Karst mountain groundwater level prediction method and system

Abstract

The application discloses a karst mountain area underground water level prediction method and system, relates to the underground water level prediction technical field, constructs the underground water level prediction model of the karst mountain area, carries out underground excavation simulation, analyzes the prediction efficiency of the prediction model, carries out subsequent optimization when the prediction efficiency is poor, when the prediction effect is good, according to the result of simulation, confirms the excavation position with large underground water level change in the excavation process, and predicts possible dangerous phenomena, then formulates a prevention scheme, provides a reference for the protection of subsequent actual excavation, the prediction efficiency of the prediction model is evaluated, the change of underground water level can be quickly and accurately obtained in subsequent actual construction, protection is carried out in time, the occurrence of water inrush, mud inrush and other accidents is reduced, the life safety and equipment safety of construction personnel are ensured, the phenomenon of construction interruption or the need for additional treatment measures is reduced, thereby the engineering progress is ensured, and the engineering cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of groundwater level prediction technology, specifically to a method and system for predicting groundwater levels in karst mountainous areas. Background Technology

[0002] Karst mountain areas are characterized by well-developed underground caves and fissures, with abundant and complex groundwater flow. Excavation in karst mountain areas may disrupt existing groundwater channels and impermeable layers, allowing groundwater to surge into the construction area and causing water inrush accidents. Furthermore, changes in groundwater levels can have long-term adverse effects on the structure, such as causing leakage and corrosion. Predicting groundwater levels during excavation helps in implementing effective maintenance measures during the project's operation phase, ensuring the long-term safe use of the project.

[0003] Existing technologies, such as the invention patent disclosed in CN119322987A, which describes a groundwater level prediction method and system based on a spatiotemporal attention-gated neural network, include: constructing a historical groundwater level spatiotemporal sequence dataset; performing principal component analysis on the spatiotemporal sequence dataset to obtain a dimensionality-reduced spatiotemporal sequence dataset; constructing multiple spatiotemporal feature matrices for the dimensionality-reduced spatiotemporal sequence dataset using a sliding window method; training a groundwater level prediction model using the multiple spatiotemporal feature matrices to obtain a trained groundwater level prediction model; acquiring spatiotemporal sequence data of historical groundwater levels at the location to be measured, constructing a spatiotemporal feature matrix of groundwater levels, and using the trained groundwater level prediction model to predict the groundwater level at the location to be measured. This invention achieves rapid and accurate prediction of groundwater levels, greatly improving the generalization ability, prediction accuracy, and interpretability of the groundwater level prediction model while reducing modeling costs.

[0004] Existing technologies, such as the invention patent disclosed in CN114881323A, "A Method for Predicting and Updating Groundwater Levels in Foundation Pit Dewatering Areas Based on Deep Neural Networks," are used for predicting and updating the 2-D spatial distribution of groundwater levels in foundation pit dewatering areas. This method includes: collecting groundwater level observation data from various observation points, constructing a deep neural network model, and establishing functional relationships through a deep neural network algorithm to predict and update the groundwater level in the target area, providing information for foundation pit construction. Based on the observation data, it integrates multiple loss functions to solve the problems of previous neural network models in groundwater level prediction, such as single spatial location of prediction points, low accuracy of observation data, large data dispersion, poor real-time update effect, and inability to timely and fully reflect the actual situation of water level changes. This leads to the inability to predict groundwater levels during construction, resulting in low construction efficiency and the inability to obtain accurate data.

[0005] The above-mentioned scheme details the construction and optimization process of the groundwater level prediction model. However, karst mountainous areas have complex geological structures, and the prediction model involves the fusion of multiple data sources, which affects the prediction efficiency of groundwater levels. The speed of groundwater level prediction affects the excavation project. However, the above-mentioned scheme does not conduct simulation tests on the prediction efficiency of the prediction model after its construction is completed. It is impossible to grasp the prediction speed of the prediction model, and it is impossible to quickly and accurately obtain groundwater level changes in subsequent actual construction. As a result, it is impossible to take timely protective measures, reduce the occurrence of accidents such as water inrush and mud inrush, and ensure the safety of construction personnel and equipment. In addition, if the groundwater level prediction is not timely, construction may be interrupted due to water level changes or additional treatment measures may be required, thereby delaying the project progress and increasing project costs. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the present invention aims to provide a method and system for predicting groundwater levels in karst mountainous areas.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: In the first aspect, the present invention provides a method for predicting groundwater level in karst mountainous areas, S1, constructing a water level prediction model, and simultaneously obtaining a three-dimensional model of a specified karst mountainous area and an underground excavation plan for the specified karst mountainous area, simulating underground excavation in the specified karst mountainous area, and using the water level prediction model to predict the water level in the underground excavation simulation.

