ physical response parameters and stress coupling mechanisms. Existing technologies for quantifying structural water control effects largely rely on expert experience scoring or semi-quantitative grading, lacking support from in-situ physical testing data. This makes it difficult to accurately distinguish different water control levels in the same structural location due to differences in observation scale and stress state. Step S2.1.1 conducts segmented pressure water tests within each water control unit to directly measure the rock mass permeability and hydraulic conductivity, forming a basic hydrological response dataset. Step S2.1.2, through triaxial mechanical testing and acoustic emission monitoring, reveals the evolution of fracture aperture and abrupt changes in hydraulic conductivity under different stress states, establishing a discriminant relationship between stress and fracture hydraulic conductivity coupling. These two methods synergistically transform the structural water control effect from empirical inference into a quantitative characterization based on in-situ testing and mechanical monitoring, eliminating subjective uncertainties, providing hard constraint data for the construction of physical mapping models, and improving the objectivity and reliability of hydrological parameter calibration in complex structural areas.

[0040] Example 5: Please refer to Figure 1In step S2.2, the specific method for simulating the fracture network is as follows; S2.21, Fracture Network Reconstruction: Based on core computer tomography and 3D reconstruction technology, the aperture, trace length, density and spatial connectivity characteristics of fractures inside the core are extracted, a discrete fracture network physical model is established, and the topological structure, geometric connectivity characteristics and seepage channel development degree of the fracture network are quantitatively characterized. S2.22, Seepage Law Mapping: Numerical seepage simulation is carried out based on a discrete fracture network model to analyze the quantitative response laws of fracture aperture, connectivity and equivalent permeability coefficient. By combining the results of in-situ pressure water test and seepage simulation, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, and a set of quantitative indicators of structural water control effect is generated.

[0041] In this embodiment: In step S2.2.1, representative rock cores are collected from each water control unit, and images of the spatial distribution of fractures inside the rock cores are obtained using computed tomography (CT) technology. Image processing is used to extract fracture aperture, trace length, density, and spatial connectivity features to establish a discrete fracture network physical model. This model quantitatively characterizes the fracture network topology, geometric connectivity features, and the degree of seepage channel development, transforming hidden fractures within the rock mass into a computable geometric topology, providing a geometric constraint basis for seepage simulation. This process completes the three-dimensional digital reconstruction of the fracture structure inside the rock core, realizing the transformation of fracture geometric parameters from qualitative description to quantitative characterization, overcoming the shortcomings of existing technologies that rely on borehole statistics and neglect spatial connectivity in fracture network characterization.

[0042] In step S2.2.2, numerical seepage simulation is conducted based on a discrete fracture network physical model. Fluid boundary conditions and seepage parameters are set, and the fluid transport law within the fracture network is solved. The quantitative response relationship between fracture aperture, connectivity, and equivalent permeability coefficient is analyzed. The equivalent permeability coefficient of the fracture network model is calibrated and corrected using the rock mass permeability and hydraulic conductivity measured by in-situ pressure water tests as hard constraints. By integrating the results of in-situ tests and seepage simulation, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, generating a quantitative index set of structural water control effects. This process completes the coupling mapping between fracture geometry and hydrological parameters, realizing the physical quantitative transformation from structural geometry to hydrological response, and providing an objective physical basis for weight allocation.

[0043] Step S2.2, through steps S2.2.1 and S2.2.2, elevates the tectonic water control effect from empirical qualitative inference to physical quantitative driving. Existing technologies for quantifying tectonic water control effects largely rely on expert experience scoring or semi-quantitative grading, lacking objective calibration based on real fracture network seepage simulations. This results in a lack of physical basis for the mapping relationship between tectonic geometric parameters and hydrological response parameters. Step S2.2.1, based on core computed tomography and 3D reconstruction technology, extracts fracture aperture, trace length, density, and spatial connectivity characteristics to establish a discrete fracture network physical model, transforming the hidden fracture structure within the rock mass into a computable geometric topology. Step S2.2.2, based on this model, conducts numerical seepage simulations to analyze the quantitative response laws of fracture aperture, connectivity, and equivalent permeability coefficient. Calibration is performed using in-situ pressure water test results as constraints, establishing a physical mapping relationship between tectonic geometric parameters and hydrological response parameters. The synergy of these two steps provides objective physical basis for the quantitative indicators, eliminating subjective uncertainty and providing a physical constraint benchmark for subsequent multi-source fusion weight allocation.

