A method for identifying roof inrush water sources based on geographic information and coupled with water quality evolution

Abstract

This invention discloses a method for identifying roof inrush water sources based on geographic information and coupled with water quality evolution, including: S1 Collection and geographic informatization of geological and hydrogeological conditions in the study area; S2 Analysis of mineral composition of surrounding rock cores along the inrush path and determination of chemical reactions in water-rock interactions; S3 Analysis of the evolution law of characteristic ions of roof water and updating of water chemical data of target aquifers; S4 Preliminary identification of water-bearing aquifers and identification of roof water sources based on PCA-SVM. The method of this invention can select the corresponding coal seam thickness, water-conducting fracture zone development height, and stratigraphic information according to the inrush coordinates to determine the aquifers that may be involved in the inrush. Combined with geochemical simulation, the evolution results of inrush water quality are determined. The water source identification database is updated based on the evolution results, and the updated database is used as the water source identification standard, thereby solving the problem of inaccurate water source identification after the evolution of inrush water quality.

Description

Technical Field

[0001] This invention relates to the field of mine water control, and in particular to a method for identifying roof water inrush sources based on geographic information and coupled with water quality evolution. Background Technology

[0002] Coal occupies a primary position in my country's primary energy structure, and the safe and efficient mining of coal resources is crucial to national energy security and development. Coal mining faces various threats during the mining process, among which mine water inrush accidents are characterized by their suddenness, wide range of occurrence, rapid flooding, high hazard, and difficulty in rescue, making them a significant factor restricting safe coal mine production. Identification of mine water sources is a crucial foundation for water control efforts; therefore, accurate and efficient identification of water inrush sources is of great importance to coal mine safety.

[0003] Common methods for identifying water sources can be categorized into water temperature and level methods, mathematical theoretical analysis methods, and hydrochemical analysis methods. Among these, hydrochemical composition is a crucial source of information for inferring aquifer characteristics. However, current methods do not integrate the evolution of mine water quality during aquifer inflow with water source identification. Hydrochemical-based mine water source identification models often rely on the original aquifer hydrochemistry as a database. In reality, aquifer water continuously interacts with the formation rock mass during migration, leading to drastic changes in water quality. Ignoring the impact of water quality evolution during aquifer migration in water source identification models often reduces the accuracy and applicability of the models, and may even render them completely inapplicable in some mining areas. Summary of the Invention

[0004] This invention provides a method for identifying roof inrush water sources based on geographic information and coupled with water quality evolution. This method is applicable to roof inrush water source identification scenarios with unstable water quality under special hydrogeological conditions.

[0005] The present invention mainly adopts the following technical means to solve the above-mentioned technical problems:

[0006] A method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution includes:

[0007] Data collection and geographic informatization of geological and hydrogeological conditions in the S1 study area: Collect data on various geological and hydrogeological conditions in the study area and geographic informatize them, that is, use various data as additional attributes of coordinate points. For information points with insufficient data in the study area, on-site sampling is required to improve them, or interpolation methods are used to make various additional attributes effectively distributed in various locations of the mining area.

[0008] S2 Water Inrush Path Surrounding Rock Core Mineral Composition Analysis and Water-Rock Interaction Chemical Reaction Judgment: When water inrush occurs, determine the location information of the water inrush, retrieve the coal seam thickness and main aquifer strata corresponding to the location from the geographic information system, collect the measured data of the mining thickness and water-conducting fracture zone of the main working faces of the surrounding coal mines, determine the development height of the water-conducting fracture zone at the location through similarity analogy method, and compare the aquifer strata to determine the aquifers involved in this water inrush.

[0009] Analysis of the evolution of characteristic ions in S3 roof water and updating of target aquifer water chemistry data: The method for updating the aquifer water chemistry database is as follows: core samples are taken from the target formation through surface or downhole exploration boreholes, the main mineral components present in the rock samples are analyzed, the possible chemical reactions between the main mineral components and roof water are determined, the changes of characteristic ions in the migration path of roof water are simulated through hydrogeochemical reactions, and the initial aquifer water chemistry database is updated based on the results of the characteristic ion changes, forming a water chemistry database after the evolution of inflow water quality.