[0008] S2. Based on the simulation of underground excavation, predict the water level and analyze the water level prediction efficiency of the water level prediction model. If the water level prediction efficiency is poor, execute S4; if the water level prediction efficiency is good, execute S3.

[0009] S3. Use the water level prediction model to predict the water level in the underground excavation simulation, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute S4.

[0010] S4. When the water level prediction efficiency is poor, feedback is sent to the visualization display terminal; preventive measures for underground excavation in designated karst mountain areas are fed back to the visualization display terminal, and preventive prompts are provided.

[0011] Secondly, the present invention provides a groundwater level prediction system for karst mountain areas, comprising: a model building and prediction module, used to build a water level prediction model, and simultaneously acquire a three-dimensional model of a specified karst mountain area and an underground excavation plan for the specified karst mountain area, simulate underground excavation in the specified karst mountain area, and use the water level prediction model to predict the water level of the underground excavation simulation.

[0012] The model prediction and evaluation module is used for water level prediction based on underground excavation simulation. It analyzes the water level prediction efficiency of the water level prediction model. When the water level prediction efficiency is poor, the prediction execution module is executed. When the water level prediction efficiency is good, the prediction result processing module is executed.

[0013] The prediction result processing module is used to predict the water level of the underground excavation simulation using the water level prediction model, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute the prediction execution module.

[0014] The prediction execution module is used to feed back information to the visualization display terminal when the water level prediction efficiency is poor; it also feeds back preventive measures for underground excavation in designated karst mountain areas to the visualization display terminal and provides preventive prompts.

[0015] The beneficial effects of this invention are as follows: This application provides a method and system for predicting groundwater levels in karst mountainous areas. A groundwater level prediction model for karst mountainous areas is constructed, and underground excavation simulation is performed. The prediction efficiency of the model is analyzed. When the prediction efficiency is poor, subsequent optimization is carried out. When the prediction effect is good, based on the simulation results, the excavation locations with significant groundwater level changes during the excavation process are identified, and potential dangerous phenomena are predicted. Then, prevention plans are formulated to provide a reference for protection during subsequent actual excavation. By evaluating the prediction efficiency of the model, this application ensures that groundwater level changes can be quickly and accurately obtained during subsequent actual construction, enabling timely protection, reducing the occurrence of accidents such as water inrush and mudslide, protecting the lives and equipment of construction personnel, and reducing the occurrence of construction interruptions or the need for additional treatment measures. This ensures project progress and reduces project costs. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0018] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example 1

[0020] See Figure 1 As shown, a groundwater level prediction method for karst mountain areas includes the following steps: S1, constructing a water level prediction model, simultaneously obtaining a three-dimensional model of a specified karst mountain area and an underground excavation plan for the specified karst mountain area, simulating underground excavation in the specified karst mountain area, and using the water level prediction model to predict the water level in the underground excavation simulation.

[0021] In a specific embodiment, the specific process of constructing the water level prediction model is as follows: S11, through geological exploration, obtain geological data and hydrogeological data of the designated karst mountain area, construct a conceptual model of the designated karst mountain area, and use numerical simulation software to divide the designated karst mountain area into grids and establish a three-dimensional groundwater flow numerical model.

[0022] Numerical simulation software such as Visual MODFLOW.

[0023] In the above, geological exploration includes drilling, geophysical exploration and other methods, geological data includes stratigraphic distribution, rock type, karst development characteristics, etc., and hydrogeological data includes water flow path and flow velocity changes, etc.

[0024] A conceptual model is an abstract and simplified qualitative description of a groundwater system in karst mountainous areas. It is mainly used to help understand and analyze the operating mechanism of the groundwater system and to provide a basic framework for subsequent model building. Specifically, it includes boundary determination, aquifer characteristic description, water flow process analysis, generalization of hydrogeological conditions, and identification of major influencing factors.

[0025] Boundary Determination: Clearly define the scope of the groundwater system in karst mountainous areas, including its horizontal and vertical boundaries. For example, determine the boundaries of the groundwater recharge zone, runoff zone, and discharge zone, as well as the top and bottom interfaces of the aquifer, based on geological structures, topography, and other factors.

[0026] Aquifer characteristics description: This involves classifying and describing the characteristics of aquifers in karst mountainous areas, including their lithology, permeability, and porosity. For example, it distinguishes between karst-developed carbonate aquifers and relatively impermeable non-carbonate strata, and describes the unique water storage and conduction space characteristics of carbonate aquifers, such as dissolution pores, fissures, and caves formed by karst processes.