[0044] Example 6: Please refer to Figure 1 In step S3, the specific method for data decoupling and adaptation is as follows: S3.1 Multi-source error decoupling: Analyze the physical field response mechanism and applicable conditions of transient electromagnetic, borehole television and hydrological monitoring in structurally complex areas, identify the sources of systematic errors and detection blind spots of each detection method in different water control units, establish a regional differentiated error correction model, decouple and map the geophysical volume response data to the corresponding water control unit, and eliminate the multi-source detection system deviation. S3.2 Scale Adaptation and Conversion: Based on the aforementioned physical mapping model, the borehole-scale fracture network parameters are extended and converted to the structural unit volume scale to establish a cross-scale bridging relationship between borehole point measurements and geophysical volume response, thereby achieving accurate adaptation and spatial consistency registration of multi-source heterogeneous data within the structural spatial framework.

[0045] In this embodiment, step S3.1 analyzes the physical field response mechanism and applicable conditions of three detection methods—transient electromagnetic, borehole television, and hydrological monitoring—in structurally complex areas. Combining the geological characteristics of fault-conducting zones, fracture-rich zones in fold axes, and tectonic perched areas, the sources of systematic errors and detection blind spots for each detection method within different water-controlling units are identified. Based on this, a zone-specific error correction model is established, decoupling and mapping the geophysical volumetric response data to the corresponding water-controlling units, eliminating multi-source detection system bias. This process achieves error identification and zone-specific correction of multi-source data in structurally complex areas, decoupling and normalizing the physical responses of different detection methods. It overcomes the shortcomings of existing technologies that directly fuse multi-source data without considering the differences in methodological principles and systematic biases in structurally complex areas, thus improving the basic quality of multi-source data fusion in key structural locations.

[0046] In step S3.2, based on the aforementioned physical mapping model, the borehole-scale fracture network parameters are extended and transformed to the structural unit volume scale. A cross-scale bridging relationship is established between borehole point measurements and geophysical volumetric responses, achieving accurate adaptation and spatial consistency registration of multi-source heterogeneous data within the structural spatial framework. This process completes the conversion and bridging between parameters at different observation scales, overcoming the scale effect distortion problem caused by directly splicing borehole-scale and geophysical-scale data in existing technologies. It ensures the spatial consistency of multi-source data within the structural spatial framework, providing a scale-unified data foundation for subsequent fusion weight optimization.

[0047] In existing technologies, multi-source detection data in structurally complex regions are typically directly spatially registered and correlated without fully considering the differences in the physical field response mechanisms, applicable boundaries, and detection blind zones of various detection methods. Furthermore, no scale bridging relationship is established between borehole point measurements and geophysical volume responses, resulting in systemic bias and scale mismatch in multi-source data fusion. Step S3, through steps S3.1 and S3.2, achieves decoupling of the physical responses of multi-source data and precise cross-scale adaptation. Step S3.1 analyzes the physical field response mechanisms of transient electromagnetic, borehole television, and hydrological monitoring, identifies the sources of systemic errors and detection blind zones within different water control units, establishes a zone-specific error correction model, and decouples and maps geophysical volume response data to the corresponding water control units, eliminating multi-source detection system bias. Step S3.2, based on the physical mapping model, extends and transforms borehole-scale fracture network parameters to the structural unit volume scale, establishing a cross-scale bridging relationship between borehole point measurements and geophysical volume responses, achieving precise adaptation and spatially consistent registration of multi-source heterogeneous data. The two work together to upgrade multi-source data from simple spatial superposition to deep fusion after physical response decoupling and scale adaptation, eliminating system errors and scale effects in multi-source data fusion in complex areas, and improving the spatial consistency and reliability of hydrological parameter fusion calibration.

