Mine electrical method integrated exploration system based on matrix electrode and use method

Abstract

The invention discloses a mine electrical method integrated exploration system based on matrix electrodes and a use method, and belongs to the technical field of mine water disaster prevention and control, the system comprises a multi-core cable, matrix electrodes, an on-site host and the like, the host integrates an intelligent data analysis module, a data visualization module and the like, and the electrodes are arranged in a double-row matrix mode. When in use, exploration is completed through the steps of system deployment and self-inspection, parameter configuration, multi-frequency data acquisition, real-time analysis and anomaly recognition and remote collaboration and decision support. The system has the beneficial effects that an intelligent algorithm is built in the field host to realize data real-time preprocessing and low-resistance abnormity identification, so that the processing speed is improved, and minute-level response of the water disaster risk is realized; the data visualization module converts the data into a three-dimensional model, and the geological condition is visually presented; and the automatic control module automatically adjusts pole distribution parameters, remote fault diagnosis is realized by combining a remote monitoring function, and the reliability and the safety of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of mine water hazard prevention and control, and in particular to a mine electrical resistivity exploration system based on matrix electrodes and its usage method. Background Technology

[0002] In the field of mine geological exploration, mine water hazard prevention has always been a core aspect of ensuring safe production. Mine electrical resistivity tomography (EPT), as an important means of detecting geological structures such as aquifers and faults, effectively identifies low-resistivity anomalies by transmitting electrical signals underground and receiving reflected waves, providing crucial data support for water hazard early warning. Traditional EPT systems typically employ a single-channel data acquisition and offline analysis mode. Their technical architecture relies heavily on external computers for data processing, and the data presentation is limited to two-dimensional profiles, making it difficult to meet the efficient and precise exploration needs of modern mines.

[0003] Common mine electrical resistivity tomography (EPT) systems generally suffer from the following technical bottlenecks: First, the data acquisition and analysis processes are fragmented. Raw electrical resistivity data must be transmitted to a ground computer via cables for processing. Limited transmission bandwidth and computing resources lead to significant delays in data processing, making real-time risk warnings impossible. Second, two-dimensional data display methods cannot intuitively reflect the three-dimensional spatial characteristics of geological structures. Operators need to infer the morphology of anomaly zones based on experience, which can easily lead to misjudgments. Third, electrode placement and parameter configuration rely on manual experience. Under complex geological conditions, repeated adjustments to electrode spacing and power supply frequency are required, resulting in low operational efficiency and the introduction of human error. Fourth, system operation and maintenance are highly dependent on on-site personnel, lacking remote monitoring and fault diagnosis capabilities. When faults such as cable breaks or poor contact occur, exploration must be interrupted for manual troubleshooting, severely restricting the system's continuous operation capability and failing to meet the requirements of mine water hazard prevention applications. Therefore, a mine electrical resistivity tomography integrated exploration system based on matrix electrodes and its usage method are proposed. Summary of the Invention