[0027] Water flow process analysis: This involves analyzing the flow process of groundwater in karst mountainous areas, including recharge, runoff, and discharge. For example, it considers the recharge of groundwater by atmospheric precipitation through surface infiltration and karst conduits, the runoff of groundwater along a certain hydraulic gradient in the aquifer, and the discharge through springs, rivers, and other pathways.

[0028] Hydrogeological condition generalization: This involves reasonably simplifying and generalizing complex actual hydrogeological conditions. For example, the topographic relief of karst mountain areas can be generalized into a relatively simple topographic model, and the heterogeneity of aquifers can be averaged to a certain extent to facilitate subsequent calculations.

[0029] Identification of key influencing factors: Determine the main factors affecting groundwater level changes in karst mountain areas, such as meteorological factors (including precipitation and evaporation), geological factors (including stratigraphic lithology and geological structure), and human activities (including groundwater extraction and engineering construction). Clarify how these factors interact with the groundwater system to influence the dynamic changes in groundwater levels.

[0030] The specific process of the conceptual model has been publicly disclosed and can be found on the Internet, so it will not be elaborated here.

[0031] S12. Based on the three-dimensional groundwater flow numerical model, long-term observation data of a specified karst mountain area are obtained from the data center. Correlation analysis is performed on the long-term observation data of the specified karst mountain area. Statistical methods are used to establish a statistical model for groundwater level prediction. The parameters of the model are determined through regression analysis.

[0032] The statistical methods mentioned above include multiple linear regression and time series analysis. Statistical models for groundwater level prediction, for example, use precipitation and river flow as independent variables and groundwater level as the dependent variable. The specific process is publicly available online and will not be elaborated upon here.

[0033] S13. Optimize and correct the model using unused data from long-term observation data to obtain a water level prediction model.

[0034] In the above, the long-term observation data is divided into the model building dataset and the model optimization dataset. The model is built using the data in the model building dataset, and the unused data in the long-term observation data is the model optimization dataset. The specific model optimization process has been disclosed in existing technology and will not be elaborated here.

[0035] The long-term observation data consists of data collected over a long period of time from multiple groundwater level monitoring points set up in designated karst mountain areas using various monitoring equipment. This data includes underground geological structure, groundwater level, groundwater pH, groundwater conductivity, groundwater dissolved oxygen, surface water level, surface water flow rate, and surface water temperature. The monitoring equipment includes water level gauges, multi-parameter water quality analyzers, and ground-penetrating radar.

[0036] In another specific embodiment, the process of simulating underground excavation in a designated karst mountain area is as follows: obtain each excavation location, the excavation parameters of each excavation location, and the excavation depths for which groundwater prediction is required at each excavation location from the underground excavation plan, and use each excavation depth for which groundwater prediction is required as the prediction depth.

[0037] The excavation parameters include the size and depth of the excavation.

[0038] In the three-dimensional model of the specified karst mountain area, excavation simulation is carried out sequentially according to each excavation location and its excavation parameters. When the underground excavation simulation is carried out, when the excavation depth of each excavation location reaches the predicted depth, the groundwater level at each excavation location is predicted by the water level prediction model. At the same time, the start time and end time of the prediction of the groundwater level at each excavation location and the predicted depth are obtained from the computer background records and used as the initial time and end time of the groundwater level prediction at each predicted depth in each excavation location.

[0039] The actual predicted depth of the predicted groundwater level at each excavation location is obtained from the three-dimensional model of the specified karst mountain area. The actual predicted depth of the predicted groundwater level at each excavation location is subtracted from the actual predicted depth of the predicted groundwater level at each excavation location to obtain the excavation depth of each excavation location at the predicted groundwater level.

[0040] The groundwater level at each excavation location when the excavation depth reaches each predicted depth, the excavation depth at each excavation location when the groundwater level at each predicted depth is predicted, and the initial and end times of the predicted groundwater level at each predicted depth in each excavation location are used as the water level prediction simulation data.

[0041] S2. Based on the simulation of underground excavation, predict the water level and analyze the water level prediction efficiency of the water level prediction model. If the water level prediction efficiency is poor, execute S4; if the water level prediction efficiency is good, execute S3.

[0042] In a specific embodiment, the water level prediction efficiency of the water level prediction model is specifically analyzed as follows: using the groundwater level at each excavation location when the excavation depth reaches each prediction depth in the water level prediction simulation data, the prediction efficiency influence coefficient of each excavation location at each prediction depth is analyzed, and the prediction duration of the groundwater level at each prediction depth in each excavation location is obtained based on the initial time and end time of the groundwater level prediction at each prediction depth in each excavation location.