[0048] Example 7: Please refer to Figure 1 In step S4, the specific method for dynamic weight optimization is as follows: S4.1 Credibility Calibration: Using the results of in-situ pressure water test as hard constraints and the data of fracture seepage simulation and rock mechanics test as soft constraints, a multi-source data credibility evaluation function is constructed. Based on the physical response accuracy and data quality of each detection method in different water control units, the confidence level of each data source is calibrated, a credibility benchmark system is established, and calibration parameters are generated. S4.2 Dynamic Weight Optimization: An adaptive weight allocation mechanism is established. In the structural composite parts, the contribution weights of each data source are dynamically adjusted and redistributed based on the differences in structural stress state and crack filling characteristics. This achieves adaptive optimization and imbalance correction of the fusion weights of key structural parts, forming a multi-source optimized fusion weight configuration scheme and outputting the fusion weights.

[0049] In this embodiment: In step S4.1, the permeability and hydraulic conductivity measured by in-situ pressure water tests are used as hard constraints, while the equivalent permeability obtained from fracture seepage simulation and the fracture hydraulic conductivity characteristics extracted from rock mechanics tests are used as soft constraints to construct a multi-source data credibility evaluation function. Based on the physical response accuracy and data quality of each detection method in different water control units, the confidence levels of each data source—transient electromagnetic, borehole television, and hydrological monitoring—are calibrated, thereby establishing a credibility benchmark system and generating calibration parameters. This process embeds the physical calibration results into the credibility evaluation stage, replacing expert experience scoring with objective physical data, eliminating the subjective uncertainty in basic probability allocation, and thus providing a reliable physical constraint benchmark for adaptive weight allocation.

[0050] In step S4.2, an adaptive weight allocation mechanism is established. In structurally complex regions, the contribution weights of each data source are dynamically adjusted and redistributed based on the differences in tectonic stress state and fracture filling characteristics. By analyzing the abrupt changes in response accuracy of different detection methods in structurally complex regions, the imbalance in the initial weight allocation is corrected, achieving adaptive optimization of the fusion weights for key structural regions, and ultimately outputting a multi-source optimized fusion weight allocation scheme. This process solves the problem of unbalanced data source weight allocation caused by differences in stress state and filling characteristics in structurally complex regions, ensuring that the fusion results accurately reflect the hydrogeological conditions of key structural regions.

[0051] In existing technologies, multi-source fusion weight allocation often employs DS evidence theory or entropy weight method, establishing basic probability allocations based on expert experience scoring. This lack of objective calibration using in-situ physical test data leads to human uncertainty in the weight coefficients of structurally complex parts, causing distortion in the fusion results. Step S4, through steps S4.1 and S4.2, transforms the fusion weights from subjective experience-based assignment to adaptive optimization under physical constraints. Step S4.1 uses in-situ pressure water test results as hard constraints and fracture seepage simulation and rock mechanics test data as soft constraints to construct a credibility evaluation function, calibrate the confidence level of each data source, and establish a physical constraint benchmark. Step S4.2 establishes an adaptive weight allocation mechanism based on this benchmark, dynamically adjusting weights at structurally complex parts by considering differences in structural stress state and fracture filling characteristics to achieve imbalance correction. The two mechanisms work together to integrate objective physical calibration results throughout the entire fusion weight allocation process, eliminating human uncertainty, resolving the imbalance problem in weight allocation at structurally complex parts, and improving the objectivity and reliability of the fusion results of hydrological parameters in complex structural coalfields.

[0052] Example 8: Please refer to Figure 1 In step S4.1, the specific method for confidence level determination is as follows: S4.11, Constraint Benchmark Construction: Using the in-situ pressure water test permeability coefficient and hydraulic conductivity as hard constraints, and the equivalent permeability coefficient of fracture seepage simulation and the fracture hydraulic conductivity characteristics of rock mechanics test as soft constraints, a credibility evaluation function is constructed, the water control unit type and the tectonic stress state correction coefficient are coupled, the response accuracy and data quality differentiation law of each detection method are analyzed, and evaluation indicators and physical constraint benchmarks are established. S4.12 Confidence Calibration: Based on the response accuracy, data quality and spatial coverage of each detection method in different water control units, and combined with the differences in tectonic stress state and fracture filling characteristics, the confidence levels of each data source of transient electromagnetic, borehole television and hydrological monitoring are graded and calibrated, multi-scale quantitative calibration parameters are generated, a confidence benchmark system is established and calibration results are formed.