[0010] Preliminary identification of S4 water-filled aquifer and identification of roof water source based on PCA-SVM: For areas with complex roof water inflow, the hydrochemical database after water quality evolution is retrieved as the standard sample for the water source identification model; the original index data is dimensionality-reduced by principal component analysis to determine the main characteristic indicators for identifying water sources, and the machine learning method of support vector machine is applied to identify the water source inflow.

[0011] Optionally, in step S1, the geographic information represents the spatial coordinates of the main points in the study area, including the boundary turning point of the mining area, the main exploration boreholes in the area, and the roof water inflow point.

[0012] The geological and hydrogeological data involved include stratigraphic lithology and thickness distribution, main coal seam thickness distribution, major stratigraphic positions, and groundwater chemical data. The groundwater chemical data primarily includes Na... + +K + Ca 2+ Mg 2+ Cl - HCO3 - SO4 2- and F - Seven characteristic ion concentration indicators.

[0013] Optionally, in step S2, it is assumed that under the premise of roughly the same geological conditions, the ratio of the water-conducting fracture zone to the coal seam thickness after mining should be basically equal in different working faces in the same mining area. Therefore, the ratio is determined by analogy method, and the development height of the water-conducting fracture zone in the roof of the working face is determined by combining the coal seam thickness attributes of different locations in the geographical information.

[0014] By comparing and analyzing the aquifer layers in the geographic information system, the aquifers that may be involved in this water inrush can be identified. When the height of the water-conducting fracture zone is greater than the distance from the coal seam roof to the aquifer, the aquifer will participate in this roof water inrush.

[0015] If the judgment result indicates that a single aquifer is involved in the water inrush, the water source identification result is output directly; if the judgment result indicates that multiple aquifers are involved in the water inrush, the aquifer water chemistry database needs to be updated.

[0016] Optionally, in step S3, the mineral composition of the main lithology in the roof water inflow path is analyzed by X-ray diffraction; the possible water-rock interaction is determined based on the main mineral composition in the rock sample, and hydrogeochemical reverse simulation is performed by PHREEQC. Different geochemical simulation paths are set, assuming that the migration process of roof water is A→B→C, where the surrounding rocks of the path from A to B are mainly composed of mudstone sections, and the surrounding rocks of the path from B to C are mainly composed of sandstone sections.

[0017] The changes in characteristic ion concentrations before and after the top water runoff passes through the mudstone and sandstone sections were analyzed to establish the influence of mudstone and sandstone of unit thickness on the concentration of major characteristic ions.

[0018] Based on this, and combined with the coal seam roof lithological combination data in the geographic information, a roof water inflow water quality evolution module with the path of "water source → runoff strata → working face" is formed. This module is used to update the initial aquifer hydrochemical database to form a hydrochemical database after the evolution of water inflow water quality.

[0019] Optionally, in step S3: the mineral saturation index in the water sample is calculated using PHREEQC, and the formula for calculating the mineral saturation index is:

[0020]

[0021] In the formula; IAP is the ion activity product; K is the equilibrium constant; SI = 0 indicates that the mineral is in equilibrium, SI < 0 indicates that the mineral is in dissolution, and SI > 0 indicates that the mineral is in precipitation.

[0022] Using PHREEQC's reverse hydrogeochemical simulation function, the amount of mineral dissolution or precipitation at two points along the groundwater flow path was determined.