[0043] The prediction efficiency influence coefficient of each excavation location at each prediction depth, the prediction time of groundwater level at each prediction depth in each excavation location, and the excavation depth of each excavation location at each prediction depth when predicting groundwater level are input into the water level prediction efficiency formula, and the water level prediction efficiency analysis results of the water level prediction model are output.

[0044] The results of the water level prediction efficiency analysis include YES and NO. When the result is YES, it indicates that the water level prediction efficiency is good, and vice versa.

[0045] The analysis process for the prediction efficiency influence coefficient of each excavation location at each prediction depth, as described above, is as follows:

[0046] The groundwater level at each excavation location when the excavation depth reaches the predicted depth is denoted as . , where r represents the number of each excavation location, y represents the number of each predicted depth, and both r and y are positive integers;

[0047] Using analytical formulas: The prediction efficiency influence coefficient of the r-th excavation location at the y-th prediction depth is obtained. In the formula, and Let represent the groundwater level at the r-th excavation location when the excavation depth reaches the (y-1)-th predicted depth and the groundwater level at the r-th predicted depth, respectively. This represents the groundwater level at the r-th excavation location before excavation. This indicates the threshold value set for the difference in groundwater level changes. Let Y represent the prediction efficiency requirement factor for the r-th excavation location, and let R represent the number of prediction depths and the number of excavation locations, respectively.

[0048] in, .

[0049] It should be noted that groundwater levels at each excavation location were collected using water level gauges before excavation. The threshold for the difference in groundwater level variation is a permissible value for groundwater level variation, jointly determined by multiple professionals based on the characteristics and capabilities of karst mountain areas. It can be 5 or 8, and no specific numerical limit is imposed here.

[0050] Preferably, the water level prediction efficiency formula is: In the formula, This indicates the results of the water level prediction efficiency analysis of the water level prediction model. This represents the excavation depth at the r-th excavation location when the groundwater level is predicted at the y-th predicted depth. This represents the depth difference between the predicted depths at the r-th excavation location, where T is the set prediction duration threshold. Indicates the prediction efficiency threshold. This represents the prediction duration of the groundwater level at the y-th predicted depth in the r-th excavation location.

[0051] It should be noted that the depth difference between the predicted depths at each excavation location is the average of the differences between adjacent predicted depths at each excavation location.

[0052] The prediction duration threshold is the maximum permitted prediction duration set by multiple professionals based on testing needs. It can be 1 minute or 3 minutes, and no specific numerical limit is set here.

[0053] The prediction efficiency threshold is a benchmark value for evaluating whether the prediction efficiency is good. When it is greater than or equal to the prediction efficiency threshold, it indicates that the prediction efficiency is good, and vice versa. It is set by multiple professionals according to the test requirements. It can be 2 or 6. There is no specific numerical limit here.

[0054] It should be noted that, Figure 1 and Figure 2 YES and NO represent water level prediction efficiency that is good and water level prediction efficiency that is poor, respectively.

[0055] S3. Use the water level prediction model to predict the water level in the underground excavation simulation, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute S4.

[0056] In a specific embodiment, the process of formulating preventive measures for underground excavation in designated karst mountain areas is as follows: S31, using water level prediction simulation data, screening out each target excavation location, and obtaining geological data, groundwater level parameters, underground environmental data and excavation parameters of each target excavation location as construction correlation data for each excavation location, obtaining construction records corresponding to each historical excavation in each karst mountain area from the data center, and predicting the set of dangerous phenomena for each target excavation location.

[0057] Preferably, the target excavation locations are selected by means of the following process: if the absolute value of the difference between the groundwater level at at least one predicted depth and the groundwater level at an adjacent predicted depth in a certain excavation location is greater than the threshold value of the difference in groundwater level changes, then the excavation location is taken as the target excavation location, thereby selecting the target excavation locations.

[0058] It should be noted that the geological data includes geological structures and topography. The groundwater level parameters are the groundwater levels at each predicted depth. The groundwater environmental data includes groundwater pH, groundwater conductivity, and dissolved oxygen.

[0059] Preferably, the specific process for predicting the set of hazardous phenomena at each target excavation location is as follows: geological data, groundwater level parameters, underground environmental data, and excavation parameters for each historical excavation location are obtained from the construction records corresponding to each historical excavation in each karst mountain area, which serve as construction correlation data for each historical excavation location. At the same time, the combination of hazardous phenomena, the hazard level of each hazardous phenomenon, prevention plan, and prevention level are obtained from the construction records corresponding to each historical excavation in each karst mountain area.