[0053] In this embodiment: In step S4.11, the permeability and hydraulic conductivity measured by in-situ pressure water tests are used as hard constraints, while the equivalent permeability obtained from fracture seepage simulation and the fracture hydraulic conductivity characteristics extracted from rock mechanics tests are used as soft constraints to construct a credibility evaluation function. The water control unit type and tectonic stress state correction coefficient are coupled to analyze the response accuracy and data quality differentiation patterns of each detection method in different water control units, establishing evaluation indicators and physical constraint benchmarks. This process transforms the in-situ physical test results into quantitative constraints for credibility evaluation, replacing expert experience scoring with objective physical data, eliminating subjective uncertainties in basic probability allocation, and providing a reliable physical benchmark for subsequent confidence level grading.

[0054] In step S4.12, based on the response accuracy, data quality, and spatial coverage of each detection method within different water control units, and considering the differences in tectonic stress state and fracture filling characteristics, the confidence levels of each data source—transient electromagnetic, borehole television, and hydrological monitoring—are graded and calibrated. Multi-scale quantitative calibration parameters are generated, a confidence benchmark system is established, and calibration results are formed. This process enables differentiated confidence assessment of each data source under different tectonic backgrounds, ensuring that the confidence level truly reflects the physical response accuracy and data quality of the detection methods. It avoids the weight allocation bias caused by uniform confidence setting in traditional methods, providing a refined grading basis for adaptive weight optimization.

[0055] In existing technologies, multi-source fusion confidence calibration often relies on expert experience scoring or simple geometric statistics, lacking objective constraints from in-situ physical testing and fracture seepage simulation, leading to subjective uncertainty in the confidence levels of each data source. Step S4.1, through steps S4.11 and S4.12, improves confidence calibration from subjective experience-based inference to graded quantification under physical constraints. Step S4.11 uses in-situ pressure water test results as hard constraints and fracture seepage simulation and rock mechanics test data as soft constraints to construct a confidence evaluation function and establish a physical constraint benchmark. Step S4.12, based on the response accuracy and data quality of each detection method in different water control units, combined with the differences in tectonic stress state and fracture filling characteristics, gradedly calibrates the confidence levels of each data source and generates quantitative calibration parameters. The two work together to integrate objective physical calibration results throughout the entire credibility evaluation process, eliminate human uncertainty, establish a refined credibility benchmark system, provide a reliable physical basis for the adaptive optimization and imbalance correction of the fusion weights of subsequent structural composite parts, and improve the objectivity and reliability of the fusion results of multi-source hydrological parameters in complex structural coalfields.

[0056] Example 9: Please refer to Figure 1 In step S4.2, the specific method for dynamic weight optimization is as follows: S4.21 Weight Initialization: Based on the confidence calibration results and the physical constraint benchmark, an adaptive weight allocation mechanism is established. According to the confidence level and data quality differences of each detection method in different water control units, the initial contribution weight of each data source is calculated to form the initial fusion weight configuration. S4.22 Dynamic Imbalance Correction: In structural composite parts, combining the differences in structural stress state and crack filling characteristics, the sudden change law of confidence of each data source is analyzed, the initial contribution weight is dynamically adjusted and redistributed, the weight imbalance of key structural parts is corrected, and adaptive optimized fusion weights are generated and output.

[0057] In this embodiment: In step S4.21, based on the confidence calibration results and using the physical constraint benchmark as the allocation basis, an adaptive weight allocation mechanism is established. According to the confidence level and data quality differences of each detection method in different water control units, the initial contribution weights of each data source—transient electromagnetic, borehole television, and hydrological monitoring—are calculated to form the initial fusion weight configuration. This process transforms the physical constraint benchmark into a quantifiable weight allocation scheme, enabling each data source to receive differentiated weight assignments based on its objective physical response accuracy. This avoids the fusion bias caused by average weight allocation or subjective setting in traditional methods, providing a reasonable initial configuration basis for subsequent dynamic imbalance correction.