[0023] Optionally, the influence of the unit thickness of mudstone and sandstone on the concentration of key characteristic ions includes:

[0024] Starting with the initial groundwater quality, x1 (mmol) of the first mineral dissolved or precipitated per unit concentration in the water is recorded. Precipitation is recorded as a negative value, solubility as a positive value, and so on, up to x. nAfter the nth mineral (mmol) was added, the "endpoint" water quality was formed; the increase in the i-th element was b. i (mmol), and its mass balance equation is:

[0025]

[0026] In the formula, a ij is the stoichiometric coefficient of the i-th element relative to the j-th mineral; n is the total number of equilibrium elements in the aqueous solution.

[0027] Optionally, in step S4, if the judgment result is water inflow from the top plate of the composite aquifer, the initial water quality data of multiple aquifers are input into the water quality evolution module, and the output results are used as the characteristic ion concentration index of each aquifer. It is considered that the water quality of the inflow is a mixed result after the water quality evolution of multiple aquifers. The mixing ratio of each aquifer is deduced from the water quality detection results of the inflow, and the main water-filling aquifer is determined.

[0028] The advantages of this invention are:

[0029] The method of this invention can select corresponding coal seam thickness, water-conducting fracture zone development height, and stratigraphic information based on the water inrush coordinates to determine the aquifers that may be involved in the water inrush. Furthermore, it applies the lithological composition and thickness of the water inrush path from geographic information, combined with geochemical simulation, to determine the water quality evolution results. Based on these evolution results, the water source identification database is updated, and the updated database serves as the standard for water source identification, thereby solving the problem of inaccurate water source identification after the evolution of water quality inrushes. Attached Figure Description

[0030] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0031] Figure 1 A flowchart illustrating the workflow of the present invention is shown;

[0032] Figure 2 A schematic diagram illustrating the geographic information system of coal thickness data in an embodiment is shown;

[0033] Figure 3 This example illustrates a geographic information diagram of the distance data between coal seam 4 and the bottom of the Yijun Formation.

[0034] Figure 4 A schematic diagram illustrating the geographic information system of the thickness data of the straight section group is shown in the embodiment.

[0035] Figure 5 A schematic diagram illustrating the geographic information system of the thickness data of the Luohe + Yijun group in an embodiment is shown.

[0036] Figure 6 A geographic information map illustrating the impact of the water-conducting fracture zone on the Luohe Formation in an embodiment is provided. Detailed Implementation

[0037] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0038] Currently, there is an urgent need for a method to identify roof-driven water inflow sources that can address the issue of water quality evolution. This is of great significance for the prevention and control of roof-driven water hazards in mining areas of western my country. To address this problem, this invention proposes a roof-driven water inflow source identification method based on geographic information and coupled with water quality evolution. This method selects corresponding coal seam thickness, water-conducting fracture zone development height, and stratigraphic information based on the water inflow coordinates to determine the aquifers that may be involved in the water inflow. Furthermore, it applies the lithological composition and thickness of the water inflow path from the geographic information, combined with geochemical simulation, to determine the water quality evolution results. The water source identification database is updated based on these evolution results, and the updated database serves as the standard for water source identification, thereby solving the problem of inaccurate water source identification after water quality evolution.

[0039] Combination Figure 1 The present invention provides a method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution, comprising:

[0040] Data collection and geographic informatization of geological and hydrogeological conditions in the S1 study area: Collect data on various geological and hydrogeological conditions in the study area and geographic informatize them, that is, use various data as additional attributes of coordinate points. For information points with insufficient data in the study area, on-site sampling is required to improve them, or interpolation methods are used to make various additional attributes effectively distributed in various locations of the mining area.

[0041] Further, in step S1, the geographic information represents the spatial coordinates of key points within the study area, including the boundary inflection points of the mining area, major exploration boreholes within the area, and roof water inflow points. The geological and hydrogeological data involved include stratigraphic lithology and thickness distribution, main coal seam thickness distribution, major stratigraphic positions, and groundwater chemical data, wherein the groundwater chemical data mainly includes Na... + +K + Ca 2+ Mg 2+ Cl - HCO3 - SO4 2- F - Seven indicators, specifically the characteristic ion concentration indicator.

