Drilling group water disaster prevention and control intelligent decision-making system and three-dimensional geologic model dynamic updating method

Abstract

The invention discloses an intelligent decision-making system for water disaster prevention and control of a drill hole group and a three-dimensional geologic model dynamic updating method, and relates to the technical field of coal mining safety. According to the system, multi-source geological and hydrological data are obtained through the borehole group cooperative detection module, a three-dimensional geological model is driven to be dynamically updated after data fusion processing, and a targeted water disaster prevention and control scheme is generated in combination with an intelligent decision algorithm. The method comprises four core steps of borehole group data acquisition, multi-source data fusion, three-dimensional geologic model dynamic updating and water disaster prevention and control intelligent decision making, and an improved GRU neural network and fracture inversion modeling technology is adopted to realize accurate pre-judgment of water disaster risks and dynamic optimization of a prevention and control scheme. The problems of geologic model staticization, decision hysteresis and the like in traditional water disaster prevention and control are solved, the intelligent level and reliability of water disaster prevention and control are improved, and the method is suitable for various mine water disaster prevention and control scenes.

Description

Technical Field

[0001] This invention relates to the field of coal mining safety technology, specifically to an intelligent decision-making system for preventing water hazards in borehole clusters and a method for dynamically updating three-dimensional geological models. Background Technology

[0002] Mine water hazards are one of the major disasters in coal mine production, seriously threatening personnel and property safety. Traditional water hazard prevention methods rely on static geological models and human experience-based decision-making, which have the following problems: First, geological models cannot reflect the deformation and fracture evolution of rock strata under the influence of mining in real time, leading to inaccurate prediction of water hazard risks; second, multi-source monitoring data lack effective fusion methods, resulting in low data utilization and susceptibility to interference; and third, decision-making schemes lack dynamic optimization mechanisms, making it difficult to adapt to real-time changes in hydrogeological conditions.

[0003] Existing borehole cluster technologies are mostly focused on single functions such as roof fracturing, fracture filling, or fault detection, such as microwave-based borehole cluster fracturing technology and borehole cluster trajectory comparison analysis methods. However, a comprehensive intelligent system for flood control has not yet been developed. Furthermore, existing flood warning models tend to focus on single signal analysis and do not fully integrate multi-dimensional geological and hydrological data from borehole cluster detection, resulting in insufficient warning accuracy and decision-making reliability. Therefore, there is an urgent need to develop an integrated flood control technology that combines borehole cluster collaborative detection, dynamic modeling, and intelligent decision-making. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent decision-making system for the prevention and control of water hazards in borehole clusters and a method for dynamically updating three-dimensional geological models, so as to solve the problems of static geological models, decision lag and low data utilization in the prior art.

[0005] The system structure is as follows: 1. Borehole Cluster Collaborative Detection Module: Utilizing a comb-style directional borehole cluster layout, combined with directional drilling technology, ensures accurate borehole trajectory. Multiple types of sensors are deployed within the borehole cluster to collect real-time data on water pressure, water volume, rock stress, fracture development, and borehole trajectory. This data is then uploaded to the data fusion module in real-time via a dual-transmission mode, providing fundamental data support for subsequent modeling and decision-making.

[0006] 2. Multi-source data fusion module: First, an intelligent comparative analysis algorithm for borehole trajectories is used to verify the consistency of trajectory data from different boreholes, eliminating abnormal data caused by borehole deviations. Then, normalization is applied to unify the data format, weights are assigned based on data reliability, and a weighted fusion algorithm is used to integrate geological, hydrological, and trajectory data to generate a standardized, highly reliable dataset, thus solving the problem of heterogeneity in multi-source data.

[0007] 3. Dynamic Update Module for 3D Geological Model: Based on initial geological survey data, a basic 3D geological model is constructed. Inversion modeling is performed using fracture data from borehole surveys to accurately reconstruct the distribution of rock fractures and the morphology of hydrological channels. When new monitoring data is received or update trigger conditions are met, the model's rock strata physical parameters, hydrological parameters, and spatial structure are adjusted through deviation analysis to achieve real-time dynamic updates, ensuring a high degree of consistency between the model and actual geological conditions.