[0058] In step S4.22, at structurally complex locations, the abrupt change patterns of confidence levels from various data sources are analyzed by considering the differences in tectonic stress state and fracture filling characteristics. The initial contribution weights are dynamically adjusted and redistributed to correct weight imbalances in key structural locations, generating and outputting adaptive optimized fusion weights. This process achieves dynamic rebalancing of weights in structurally complex locations, enabling the fusion results to adapt to spatial variations in tectonic stress state and fracture filling characteristics. It overcomes the shortcomings of existing technologies where fixed weight configurations cannot respond to local structural anomalies, ensuring that the fusion weights for key structural locations match actual hydrogeological conditions.

[0059] In existing technologies, multi-source fusion weight allocation often employs fixed weights or statistical methods based on entropy weighting, failing to fully consider the abrupt changes in data source response accuracy caused by differences in stress state and fracture filling characteristics in structurally complex regions. This leads to imbalanced weight allocation in key structural components and distorted fusion results. Step S4.2, through steps S4.21 and S4.22, transforms the fusion weights from static configuration to dynamic adaptive optimization. Step S4.21 establishes an adaptive weight allocation mechanism based on the confidence level calibration results, calculating initial contribution weights according to the confidence level and data quality differences of each detection method within different water control units, forming a reasonable initial configuration. Step S4.22 analyzes the abrupt changes in confidence level in structurally complex regions by considering the differences in structural stress state and fracture filling characteristics, dynamically adjusting and redistributing the initial weights to correct weight imbalances. The synergy of these two steps enables the fusion weights to respond to variations in local hydrogeological conditions within the structure, eliminating the problem of unbalanced weight allocation in key structural components and improving the objectivity and reliability of hydrological parameter fusion results in complex structural coalfields.

[0060] Example 10: Please refer to Figure 1 In step S5, the specific method for closed-loop iterative correction is as follows: S5.1 Numerical simulation verification: Input the fused and calibrated hydrological parameters into the groundwater flow numerical simulation model, perform multi-objective fitting verification with measured water level, inflow and water quality data, establish an uncertainty evaluation system, calculate the confidence interval and sensitivity coefficient of hydrological parameters of each water control unit, analyze the source of verification deviation and extract deviation feature information; S5.2 Closed-loop iterative correction: If the verification error exceeds the preset threshold, the deviation information is fed back to step S2 to correct the physical mapping model parameters, and then fed back to step S4 to adjust the weight allocation strategy, forming a closed-loop iterative correction process of physical calibration, fusion optimization and verification feedback, so as to realize the continuous dynamic calibration of hydrological parameters.

[0061] In this embodiment: In step S5.1, the fused and calibrated hydrological parameters are input into the groundwater flow numerical simulation model. This model uses the finite difference method to solve the three-dimensional unsteady groundwater flow equation. Spatial discretization uses an unstructured grid, and the grid density is adaptively adjusted according to the scale of the water-controlling unit. The grid is refined in the fault-conducting zone and the fracture-rich area of ​​the fold axis, and sparse in the bedrock impermeable zone. Boundary conditions are set: the upper boundary is the water level boundary, the lower boundary is the impermeable bottom boundary, and the lateral boundary is the flow boundary. The measured water level, inflow, and water quality data are used as validation samples to construct a multi-objective fitting function. The function form is that the fitting error is equal to the weighted sum of the squared terms of the water level error, the squared terms of the inflow error, and the squared terms of the water quality concentration error. The weight coefficients of each term are determined according to the monitoring accuracy and sampling frequency of the corresponding data. The Monte Carlo method is used to evaluate the uncertainty. A random sample set is generated with the fused parameters as the mean and the physical calibration error as the variance. The numerical model is run repeatedly, the distribution interval of the output results of each water-controlling unit is statistically analyzed, and the confidence interval is calculated. Local sensitivity analysis is employed to calculate the sensitivity coefficient by changing the parameters of a specific water control unit once and observing the rate of change of the objective function. The sources of deviation are analyzed and verified, and deviation characteristic information is extracted to form a spatial distribution map of the parameter deviations for each water control unit. This process verifies the accuracy of the fusion results and quantifies the uncertainty. Numerical simulation is used to verify the engineering applicability of the fusion parameters, identify the spatial distribution patterns of parameter deviations, and provide a basis for subsequent iterative corrections. This overcomes the shortcomings of existing technologies that only perform simple comparisons of fusion results and lack systematic uncertainty evaluation.