[0060] Calculate the similarity between the construction association data of each historical excavation location in each karst mountain area and the construction association data of each target excavation location. Use the historical excavation locations of each historical excavation location in each karst mountain area with a similarity greater than a preset similarity threshold as reference excavation locations for each target excavation location.

[0061] It should be noted that the similarity threshold and the threshold for groundwater level change difference are set similarly, and will not be elaborated here. The similarity between the construction association data of each historical excavation location corresponding to each historical excavation in each karst mountain area and the construction association data of each target excavation location is 1 minus the absolute value of the difference between the construction association data of each historical excavation location corresponding to each historical excavation in each karst mountain area and the construction association data of each target excavation location, and the ratio of the construction association data of each target excavation location.

[0062] Extract the hazardous phenomenon combinations corresponding to each reference excavation location for each target excavation location, integrate the hazardous phenomenon combinations corresponding to each reference excavation location for each target excavation location, and use the integrated result as the hazardous phenomenon set for each target excavation location.

[0063] S32. Based on the set of dangerous phenomena at each target excavation location, select a set of target prediction schemes for each target excavation location as preventive measures.

[0064] Preferably, the specific process of S32 is as follows: S321, based on the combination of hazardous phenomena corresponding to each reference excavation location, the hazard level of each hazardous phenomenon, the prevention plan and the prevention level, obtain the hazardous phenomena and the prevention level corresponding to each prevention plan in each target excavation location; at the same time, obtain the hazard level of each hazardous phenomenon corresponding to each prevention plan in each target excavation location, and cluster the hazard levels, and use the clustering result as the target hazard level of each hazardous phenomenon, thereby obtaining the target hazard level and the prevention level of each hazardous phenomenon corresponding to each prevention plan in each target excavation location.

[0065] It should be noted that dangerous phenomena include sudden water and mud inrush, ground subsidence, and instability of surrounding rock. Prevention measures include strengthening the support structure, setting up a drainage system, and grouting to seal groundwater.

[0066] Hazard Level Assessment: An assessment index system is established, specifically by selecting key indicators related to hazardous phenomena, such as karst ratio, uniaxial compressive strength of rock, groundwater pressure, and overburden thickness. Professionals assign corresponding weights based on their impact on the degree of hazard. For example: karst ratio weight 0.3, uniaxial compressive strength of rock weight 0.25, groundwater pressure weight 0.3, and overburden thickness weight 0.15. Karst ratio, uniaxial compressive strength of rock, groundwater pressure, and overburden thickness can be obtained using ground-penetrating radar, pressure testing machines, hydraulic gauges, and drilling equipment, respectively.

[0067] Taking karst rate as an example, karst rate is divided into four levels: less than 5%, 5%-15%, 15%-30%, and greater than 30%, corresponding to danger scores of 100, 80, 60, and 30, respectively. The danger score corresponding to the karst rate is obtained based on the karst rate, and then multiplied by the karst rate weight of 0.3 to obtain the danger score of the karst rate. Following the process of obtaining the danger score of the karst rate, the danger scores of the uniaxial compressive strength of the rock, the groundwater pressure, and the overburden thickness are obtained separately, and then accumulated to obtain the total danger score.

[0068] Assume there are four hazard levels: Level 1 has a total hazard score range of 0-30, Level 2 has a total hazard score range of 31-60, Level 3 has a total hazard score range of 61-80, and Level 4 has a total hazard score range of 81-100. By comparing the total hazard score with the total hazard score range of each level, the hazard level is obtained.

[0069] The process of obtaining prevention levels is similar to that of obtaining hazard levels, and will not be elaborated upon here.

[0070] S322. Compare the hazardous phenomena prevented by each prevention scheme at each target excavation location with the hazardous phenomena in the hazardous phenomenon set at each target excavation location. Obtain the number of hazardous phenomena prevented by each prevention scheme at each target excavation location that are the same as the hazardous phenomena in the hazardous phenomenon set. Take this number as the number of effective hazardous phenomena prevented by each prevention scheme at each target excavation location. Take each hazardous phenomenon that is the same as the hazardous phenomena in the hazardous phenomenon set as each effective hazardous phenomenon. Extract the target hazard level and prevention level of each effective hazardous phenomenon prevented by each prevention scheme at each target excavation location. Calculate the priority value of each prevention scheme at each target excavation location. Select the prevention scheme with the highest priority value as the first target prevention scheme for each target excavation location.