[0042] S2 Water Inrush Path Surrounding Rock Core Mineral Composition Analysis and Water-Rock Interaction Chemical Reaction Judgment: When water inrush occurs, determine the location information of the water inrush, retrieve the coal seam thickness and main aquifer strata corresponding to that location from the geographic information system, statistically analyze the mining thickness and measured data of the water-conducting fracture zone of the main working faces of surrounding coal mines, determine the development height of the water-conducting fracture zone at that location through similarity analogy method, and compare the aquifer strata to determine the aquifers involved in this water inrush.

[0043] Furthermore, in step S2, it is assumed that under the premise of roughly the same geological conditions, the ratio of the water-conducting fracture zone to the coal seam thickness after mining should be basically equal in different working faces of the same mining area. Therefore, this ratio is determined by analogy method, and the development height of the water-conducting fracture zone in the roof of the working face is determined by combining the coal seam thickness attributes of different locations in the geographical information.

[0044] Furthermore, a comparative analysis is performed in the geographic information system with the aquifer layers to determine the aquifers that may be involved in this water inrush. When the height of the water-conducting fracture zone is greater than the distance from the coal seam roof to the aquifer, the aquifer will participate in this roof water inrush.

[0045] If the determination result indicates that a single aquifer is involved in the water inrush, the water source identification result is output directly. If the determination result indicates that multiple aquifers are involved in the water inrush, the aquifer water chemistry database needs to be updated.

[0046] Analysis of the evolution of characteristic ions in S3 roof water and updating of target aquifer water chemistry data: The method for updating the aquifer water chemistry database is as follows: core samples are taken from the target formation through surface or downhole exploration boreholes, the main mineral components present in the rock samples are analyzed, the possible chemical reactions between the main mineral components and roof water are determined, the changes of characteristic ions in the migration path of roof water are simulated through hydrogeochemical reactions, and the initial aquifer water chemistry database is updated based on the results of the characteristic ion changes, forming a water chemistry database after the evolution of inflow water quality.

[0047] Further, in step S3, the mineral composition of the main lithologies in the roof water inflow path is analyzed by X-ray diffraction (XRD). For example, the main mineral components of mudstone include (Al2Si2O5(OH)4) kaolinite, (CaCO3) calcite, (CaMg(CO3)2) dolomite, (NaAlSi3O8) albite, and (FeS2) pyrite, while the main mineral components of sandstone include quartz (SiO2), albite, and calcium feldspar (CaAl2Si2O8).

[0048] Furthermore, based on the main mineral composition of the rock samples, possible water-rock interactions were identified. For example, the oxidation of sulfides such as pyrite produces free acids, which then neutralize calcite, dolomite, and other minerals in the strata, thereby increasing the Ca content in the topwater. 2+Mg 2+ SO4 2- The ion content increases during migration, and the chemical formula for the reaction is as follows:

[0049] 4FeS2+14H2O+15O2=4Fe(OH)3+8SO4 2- +16H +

[0050] CaCO3 + 4H + =Ca 2+ +2H₂O + CO₂↑

[0051] MgCa(CO3)2+4H + =Mg 2+ +Ca 2+ +2H₂O + 2CO₂↑

[0052] Furthermore, since primary rock-forming minerals such as albite and calcium feldspar are not saturated in water, they will still dissolve in the roof water inflow, producing Ca. 2+ Mg 2+ Plasma, the chemical formula of the reaction is as follows:

[0053] 4NaAlSi3O8+4CO2+22H2O=Al4Si4O 10 (OH)₈ + 4Na + +8H4SiO4+4HCO3 -

[0054] 2CaAl2Si2O8+4CO2+6H2O=Al4Si4O 10 (OH)₈ + 2Ca 2+ +4HCO3 -

[0055] Furthermore, hydrogeochemical inverse simulations were conducted using PHREEQC, setting different geochemical simulation paths. It was assumed that the migration process of the roof water was A→B→C, where the surrounding rock along the path from A to B mainly consisted of a mudstone section of a certain thickness, and the surrounding rock along the path from B to C mainly consisted of a sandstone section of a certain thickness.