[0004] This invention provides the following technical solution: a mine electrical resistivity exploration system based on matrix electrodes, comprising: The multi-core cable includes a connecting clamp, a matrix electrode, a field host, and a remote terminal. One end of the multi-core cable is connected to the field host, and the other end is connected to the matrix electrode via the connecting clamp. The multi-core cable is used for the transmission of electrical signals. The connecting clamp is used to realize the physical and electrical connection between the multi-core cable and the matrix electrode to ensure the stable transmission of electrical signals. The matrix electrode is arranged in a matrix in the top and bottom plates of the mine roadway. The field host integrates an intelligent data analysis module, a data visualization module, a wireless communication module and an automatic control module. The intelligent data analysis module is used for real-time preprocessing, 3D geological modeling, low-resistivity anomaly identification, and multi-source data fusion analysis of mine electrical resistivity exploration data. The module incorporates machine learning algorithms and a fault diagnosis submodule. The fault diagnosis submodule automatically detects potential water hazard risks using machine learning algorithms and automatically locates faults such as cable breaks and poor electrode contact, generating analysis reports with repair suggestions. This built-in intelligent data analysis module changes the inefficient mode of traditional mine electrical resistivity exploration systems that require data transmission to external computers for analysis. Furthermore, through the integrated machine learning algorithms and fault diagnosis submodule, real-time multi-level filtering preprocessing, 3D resistivity inversion, and low-resistivity anomaly identification of electrical resistivity data can be performed on-site. Compared to traditional offline analysis, this significantly improves data processing speed. When the intelligent algorithm determines, through a support vector machine classifier, that the resistivity value is below a preset threshold and the morphology matches the characteristics of a water-bearing body, the system immediately triggers an audible and visual alarm and pushes the abnormal coordinates to a remote terminal via a wireless communication module, achieving minute-level response to mine water hazard risks and providing real-time decision support for on-site construction personnel to suspend excavation and investigate potential hazards. The data visualization module is used to convert electrical resistivity tomography data into 3D geological models, contour maps, and pseudo-color profile maps. The wireless communication module is used to realize real-time data interaction between the field host and the remote terminal. The automation control module is used to automatically adjust the matrix electrode arrangement parameters and optimize the data acquisition strategy according to the preset exploration plan. Through the data visualization module and the automation control module, the data visualization module can convert electrical resistivity tomography data into interactive 3D geological models, supporting users to rotate, section, and query the attributes of the models. Compared with traditional 2D displays, it more intuitively presents the spatial distribution of aquifers, faults, and other structures, making the basis for water hazard prevention and control more accurate.

[0005] This invention provides a method for using a matrix electrode-based integrated electrical resistivity exploration system for mines. The method includes the following steps: S1 System Deployment and Self-Check: A matrix electrode array is arranged at a preset interval on the top and bottom plates of the mine roadway, and a multi-core cable is connected to the matrix electrode through a connecting clamp. After the on-site host is started, the automation control module automatically performs equipment self-test to verify cable conductivity, electrode contact resistance and inclinometer status. When poor contact and abnormal inclinometer are detected, the fault diagnosis submodule generates a repair suggestion report containing the coordinates of the fault point. S2 parameter configuration and task loading: The XML format task script is loaded through remote terminal and local interface. After the exploration plan parser parses the script, it generates electrode deployment instructions. The user can dynamically adjust the electrode spacing, arrangement and power supply electrode spacing according to geological conditions, and configure the power supply frequency combination and data quality threshold. At the same time, the data acquisition strategy optimization engine combines geological noise characteristics to automatically determine the multi-frequency superposition acquisition mode in high-resistivity areas. S3 Multi-Frequency Data Acquisition: The automated control module sends electrical signals to the matrix electrodes in a preset frequency band sequence, synchronously triggering the data acquisition process. At the same time, the intelligent data analysis module uses a multi-level filtering algorithm to preprocess the raw electrical resistivity data. S4 Real-time Analysis and Anomaly Detection: The intelligent data analysis module constructs a three-dimensional resistivity inversion model based on the finite difference method, and combines geostatistical methods to locate low resistivity anomaly areas. At the same time, the machine learning algorithm classifies water hazard risk levels through a support vector machine classifier. If the resistivity value is lower than the preset threshold and the morphology is consistent with the characteristics of a water-bearing body, an audible and visual alarm is triggered and the abnormal coordinates are pushed to a remote terminal through the wireless communication module. S5 Remote Collaboration and Decision Support: The system displays the operating status of the on-site host, data quality heatmap, and exploration progress bar in real time via a remote terminal. Combined with the real-time monitoring function of the remote terminal, it enables remote viewing of equipment operating status and data quality heatmap, and allows for remote diagnosis and repair of faults such as cable breaks or poor electrode contact, thereby improving system reliability and construction safety.