[0071] Preferably, the target hazard level and prevention level of each effective hazard phenomenon prevented by each prevention scheme at each target excavation location are calculated by weighted average. The calculation result is used as the target hazard level and prevention level of each effective hazard phenomenon prevented by each prevention scheme at each target excavation location. The number of effective hazard phenomena prevented by each prevention scheme at each target excavation location, the target hazard level of the effective hazard phenomenon, and the prevention level of the effective hazard phenomenon are respectively denoted as... , and Where f represents the number of each target excavation location, and p represents the number of each prevention plan, both f and p are positive integers. The priority value is calculated using the following formula: In the formula, This represents the priority value of the p-th prevention plan among the f-th target excavation locations. , These represent the weighting factors for the target hazard level and the prevention level, respectively.

[0072] in, , In the formula, This represents the maximum difference between the target hazard levels of each effective hazard phenomenon prevented by the p-th prevention scheme at the f-th target excavation location. This represents the minimum target hazard level among all the effective hazard phenomena prevented by the p-th prevention scheme at the f-th target excavation location. This represents the maximum difference between the prevention levels of each effective hazard phenomenon corresponding to the p-th prevention scheme at the f-th target excavation location. F represents the minimum prevention level among the prevention levels of each effective hazard phenomenon corresponding to the p-th prevention scheme in the f-th target excavation location, where F and P represent the number of target excavation locations and the number of prevention schemes, respectively.

[0073] S323. Remove the effective hazardous phenomena corresponding to the target prevention scheme from the hazardous phenomenon set of each target excavation location to obtain the second hazardous phenomenon set of each target excavation location. Then, compare the remaining hazardous phenomena corresponding to each prevention scheme in each target excavation location with the second hazardous phenomenon set of each target excavation location, and obtain the second target prevention scheme of each target excavation location according to the steps of S322.

[0074] S324. Repeat S322-S323 until all hazardous phenomena in the hazardous phenomenon set of each target excavation location are removed. Stop repeating and obtain the target prevention plan for each target excavation location. The set of target prevention plans is used as the target prediction plan set.

[0075] This application's embodiments provide a reference for subsequent excavation construction protection by formulating preventive measures, which is conducive to preparing protective equipment in advance, reducing the negative impact of groundwater level changes, improving construction response speed, and ensuring construction safety.

[0076] S4. When the water level prediction efficiency is poor, feedback is sent to the visualization display terminal; preventive measures for underground excavation in designated karst mountain areas are fed back to the visualization display terminal, and preventive prompts are provided. Example 2

[0077] See Figure 2 As shown, a groundwater level prediction system for karst mountain areas includes:

[0078] The model building and prediction module is used to build a water level prediction model, obtain a 3D model of a specified karst mountain area and an underground excavation plan for the specified karst mountain area, simulate underground excavation in the specified karst mountain area, and use the water level prediction model to predict the water level of the underground excavation simulation.

[0079] The model prediction and evaluation module is used for water level prediction based on underground excavation simulation. It analyzes the water level prediction efficiency of the water level prediction model. When the water level prediction efficiency is poor, the prediction execution module is executed. When the water level prediction efficiency is good, the prediction result processing module is executed.

[0080] The prediction result processing module is used to predict the water level of the underground excavation simulation using the water level prediction model, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute the prediction execution module.

[0081] The prediction execution module is used to feed back information to the visualization display terminal when the water level prediction efficiency is poor; it also feeds back preventive measures for underground excavation in designated karst mountain areas to the visualization display terminal and provides preventive prompts.

[0082] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A method for predicting groundwater levels in karst mountainous areas, characterized in that, Includes the following steps: S1. Construct a water level prediction model, and simultaneously obtain a 3D model of the specified karst mountain area and an underground excavation plan for the specified karst mountain area. Simulate underground excavation in the specified karst mountain area and use the water level prediction model to predict the water level in the underground excavation simulation. S2. Based on the simulation of underground excavation, predict the water level and analyze the water level prediction efficiency of the water level prediction model. If the water level prediction efficiency is poor, execute S4; if the water level prediction efficiency is good, execute S3. The efficiency of the water level prediction model was analyzed, and the specific analysis process is as follows: By using the groundwater level at each excavation location when the excavation depth reaches each predicted depth in the water level prediction simulation data, the prediction efficiency influence coefficient of each excavation location at each predicted depth is analyzed. Based on the initial time and end time of the groundwater level prediction at each predicted depth in each excavation location, the prediction duration of the groundwater level at each predicted depth in each excavation location is obtained. The prediction efficiency influence coefficient of each excavation location at each prediction depth, the prediction time of groundwater level at each prediction depth in each excavation location, and the excavation depth of each excavation location at each prediction depth when predicting groundwater level are input into the water level prediction efficiency formula, and the water level prediction efficiency analysis results of the water level prediction model are output. The results of the water level prediction efficiency analysis include YES and NO. When the result is YES, it indicates that the water level prediction efficiency is good, and vice versa. In this process, each excavation depth that requires groundwater prediction is taken as the prediction depth; the actual prediction depth at each excavation location is subtracted from the prediction depth at each prediction depth to obtain the excavation depth at each excavation location when the groundwater level is predicted at each prediction depth. The formula for water level prediction efficiency is as follows: In the formula, This indicates the results of the water level prediction efficiency analysis of the water level prediction model. This represents the excavation depth at the r-th excavation location when the groundwater level is predicted at the y-th predicted depth. This represents the depth difference between the predicted depths at the r-th excavation location, where T is the set prediction duration threshold. Indicates the prediction efficiency threshold. This represents the prediction duration of the groundwater level at the y-th predicted depth in the r-th excavation location; S3. Use the water level prediction model to predict the water level in the underground excavation simulation, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute S4. S4. When the water level prediction efficiency is poor, feedback is sent to the visualization display terminal; when the water level prediction efficiency is good, the preventive measures for underground excavation in designated karst mountain areas are fed back to the visualization display terminal, and preventive prompts are given.