[0062] In step S5.2, a preset threshold for verification error is set. When the root mean square error of the multi-objective fitting exceeds this threshold, the fusion result is determined to require iterative correction. Deviation information is fed back through a dual-channel feedback mechanism. The first channel feeds back the deviation information to step S2 to correct the equivalent permeability coefficient of the fracture network, the stress-fracture coupling relationship parameters, and the scale conversion factor in the physical mapping model. The second channel feeds back the deviation information to step S4 to adjust the structural stress state correction coefficient, fracture filling characteristic correction coefficient, and weight dynamic adjustment factor in the credibility evaluation function. During the iteration process, the quantitative index set of structural water control effect is updated with the corrected physical mapping model parameters, the fusion weights are recalculated using the adjusted weight allocation strategy, and fusion calibration and numerical simulation verification are performed again, forming a closed-loop iterative correction process of physical calibration, fusion optimization, and verification feedback. The iteration termination condition is set as either the verification error falling below the preset threshold or the number of iterations reaching the upper limit. This process establishes a reverse propagation channel from verification bias to model correction, enabling the physical mapping model and weight allocation strategy to adaptively adjust based on the verification results. This avoids the defects in existing technologies, such as the disconnect between fusion results and model parameters and unclear correction directions, ensuring the reliability of the hydrogeological model during long-term mining operations.

[0063] Step S5, through steps S5.1 and S5.2, transforms the fusion calibration results from static output to dynamic continuous optimization. In existing technologies, multi-source fusion calibration results are typically output once and then not corrected, lacking verification feedback and iterative optimization mechanisms. This causes the model to gradually deviate from actual hydrological conditions due to mining disturbances. Step S5.1 inputs the fusion parameters into the numerical simulation model for multi-objective fitting verification, uses the Monte Carlo method to calculate confidence intervals, and employs local sensitivity analysis to calculate sensitivity coefficients, quantifying uncertainty and extracting deviation information. Step S5.2 establishes a dual-channel feedback mechanism for deviation information to the physical mapping model and weight allocation strategy. Iteration is triggered based on a preset threshold to correct the physical mapping model parameters and weight allocation strategy, forming a closed-loop iterative correction process. The synergy of these two mechanisms enables continuous calibration of hydrological parameters based on measured data, eliminating accumulated model errors, improving the long-term reliability and dynamic adaptability of hydrogeological models for complex coalfield structures, and providing continuous and effective technical support for mine water hazard prevention and control.

[0064] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A complex structure coalfield hydrological parameter multi-source fusion calibration method, characterized by: The specific steps are as follows: S1. Detailed structural analysis: A three-dimensional structural model is constructed based on coalfield exploration data, and water-controlling units are divided according to structural mechanical properties and fracture characteristics; S2. Water control effect calibration: In-situ pressure water tests and fissure seepage simulations are carried out in each water control unit to establish a physical mapping between structural parameters and hydrological response; Specific methods include: S2.1, In-situ multi-parameter testing: In-situ pressure water tests are conducted in each water-controlling unit to obtain the rock mass permeability and hydraulic conductivity. Simultaneously, triaxial mechanical tests and acoustic emission monitoring of the tectonic rock mass are carried out to analyze the evolution law of fracture aperture and the characteristics of abrupt changes in hydraulic conductivity under different stress states. Physical response parameters of tectonic water control effect are obtained and water control level thresholds are calibrated; S2.2, Fracture network simulation: A physical model of fracture network is established based on core scanning imaging and three-dimensional reconstruction. Seepage simulation is carried out to obtain the quantitative relationship between fracture aperture, connectivity and permeability. By combining the results of in-situ testing and simulation, a physical mapping model of tectonic geometric parameters and hydrological response parameters is established to generate a set of quantitative indicators of tectonic water control effect. S3. Data decoupling and adaptation: Establish an error correction model for multi-source detection within the water control unit, convert borehole-scale parameters to unit-scale parameters, and achieve multi-source data decoupling; S4. Dynamic Weight Optimization: A credibility function is constructed based on physical calibration results as constraints. An adaptive weighting mechanism is established based on response accuracy, and the weights of data sources are adjusted in structural composite parts. The specific methods of dynamic weight optimization are as follows: S4.