[0056] Furthermore, the changes in characteristic ion concentrations before and after the top surface water runoff passes through the mudstone and sandstone sections were analyzed to establish the influence of mudstone and sandstone of unit thickness on the concentration of major characteristic ions.

[0057] Assuming the initial groundwater quality is taken as the "starting point," x1 (mmol) of the first mineral is dissolved (or precipitated) per unit concentration in the water (precipitation is negative, dissolution is positive), ..., x nAfter the nth mineral (mmol) was added, the "final" water quality was formed. The increase in the i-th element was b. i (mmol), and its mass balance equation is:

[0058]

[0059] In the formula, a ij is the stoichiometric coefficient of the i-th element relative to the j-th mineral; n is the total number of equilibrium elements in the aqueous solution.

[0060] Furthermore, based on this, and combined with the coal seam roof lithological combination data in the geographic information, a roof water inflow water quality evolution module with the path of "water source → runoff strata → working face" is formed. This module is used to update the initial aquifer hydrochemical database to form a hydrochemical database after the evolution of water inflow water quality.

[0061] Preliminary identification of the S4 aquifer and roof water source identification based on PCA-SVM: For areas with complex roof water inflow, the hydrochemical database after water quality evolution was retrieved as the standard sample for the water source identification model. Principal component analysis was used to reduce the dimensionality of the original index data to determine the main characteristic indicators for identifying water sources. The support vector machine learning method was then applied to identify the water inflow sources.

[0062] Furthermore, if the judgment result is that water is flowing from the top of a composite aquifer, the initial water quality data of multiple aquifers are input into the water quality evolution module, and the output results are used as the characteristic ion concentration index of each aquifer. It is believed that the water quality of the water inflow is a mixed result of the water quality evolution of multiple aquifers. The mixing ratio of each aquifer is deduced from the water quality detection results of the water inflow, and the main water-filling aquifer is determined.

[0063] Example 1:

[0064] This example uses a water volume of 500m³ from a specific mine in the Binchang mining area. 3 Taking the roof inrush event as an example, the Cretaceous Luohe Formation aquifer is the main water-bearing aquifer in this area. This aquifer has low mineralization (510-550 mg / L) and belongs to HCO3-. - Ca·Na·Mg type water; the Jurassic Yan'an-Zhiluo Formation aquifer is also present, which generally has high mineralization, ranging from 2400 to 4700 mg / L, and the water quality is mostly SO4·Cl. - Na. Generally, water source identification models based on water quality analysis indicate that the water inflow at the working face is biased towards the Jurassic aquifer. However, the water level of the Luohe Formation aquifer showed a significant downward trend after the inflow, and previous hydrological work considered the Jurassic aquifer to have extremely weak water-bearing capacity, making it difficult to support a large flow inflow. Therefore, the water source identification results are not convincing.

[0065] Based on this, this example explains the phenomenon using a roof-mounted water inrush source identification method that combines geographic information with water quality evolution, including the following steps:

[0066] S1: Collect data on the thickness of the main coal seam, groundwater level, thickness of the main aquifer (aquifer), distance from the main coal seam to the main aquifer, aquifer water quality, and Jurassic strata lithology in the study area, and digitize them using geographic information systems (e.g., Figures 2 to 5 As shown in the figure, Table 1 lists the stratigraphic lithology combinations revealed by some boreholes.

[0067] Table 1. Statistical Table of Lithological Assemblages of Jurassic Strata The results of special exploration and mining investigations conducted in multiple mines in the Binchang mining area were collected and compiled. The measured results show that the ratio of the development height of the water-conducting fracture zone to the working face thickness is between 22.3 and 38.5. The statistical results are shown in Table 2.