2. The method for predicting groundwater levels in karst mountainous areas according to claim 1, characterized in that, The specific process for constructing the water level prediction model is as follows: S11. Through geological exploration, obtain geological and hydrogeological data of the designated karst mountain area, construct a conceptual model of the designated karst mountain area, and use numerical simulation software to divide the designated karst mountain area into grids and establish a three-dimensional groundwater flow numerical model. S12. Based on the three-dimensional groundwater flow numerical model, long-term observation data of a specified karst mountain area are obtained from the data center. Correlation analysis is performed on the long-term observation data of the specified karst mountain area. Statistical methods are used to establish a statistical model for groundwater level prediction. The parameters of the model are determined through regression analysis. S13. Optimize and correct the model using unused data from long-term observation data to obtain a water level prediction model.

3. The method for predicting groundwater levels in karst mountainous areas according to claim 1, characterized in that, The specific process of simulating underground excavation in a designated karst mountain area is as follows: Obtain the excavation locations and excavation parameters from the underground excavation plan, as well as the excavation depths at each location where groundwater prediction is required, and use these excavation depths as the prediction depths. In the three-dimensional model of the specified karst mountain area, excavation simulation is carried out sequentially according to each excavation location and its excavation parameters. When the underground excavation simulation is carried out, when the excavation depth of each excavation location reaches the predicted depth, the groundwater level at each excavation location is predicted by the water level prediction model. At the same time, the start time and end time of the prediction of the groundwater level at each excavation location and the predicted depth are obtained from the computer background records and used as the initial time and end time of the groundwater level prediction at each predicted depth in each excavation location. The actual predicted depth of the predicted groundwater level at each excavation location is obtained from the three-dimensional model of the specified karst mountain area. The actual predicted depth of the predicted groundwater level at each excavation location is subtracted from the actual predicted depth of the predicted groundwater level at each excavation location to obtain the excavation depth of each excavation location at the predicted groundwater level. The groundwater level at each excavation location when the excavation depth reaches each predicted depth, the excavation depth at each excavation location when the groundwater level at each predicted depth is predicted, and the initial and end times of the predicted groundwater level at each predicted depth in each excavation location are used as the water level prediction simulation data.

4. The method for predicting groundwater levels in karst mountainous areas according to claim 1, characterized in that, The analysis process for the prediction efficiency influence coefficient of each excavation location at each prediction depth is as follows: The groundwater level at each excavation location when the excavation depth reaches the predicted depth is denoted as . , where r represents the number of each excavation location, y represents the number of each predicted depth, and both r and y are positive integers; Using analytical formulas: The prediction efficiency influence coefficient of the r-th excavation location at the y-th prediction depth is obtained. In the formula, and Let represent the groundwater level at the r-th excavation location when the excavation depth reaches the (y-1)-th predicted depth and the groundwater level at the r-th predicted depth, respectively. This represents the groundwater level at the r-th excavation location before excavation. This indicates the threshold value set for the difference in groundwater level changes. Let Y represent the prediction efficiency requirement factor for the r-th excavation location, and let R represent the number of prediction depths and the number of excavation locations, respectively. in, .