[0008] 4. Intelligent Decision-Making Module: This module incorporates an improved GRU neural network early warning model trained on a large number of samples. Inputting dynamically updated model data, it extracts water hazard evolution characteristics and predicts the probability and risk level of water hazards. Combined with a prevention and control scheme database, it generates personalized solutions for different risk levels and geological conditions, including borehole grouting and sealing, drainage and pressure reduction, and fracture filling. The module dynamically optimizes scheme parameters based on model updates, achieving closed-loop management of water hazard prevention and control.

[0009] The process is as follows: 1. Borehole Layout and Data Acquisition: Based on the hydrogeological exploration results of the mining area, the layout location, depth, and spacing of the comb-type directional borehole group are designed to ensure that the borehole group can fully cover areas with potential water hazard risks. The sensor group is activated to continuously collect data, including real-time water pressure, water inflow, changes in rock stress, and actual borehole trajectory data. The data sampling frequency is dynamically adjusted according to the complexity of geological conditions.

[0010] 2. Multi-source data fusion processing: The collected multi-source data is preprocessed to remove noise interference and outliers. A borehole group trajectory intelligent comparison analysis method is used to verify the consistency of data from different boreholes and correct trajectory deviations. A weighted fusion strategy is adopted, allocating weights according to the reliability of data types, to fuse the processed geological, hydrological, and trajectory data into a standardized dataset, providing high-quality data input for model updates.

[0011] 3. Dynamic Update of 3D Geological Model: Utilizing rock strata parameters and fracture data from standardized datasets, and combining inversion modeling techniques, the initial 3D geological model is optimized to clarify the spatial distribution of hydrological channels. When abrupt changes in hydrological parameters, rock strata deformation, or the addition of new borehole data are detected, the model is dynamically updated. Through iterative calculations, model parameters are adjusted, enabling the model to reflect changes in geological and hydrological conditions in real time and improving the accuracy of flood risk prediction.

[0012] 4. Intelligent Decision-Making for Flood Control: The dynamically updated 3D geological model data is input into an improved GRU neural network early warning model, which outputs the flood risk level (high, medium, low) and potential hazard areas. The intelligent decision-making module calls upon the prevention and control scheme database, combining technologies such as borehole group fracturing and filling, to generate targeted prevention and control schemes, including optimized borehole layout, grouting material selection, grouting pressure parameters, and drainage system design. Based on subsequent model updates and the effectiveness of scheme implementation, the prevention and control strategy is continuously optimized to ensure the effectiveness and timeliness of flood control.

[0013] This invention relies on borehole cluster collaborative detection and multi-source data fusion technology. Leveraging the spatial coverage characteristics of borehole clusters, it enhances the comprehensiveness and reliability of data acquisition, effectively overcoming the limitations of single data sources. Through a dynamic update mechanism of the three-dimensional geological model, it maps the deformation of rock strata and the evolution of hydrological conditions under mining influence in real time, breaking through the application bottleneck of traditional static models and providing solid technical support for accurate prediction of water hazard risks. Combining an improved GRU neural network, intelligent decision-making algorithm, and borehole cluster prevention technology, it generates personalized and dynamically optimized water hazard prevention solutions, significantly improving the intelligence level and decision-making efficiency of water hazard prevention and reducing the risk of manual intervention. Simultaneously, the system has strong compatibility, adapting to the hydrogeological conditions of different types of mines by adjusting the borehole cluster layout and model parameters, possessing broad practicality and promotional value. Attached Figure Description

[0014] Figure 1 This is a flowchart of the module relationships of the intelligent decision-making system for borehole group water hazard prevention and control of the present invention; Figure 2 is a flowchart of the dynamic updating of the three-dimensional geological model and the decision-making process for water hazard prevention in this invention; Figure 3 is an interactive flowchart of the core module of the intelligent decision-making system for water hazard prevention and control in borehole clusters; Figure 1 This diagram is a flowchart of the module relationships in the intelligent decision-making system for borehole cluster water hazard prevention. It clearly shows the hierarchical relationship between the borehole cluster collaborative detection module, the multi-source data fusion module, the 3D geological model dynamic update module, and the intelligent decision-making module, as well as the information flow path of "real-time data - fused data - updated model data". The feedback arrows from the prevention and control plan to the borehole cluster collaborative detection module demonstrate the closed-loop management logic of the system.

[0015] Figure 2 The flowchart for dynamic updating of 3D geological models and decision-making for water hazard prevention demonstrates the entire process from acquiring initial geological survey data and constructing an initial 3D geological model, through borehole group data acquisition and multi-source data fusion processing, to determining whether the update trigger conditions are met, thus realizing dynamic updating of the 3D geological model (if the conditions are met) or continuous monitoring (if the conditions are not met), and finally outputting the updated model and completing intelligent decision-making for water hazard prevention.