[0006] Preferably, the connecting clamp adopts a combination structure of a waterproof and corrosion-resistant engineering plastic shell and copper conductive terminals. The inner side of the engineering plastic shell is provided with an elastic sealing ring, the surface of the conductive terminals is gold-plated, and both ends of the connecting clamp are respectively provided with anti-misinsertion interfaces, which effectively resists corrosion from the humid environment of the mine, reduces signal transmission loss, and the design of elastic sealing ring and anti-misinsertion interface can prevent poor contact caused by water seepage, improve connection reliability, and reduce the risk of data interruption caused by cable failure.

[0007] Preferably, each row of the matrix electrode contains 16-32 electrode units. The row spacing and column spacing of the matrix electrode are automatically calibrated by an automated control module. Each electrode unit of the matrix electrode is equipped with a pressure sensor and an inclinometer. Each row of electrodes in the matrix electrode is equipped with an independent power supply circuit, which can monitor the stability of the electrode installation in real time and avoid data distortion caused by loosening. The independent power supply circuit supports hot-swapping of faulty units. The failure of a single row of electrodes does not affect the overall operation, which significantly improves the system's fault tolerance and continuous operation capability.

[0008] Preferably, the intelligent data analysis module preprocesses the raw electrical resistivity data using a multi-level filtering algorithm. It also constructs a three-dimensional resistivity inversion model using the finite difference method and combines geostatistical methods to spatially locate and characterize low-resistivity anomaly zones. The machine learning algorithm is built upon a support vector machine classifier and trained using historical exploration data to classify water hazard risk levels. The fault diagnosis submodule integrates an expert knowledge base and improves fault diagnosis accuracy through historical case training, accurately distinguishing between cable breaks and poor contact types. Furthermore, the combination of geostatistical methods reduces the misjudgment rate of low-resistivity anomaly zones, providing a more reliable quantitative basis for water hazard mitigation solutions.

[0009] Preferably, the data visualization module constructs an interactive three-dimensional geological model based on WebGL technology. The module employs a double-buffered rendering mechanism to dynamically update contour maps and pseudo-color profile maps, and can overlay and display measuring point coordinates, resistivity values, and anomaly zone boundaries. The module includes an internal visualization report generation engine that exports PDF documents containing the three-dimensional coordinates of the anomaly zone, resistivity threshold ranges, and risk assessment conclusions. It automatically adds exploration timestamps and equipment serial numbers, achieving standardized management of the results documents. The integrated display of three-dimensional coordinates and risk conclusions facilitates technical briefings and archiving, reducing the time-consuming process of manually organizing information.

[0010] Preferably, the automated control module incorporates an exploration plan parser based on a rule engine, an electrode placement parameter adjustment mechanism, and a data acquisition strategy optimization engine. The exploration plan parser parses XML-formatted task scripts and automatically generates electrode placement instructions. The electrode placement parameter adjustment mechanism employs a closed-loop control algorithm to dynamically optimize electrode spacing, arrangement, and supply electrode spacing based on geological conditions. The data acquisition strategy optimization engine combines Shannon's sampling theorem with geological noise characteristics to automatically determine the optimal sampling frequency and integration time. In high-resistivity areas, it enables a multi-frequency overlay acquisition mode, allowing for dynamic adjustment of the sampling frequency based on real-time geological noise. The multi-frequency overlay mode in high-resistivity areas improves the data signal-to-noise ratio. Furthermore, the exploration plan parser is compatible with XML scripts, enabling standardized and rapid reuse of parameter configurations.

[0011] Preferably, in step S1, the automation control module verifies the cable continuity by sending a low-frequency pulse signal to the multi-core cable. When the signal attenuation exceeds a preset threshold, the fault diagnosis submodule locates the electrode with poor contact based on the electrode potential difference distribution pattern. At the same time, it compares the inclinometer data with the preset burial angle. If the deviation exceeds the allowable range, it generates a visual repair report containing the electrode number, actual inclinometer angle, and calibration suggestions. This allows for the rapid location of the poor contact point through the low-frequency pulse signal attenuation threshold. Combined with the inclinometer data, a visual repair report is generated, shortening the equipment self-inspection time and reducing the workload of manual troubleshooting.