1. Credibility Calibration: Using in-situ pressure water test results as hard constraints and fracture seepage simulation and rock mechanics test data as soft constraints, a multi-source data credibility evaluation function is constructed. The credibility level of each data source is calibrated based on the physical response accuracy and data quality of each detection method in different water control units, a credibility benchmark system is established, and calibration parameters are generated. S4.

2. Dynamic Weight Optimization: An adaptive weight allocation mechanism is established. In structural composite parts, the contribution weights of each data source are dynamically adjusted and redistributed based on the differences in structural stress state and fracture filling characteristics. This achieves adaptive optimization and imbalance correction of the fusion weights in key structural parts, forming a multi-source optimized fusion weight configuration scheme and outputting the fusion weights. S5. Closed-loop iterative correction: Input the fused and calibrated hydrological parameters into the numerical simulation model for verification. When the error exceeds the threshold, it is fed back to step S2 and step S4 for iterative correction.

2. The complex structure coalfield hydrological parameter multi-source fusion calibration method according to claim 1, characterized in that, In step S1, the specific method for constructing a refined parsing is as follows: S1.1 Structural geometric modeling: Based on coalfield drilling, 3D seismic and mining geological data, fault attitude, fold morphology and structural intersection relationship are extracted to construct a 3D structural geological model, quantitatively characterize the degree of stratigraphic fragmentation and the spatial distribution law of structural elements, establish the structural framework of the study area, and provide geometric basis for the division of water-controlling units. S1.2 Water Control Unit Division: Based on the tectonic mechanical properties, fracture development degree and stratigraphic fracture characteristics, combined with the fault water conductivity, fold axis fracture enrichment degree and tectonic composite water-holding conditions in the three-dimensional tectonic model, the water control level of the study area is evaluated, and the fault water-conducting zone, fold axis fracture enrichment zone, tectonic composite water-holding zone and intact bedrock water-retaining zone are divided to form a spatial framework of water control units.

3. The multi-source fusion calibration method for hydrological parameters in complex coalfields according to claim 1, characterized in that, In step S2.1, the specific method for in-situ multi-parameter testing is as follows: S2.11 In-situ pressure water test: Select representative test sections according to structural parts within each water control unit, conduct segmented pressure water tests, determine the permeability coefficient and hydraulic conductivity of the rock mass, analyze the characteristics of the pressure-flow dynamic response curve, extract the true hydraulic conductivity parameters of the rock mass in different structural parts, and integrate them to form a basic hydrological response dataset for the water control unit. S2.12 Mechanical Coupling Monitoring: Simultaneously conduct triaxial mechanical tests on the tectonic rock mass, apply gradient confining pressure and axial pressure loads, use acoustic emission to capture the evolution sequence of fracture initiation, propagation and penetration in real time, analyze the evolution law of fracture aperture and the characteristics of abrupt changes in water conductivity under different stress states, calibrate the threshold of water control level, and establish the coupling discrimination relationship between stress and fracture water conductivity.