[0068] Table 2. Statistics on the Fracturing Ratio of Surrounding Coal Mines

[0069]

[0070]

[0071] The working face has a thickness of approximately 8m. Based on analogy analysis, the height of the water-conducting fracture zone at this location should be between 178.4 and 236m. The distance from the top of the working face to the Luohe Formation aquifer is 143m. Using drilling data from various boreholes, a geographical distribution map showing whether the water-conducting fracture zone in this coal mine can affect the Luohe Formation aquifer was drawn using the natural nearest neighbor interpolation method (e.g., [missing data]). Figure 6 As shown in the figure, the black part is the area where the water-conducting fracture zone can affect the Luohe Formation aquifer. The top surface water inrush that occurs in this area after mining is a mixed water inrush involving both the Zhiluo Formation and the Luohe Formation. The Luohe Formation hydrochemical database needs to be updated before it can be used for water source identification.

[0072] S2: Samples were taken from the Jurassic strata through surface borehole D47. X-ray diffraction (XRD) analysis revealed that the main mineral components of the sandy mudstone were quartz and albite, the main mineral components of the mudstone were montmorillonite, kaolinite, and a small amount of quartz, and the main mineral components of the sandstone were quartz, feldspar, and a small amount of calcite and pyrite. Therefore, when roof water inrush occurs at the working face near borehole D47, the water from the Luohe Formation aquifer will react with the 51-meter sandy mudstone section, the 29-meter sandstone section, and the 63-meter mudstone section as it flows through the Jurassic strata. The main minerals participating in the water-rock reaction are albite, calcite, and pyrite. The possible chemical reactions are as follows: albite dissolves in the unsaturated Luohe water to produce kaolinite, generating sodium ions and bicarbonate ions; pyrite oxidation produces free hydrogen ions and sulfate ions; and calcite reacts with free hydrogen ions to produce calcium ions and carbon dioxide.

[0073] The main reaction equations that occur when the Luohe River flows through the Jurassic strata are as follows:

[0074] 4NaAlSi3O8+4CO2+22H2O=Al4Si4O 10 (OH)₈ + 4Na + +8H4SiO4+4HCO3 -

[0075] 4FeS2+14H2O+15O2=4Fe(OH)3+8SO4 2- +16H +

[0076] CaCO3 + 4H + =Ca 2+ +2H₂O + CO₂↑

[0077] S3: The saturation index of minerals such as albite, calcite, and pyrite in the water sample was calculated using PHREEQC. The formula for calculating the mineral saturation index is as follows:

[0078]

[0079] In the formula: IAP is the ion activity product; K is the equilibrium constant. SI = 0 indicates that the mineral is in equilibrium, SI < 0 indicates that the mineral is in dissolution, and SI > 0 indicates that the mineral is in precipitation.

[0080] Furthermore, using PHREEQC's reverse hydrogeochemical simulation function, the amount of mineral dissolution or precipitation at two points along the groundwater flow path was determined. The dissolution equilibrium reaction of the mineral phases is expressed as follows:

[0081] SOLUTION 1: Luo He aquifer names solutions and defines the ion content of solutions:

[0082]

[0083]

[0084] Sulfate 61.91

[0085] Phases of the reaction

[0086] Halite

[0087] Gypsum

[0088]

[0089] PHASES

[0090] Albite

[0091] 4NaAlSi3O8+4CO2+22H2O=Al4Si4O 10 (OH)₈ + 4Na + +8H4SiO4+4HCO3 - Logk 0.0

[0092] Marcasite

[0093] 4FeS2+14H2O+15O2=4Fe(OH)3+8SO4 2- +16H +

[0094] Calcite

[0095] CaCO3 + 4H + =Ca 2+ +2H₂O + CO₂↑

[0096] END.

[0097] The reactant composition and molar conversion results of the Luohe Formation aquifer water flowing through the Jurassic strata are shown in Table 3 below.