[0016] Figure 3 This system establishes the data interaction and control logic relationships between the core modules during system operation. It reflects the collaborative exploration of borehole groups to acquire multi-source geological and hydrological data. After multi-source data fusion processing, it drives the dynamic updating of the three-dimensional geological model. The intelligent decision-making module generates water hazard prevention and control plans based on the updated model, forming a closed-loop operation process in which data acquisition, model updating, and decision feedback are interconnected.

Claims

1. A smart decision-making system for preventing and controlling water hazards in borehole clusters, characterized in that, include: (1) Borehole group collaborative detection module: It consists of a directional borehole group, a sensor group and a data transmission unit. The directional borehole group is arranged in a comb-like layout. The sensor group is used to collect geological and rock strata parameters, hydrological dynamic parameters and borehole trajectory parameters. (2) Multi-source data fusion module: Receives data transmitted by the borehole group collaborative detection module, removes abnormal data through trajectory intelligent comparison analysis algorithm, integrates multi-dimensional data using weighted fusion strategy, and outputs standardized dataset; (3) Dynamic update module of three-dimensional geological model: Based on standardized datasets and combined with rock fracture inversion modeling technology, an initial three-dimensional geological model is constructed, and the model parameters are dynamically adjusted and the spatial morphology is corrected through real-time data feedback; (4) Intelligent decision-making module: It has an improved GRU neural network early warning model and prevention and control scheme database. Based on the updated three-dimensional geological model, it analyzes the risk level of water hazards, generates personalized water hazard prevention and control schemes and dynamically optimizes them.

2. The system according to claim 1, characterized in that, The sensor group includes a water pressure sensor, a flow sensor, a rock stress sensor, and a borehole trajectory measurement sensor. The data transmission unit adopts a dual transmission mode combining wireless and wired connections to ensure the stability of data transmission.

3. The system according to claim 1, characterized in that, The improved GRU neural network early warning model is generated through training on sample water hazard data. It incorporates rock fracture parameters as feature inputs, has the ability to remember the evolution of water hazards over a long period of time, and can output the predicted probability of different water hazard types.

4. A method for dynamically updating a three-dimensional geological model based on the system described in any one of claims 1-3, characterized in that, Includes the following steps: S1: Drilling group layout and data acquisition. The layout of the comb-type directional drilling group is designed according to the hydrogeological conditions of the mining area. Geological, hydrological and trajectory data are collected in real time through the sensor group and continuously transmitted to the multi-source data fusion module. S2: Multi-source data fusion processing, using the borehole group trajectory intelligent comparison analysis method to verify the validity of the data, normalizing the rock layer parameters, hydrological parameters and trajectory parameters, and generating a standardized dataset through a weighted fusion algorithm; S3: Dynamic updating of the three-dimensional geological model. A basic three-dimensional model is constructed based on the initial geological data. The fracture information in the standardized dataset is used for inversion modeling. Through the deviation analysis between real-time data and the model, the rock structure, hydrological channels and fracture distribution parameters of the model are dynamically adjusted. S4: Intelligent decision-making for water hazard prevention. The updated 3D geological model data is input into the improved GRU neural network early warning model to determine the water hazard risk level and potential hazard areas. The prevention and control scheme database is called up, and combined with technologies such as borehole group fracturing and filling, a targeted scheme including borehole optimization layout and grouting parameter design is generated.

5. The method according to claim 4, characterized in that, The triggering conditions for dynamic updating of the three-dimensional geological model in step S3 include: supplementation of new borehole data, sudden changes in hydrological parameters, model prediction deviation exceeding the threshold, and deformation of rock strata caused by mining.

6. The method according to claim 4, characterized in that, In step S4, the intelligent decision-making module can dynamically adjust the prevention and control strategy according to the water hazard risk level. In high-risk areas, a borehole group collaborative grouting and sealing scheme is adopted, while in medium- and low-risk areas, a scheme combining borehole drainage and pressure reduction with crack monitoring is adopted.

7. The method according to claim 4, characterized in that, In step S2, the weight coefficients of the weighted fusion algorithm are determined based on the reliability of different data types. The weight of borehole trajectory data is higher than that of indirect monitoring data, and the weight of real-time hydrological data is dynamically adjusted over time.

Images