5. The method for predicting groundwater levels in karst mountainous areas according to claim 3, characterized in that, The specific process for formulating preventive measures for underground excavation in designated karst mountain areas is as follows: S31. Using water level prediction simulation data, select the excavation locations of each target and obtain the geological data, groundwater level parameters, underground environmental data and excavation parameters of each target excavation location as construction correlation data for each excavation location. Obtain the construction records corresponding to each historical excavation in each karst mountain area from the data center and predict the set of dangerous phenomena for each target excavation location. S32. Based on the set of dangerous phenomena at each target excavation location, select a set of target prediction schemes for each target excavation location as preventive measures.

6. The method for predicting groundwater levels in karst mountainous areas according to claim 5, characterized in that, The specific process for predicting the set of hazardous phenomena at each target excavation location is as follows: Geological data, groundwater level parameters, underground environmental data, and excavation parameters for each historical excavation location were obtained from the construction records corresponding to each historical excavation in each karst mountain area. These data served as construction correlation data for each historical excavation location. At the same time, the combination of hazardous phenomena, the hazard level of each hazardous phenomenon, prevention plans, and prevention levels were obtained from the construction records corresponding to each historical excavation in each karst mountain area. Calculate the similarity between the construction association data of each historical excavation location in each karst mountain area and the construction association data of each target excavation location. Use the historical excavation locations of each historical excavation location in each karst mountain area with a similarity greater than a preset similarity threshold as reference excavation locations for each target excavation location. Extract the hazardous phenomenon combinations corresponding to each reference excavation location for each target excavation location, integrate the hazardous phenomenon combinations corresponding to each reference excavation location for each target excavation location, and use the integrated result as the hazardous phenomenon set for each target excavation location.

7. The method for predicting groundwater levels in karst mountainous areas according to claim 6, characterized in that, The specific process of S32 is as follows: S321. Based on the combination of hazardous phenomena corresponding to each reference excavation location at each target excavation location, the hazard level of each hazardous phenomenon, the prevention plan and the prevention level, obtain the hazardous phenomena and the prevention level corresponding to each prevention plan at each target excavation location; at the same time, obtain the hazard level of each hazardous phenomenon corresponding to each prevention plan at each target excavation location, and cluster the hazard levels. The clustering result is used as the target hazard level of each hazardous phenomenon, thereby obtaining the target hazard level and the prevention level of each hazardous phenomenon corresponding to each prevention plan at each target excavation location. S322. Compare the hazardous phenomena prevented by each prevention scheme at each target excavation location with the hazardous phenomena in the hazardous phenomenon set at each target excavation location. Obtain the number of hazardous phenomena prevented by each prevention scheme at each target excavation location that are the same as the hazardous phenomena in the hazardous phenomenon set. Take this number as the number of effective hazardous phenomena prevented by each prevention scheme at each target excavation location. Take each hazardous phenomenon that is the same as the hazardous phenomena in the hazardous phenomenon set as each effective hazardous phenomenon. Extract the target hazard level and prevention level of each effective hazardous phenomenon prevented by each prevention scheme at each target excavation location. Calculate the priority value of each prevention scheme at each target excavation location. Select the prevention scheme with the highest priority value as the first target prevention scheme for each target excavation location. S323. Remove the effective hazardous phenomena corresponding to the target prevention scheme from the hazardous phenomenon set of each target excavation location to obtain the second hazardous phenomenon set of each target excavation location. Then, compare the remaining hazardous phenomena corresponding to each prevention scheme in each target excavation location with the second hazardous phenomenon set of each target excavation location, and obtain the second target prevention scheme of each target excavation location according to the steps of S322. S324. Repeat S322-S323 until all hazardous phenomena in the hazardous phenomenon set of each target excavation location are removed. Stop repeating and obtain the target prevention plan for each target excavation location. The set of target prevention plans is used as the target prediction plan set.

8. A groundwater level prediction system for karst mountain areas, implementing the groundwater level prediction method for any one of claims 1-7, characterized in that, include: The model building and prediction module is used to build a water level prediction model, obtain a 3D model of a specified karst mountain area and an underground excavation plan for the specified karst mountain area, simulate underground excavation in the specified karst mountain area, and use the water level prediction model to predict the water level of the underground excavation simulation. The model prediction and evaluation module is used for water level prediction based on underground excavation simulation. It analyzes the water level prediction efficiency of the water level prediction model. When the water level prediction efficiency is poor, the prediction execution module is executed. When the water level prediction efficiency is good, the prediction result processing module is executed. The prediction result processing module is used to predict the water level of the underground excavation simulation using the water level prediction model, formulate preventive measures for underground excavation in designated karst mountainous areas, and then execute the prediction execution module. The prediction execution module is used to feed back information to the visualization display terminal when the water level prediction efficiency is poor; it also feeds back preventive measures for underground excavation in designated karst mountain areas to the visualization display terminal and provides preventive prompts.

Images