[0098] Table 3. Results of reactant composition and molar conversion during water inrush.

[0099]

[0100]

[0101] S4: Based on the reactant composition and molar conversion results, an evolved hydrochemical database of roof inrush water was established; and ion correlation analysis was used to determine whether there was information overlap among multiple indicators. Principal component analysis was performed on the samples to reduce information overlap, thereby reducing computational load and improving the accuracy of water source identification. The indicators were standardized through neural network training, and TDS was finally selected as the characteristic factor for this water source identification. Table 4 is the explanatory variance table of this principal component analysis.

[0102] Table 4. Explained Variance of Principal Components

[0103]

[0104] The water quality evolution calculation results show that the initial TDS of the water in the Luohe Formation aquifer near borehole D47 was 516.3 mg / L. After passing through 143m of Jurassic strata, its TDS evolved to 2721.1 mg / L. Using a support vector machine neural network model to identify the complex roof inrush, it was determined that the main recharge aquifer for this inrush was the Luohe Formation aquifer, which accounted for 87.9% of the total water volume. The judgment result is consistent with the water-bearing capacity and water level observation analysis results of the main aquifers in the mining area, and the water source identification result is considered accurate.

[0105] This invention adds factors such as coal seam thickness, stratigraphic lithology, and aquifer hydrochemical characteristics to the geographical location, allowing for the inclusion of more identification factors compared to traditional water source identification methods. Furthermore, it incorporates a water quality evolution module into the traditional water source discrimination model, solving the problems of large water quality variations under special hydrogeological conditions and low accuracy of traditional water source identification models. This achieves rapid and efficient identification of roof-mounted water sources under water quality evolution conditions.

[0106] Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution, characterized in that, include: Data collection and geographic informatization of geological and hydrogeological conditions in the S1 study area: Collect data on various geological and hydrogeological conditions in the study area and geographic informatize them, that is, use various data as additional attributes of coordinate points. For information points with insufficient data in the study area, on-site sampling is required to improve them, or interpolation methods are used to make various additional attributes effectively distributed in various locations of the mining area. S2 Water Inrush Path Surrounding Rock Core Mineral Composition Analysis and Water-Rock Interaction Chemical Reaction Judgment: When water inrush occurs, determine the location information of the water inrush, retrieve the coal seam thickness and main aquifer strata corresponding to the location from the geographic information system, collect the measured data of the mining thickness and water-conducting fracture zone of the main working faces of the surrounding coal mines, determine the development height of the water-conducting fracture zone at the location through similarity analogy method, and compare the aquifer strata to determine the aquifers involved in this water inrush. S3 Top Water Characteristic Ion Evolution Law Analysis and Target Aquifer Hydrochemical Data Update: The aquifer hydrochemical database update method is as follows: core samples are taken from the target formation through surface or downhole exploration boreholes, the main mineral components present in the rock samples are analyzed, the possible chemical reactions between the main mineral components and the top water are determined, the changes of characteristic ions of the top water in the migration path are simulated through hydrogeochemical reactions, and the initial aquifer hydrochemical database is updated based on the results of characteristic ion changes to form a hydrochemical database after the evolution of the inflow water quality; Preliminary identification of S4 water-filled aquifer and identification of roof water source based on PCA-SVM: For areas with complex roof water inflow, the hydrochemical database after water quality evolution is retrieved as the standard sample for the water source identification model; the original index data is dimensionality-reduced by principal component analysis to determine the main characteristic indicators for identifying water sources, and the machine learning method of support vector machine is applied to identify the water source inflow.

2. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution as described in claim 1, characterized in that, In step S1, the geographic information refers to the spatial coordinates of the main points in the study area, including the boundary inflection points of the mining area, the main exploration boreholes in the area, and the roof water inflow points. The geological and hydrogeological data involved include stratigraphic lithology and thickness distribution, main coal seam thickness distribution, major stratigraphic positions, and groundwater chemical data. The groundwater chemical data primarily includes Na+. + +K + Ca 2+ Mg 2+ Cl - HCO3 - SO4 2- and F - Seven characteristic ion concentration indicators.

3. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution according to claim 1 or 2, characterized in that, In step S2, it is assumed that under the premise of roughly the same geological conditions, the ratio of the water-conducting fracture zone to the coal seam thickness after mining should be basically equal in different working faces in the same mining area. Therefore, the ratio is determined by analogy method, and the development height of the water-conducting fracture zone in the roof of the working face is determined by combining the coal seam thickness attributes of different locations in the geographical information. By comparing and analyzing the aquifer layers in the geographic information system, the aquifers that may be involved in this water inrush can be identified. When the height of the water-conducting fracture zone is greater than the distance from the coal seam roof to the aquifer, the aquifer will participate in this roof water inrush. If the judgment result indicates that a single aquifer is involved in the water inrush, the water source identification result is output directly; if the judgment result indicates that multiple aquifers are involved in the water inrush, the aquifer water chemistry database needs to be updated.

4. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution according to claim 1 or 2, characterized in that, In step S3, the mineral composition of the main lithology in the roof water inflow path is analyzed by X-ray diffraction; the possible water-rock interaction is determined based on the main mineral composition in the rock sample; hydrogeochemical reverse simulation is performed by PHREEQC, and different geochemical simulation paths are set. It is assumed that the migration process of roof water is A→B→C, where the surrounding rocks of the path from A to B are mainly composed of mudstone sections, and the surrounding rocks of the path from B to C are mainly composed of sandstone sections. The changes in characteristic ion concentrations before and after the top water runoff passes through the mudstone and sandstone sections were analyzed to establish the influence of mudstone and sandstone of unit thickness on the concentration of major characteristic ions. Based on this, and combined with the coal seam roof lithological combination data in the geographic information, a roof water inflow water quality evolution module with the path "water source → runoff strata → working face" is formed. This module is used to update the initial aquifer hydrochemical database to form a hydrochemical database after the evolution of water inflow water quality.

5. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution according to claim 4, characterized in that, In step S3: the mineral saturation index in the water sample is calculated using PHREEQC. The formula for calculating the mineral saturation index is: In the formula; IAP is the ion activity product; K is the equilibrium constant; SI = 0 indicates that the mineral is in equilibrium, SI < 0 indicates that the mineral is in dissolution, and SI > 0 indicates that the mineral is in precipitation. Using PHREEQC's reverse hydrogeochemical simulation function, the amount of mineral dissolution or precipitation at two points along the groundwater flow path was determined.

6. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution according to claim 4, characterized in that, The influence of mudstone and sandstone of unit thickness on the concentration of key characteristic ions includes: Starting with the initial groundwater quality, x1 (mmol) of the first mineral dissolved or precipitated per unit concentration in the water is recorded. Precipitation is recorded as a negative value, dissolution as a positive value, and so on, up to x. n After the nth mineral (mmol) was added, the "endpoint" water quality was formed; the increase in the i-th element was b. i (mmol), its mass balance equation is: In the formula, a ij is the stoichiometric coefficient of the i-th element relative to the j-th mineral; n is the total number of equilibrium elements in the aqueous solution.

7. The method for identifying roof-mounted water inrush sources based on geographic information and coupled with water quality evolution according to claim 1 or 2, characterized in that, In step S4, if the judgment result is water inflow from the top plate of the composite aquifer, the initial water quality data of multiple aquifers are input into the water quality evolution module, and the output results are used as the characteristic ion concentration index of each aquifer. It is considered that the water quality of the inflow is a mixture of the water quality of multiple aquifers after evolution. The mixing ratio of each aquifer is deduced from the water quality detection results of the inflow, and the main water-filling aquifer is determined.

Images