4. The multi-source fusion calibration method for hydrological parameters in complex coalfields according to claim 3, characterized in that, In step S2.2, the specific method for simulating the fracture network is as follows; S2.21, Fracture Network Reconstruction: Based on core computer tomography and 3D reconstruction technology, the aperture, trace length, density and spatial connectivity characteristics of fractures inside the core are extracted, a discrete fracture network physical model is established, and the topological structure, geometric connectivity characteristics and seepage channel development degree of the fracture network are quantitatively characterized. S2.22, Seepage Law Mapping: Numerical seepage simulation is carried out based on a discrete fracture network model to analyze the quantitative response laws of fracture aperture, connectivity and equivalent permeability coefficient. By combining the results of in-situ pressure water test and seepage simulation, a physical mapping relationship between structural geometric parameters and hydrological response parameters is established, and a set of quantitative indicators of structural water control effect is generated.

5. The multi-source fusion calibration method for hydrological parameters in complex coalfields according to claim 4, characterized in that, In step S3, the specific method for data decoupling and adaptation is as follows: S3.1 Multi-source error decoupling: Analyze the physical field response mechanism and applicable conditions of transient electromagnetic, borehole television and hydrological monitoring in structurally complex areas, identify the sources of systematic errors and detection blind spots of each detection method in different water control units, establish a regional differentiated error correction model, decouple and map the geophysical volume response data to the corresponding water control unit, and eliminate the multi-source detection system deviation. S3.2 Scale Adaptation and Conversion: Based on the aforementioned physical mapping model, the borehole-scale fracture network parameters are extended and converted to the structural unit volume scale to establish a cross-scale bridging relationship between borehole point measurements and geophysical volume response, thereby achieving accurate adaptation and spatial consistency registration of multi-source heterogeneous data within the structural spatial framework.

6. The multi-source fusion calibration method for hydrological parameters in complex coalfields according to claim 5, characterized in that, In step S4.1, the specific method for confidence level determination is as follows: S4.11, Constraint Benchmark Construction: Using the in-situ pressure water test permeability coefficient and hydraulic conductivity as hard constraints, and the equivalent permeability coefficient of fracture seepage simulation and the fracture hydraulic conductivity characteristics of rock mechanics test as soft constraints, a credibility evaluation function is constructed, the water control unit type and the tectonic stress state correction coefficient are coupled, the response accuracy and data quality differentiation law of each detection method are analyzed, and evaluation indicators and physical constraint benchmarks are established. S4.12 Confidence Calibration: Based on the response accuracy, data quality and spatial coverage of each detection method in different water control units, and combined with the differences in tectonic stress state and fracture filling characteristics, the confidence levels of each data source of transient electromagnetic, borehole television and hydrological monitoring are graded and calibrated, multi-scale quantitative calibration parameters are generated, a confidence benchmark system is established and calibration results are formed.

7. The multi-source fusion calibration method for hydrological parameters in a complex coalfield according to claim 6, characterized in that, In step S4.2, the specific method for dynamic weight optimization is as follows: S4.21 Weight Initialization: Based on the confidence calibration results and the physical constraint benchmark, an adaptive weight allocation mechanism is established. According to the confidence level and data quality differences of each detection method in different water control units, the initial contribution weight of each data source is calculated to form the initial fusion weight configuration. S4.22 Dynamic Imbalance Correction: In structural composite parts, combining the differences in structural stress state and crack filling characteristics, the sudden change law of confidence of each data source is analyzed, the initial contribution weight is dynamically adjusted and redistributed, the weight imbalance of key structural parts is corrected, and adaptive optimized fusion weights are generated and output.

8. The multi-source fusion calibration method for hydrological parameters in a complex coalfield according to claim 7, characterized in that, In step S5, the specific method for closed-loop iterative correction is as follows: S5.1 Numerical simulation verification: Input the fused and calibrated hydrological parameters into the groundwater flow numerical simulation model, perform multi-objective fitting verification with measured water level, inflow and water quality data, establish an uncertainty evaluation system, calculate the confidence interval and sensitivity coefficient of hydrological parameters of each water control unit, analyze the source of verification deviation and extract deviation feature information; S5.2 Closed-loop iterative correction: If the verification error exceeds the preset threshold, the deviation information is fed back to step S2 to correct the physical mapping model parameters, and then fed back to step S4 to adjust the weight allocation strategy, forming a closed-loop iterative correction process of physical calibration, fusion optimization and verification feedback, so as to realize the continuous dynamic calibration of hydrological parameters.

Images