[0012] Preferably, in step S3, the automated control module supplies power to the matrix electrodes according to a preset sequence from low frequency to high frequency. When power is supplied to each frequency band, multi-channel data acquisition is triggered synchronously. The intelligent data analysis module performs multi-level filtering processing on the raw data. That is, first, the power frequency interference is removed by bandpass filtering, then random noise is removed by wavelet transform, and finally the data curve is smoothed by sliding window algorithm. The cascaded denoising of bandpass filtering and wavelet transform improves the depth of power frequency interference suppression. Furthermore, the sliding window smoothing algorithm eliminates data spikes, providing higher quality basic data for subsequent three-dimensional inversion.

[0013] Preferably, in step S4, when the intelligent data analysis module performs three-dimensional resistivity inversion on the preprocessed data using the finite difference method, the exploration area is first discretized into a grid model. The resistivity value of each grid point is optimized through iterative calculation. Then, combined with the Kriging interpolation method of geostatistics, the spatial morphology of the low-resistivity anomaly area is characterized. At the same time, the machine learning algorithm, based on the trained support vector machine model, divides the anomaly area into three risk levels: safe, warning, and dangerous. When it is determined to be dangerous, the on-site host immediately activates the audible and visual alarm and marks the three-dimensional coordinates of the anomaly area. The accuracy of the boundary characterization of the low-resistivity anomaly area is improved through iterative calculation of the grid model, and the response time of the three-level risk classification of the support vector machine is shortened. After the danger level is determined, the alarm triggering and coordinate marking are completed simultaneously.

[0014] In summary, compared with the prior art, the present invention provides a mine electrical resistivity tomography integrated exploration system and its application method based on matrix electrodes, which has the following beneficial effects: 1. This invention, by integrating an intelligent data analysis module into the on-site host, changes the inefficient mode of traditional mine electrical resistivity exploration systems that require data transmission to an external computer for analysis. Furthermore, through the machine learning algorithm and fault diagnosis sub-module integrated within the intelligent data analysis module, real-time multi-level filtering preprocessing, three-dimensional resistivity inversion, and low-resistivity anomaly zone identification of electrical resistivity data can be performed on-site. Compared with traditional offline analysis, this improves data processing speed. Moreover, when the intelligent algorithm determines through the support vector machine classifier that the resistivity value is lower than a preset threshold and the morphology conforms to the characteristics of a water-bearing body, the system immediately triggers an audible and visual alarm and pushes the abnormal coordinates to a remote terminal via the wireless communication module, achieving minute-level response to mine water hazard risks and providing real-time decision support for on-site construction personnel to suspend excavation and investigate potential hazards. 2. This invention utilizes a data visualization module and an automated control module. The data visualization module can transform electrical resistivity data into an interactive three-dimensional geological model, allowing users to rotate, section, and query the model's attributes. Compared to traditional two-dimensional displays, it more intuitively presents the spatial distribution of aquifers, faults, and other structures, making water hazard prevention more precise. Furthermore, the automated control module can automatically analyze geological parameters and dynamically adjust the electrode spacing, arrangement, and power supply frequency sequence of the matrix electrodes. This enables intelligent optimization of exploration parameters without manual intervention. Combined with the real-time monitoring function of a remote terminal, it allows for remote viewing of equipment operating status and data quality heatmaps, and remote diagnosis and repair of faults such as cable breaks or poor electrode contact, improving system reliability and construction safety. Attached Figure Description

[0015] Figure 1 This is a system structure block diagram of the present invention.

[0016] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0017] 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.

[0018] Please see Figure 1 This invention provides a technical solution: a mine electrical resistivity tomography integrated exploration system based on matrix electrodes, comprising: Multi-core cable, connecting clamp, matrix electrode, field host and remote terminal. One end of the multi-core cable is connected to the field host, and the other end is connected to the matrix electrode through the connecting clamp. The multi-core cable is used for the transmission of electrical signals. The connecting clamp is used to realize the physical and electrical connection between the multi-core cable and the matrix electrode, ensuring the stable transmission of electrical signals. The connecting clamp adopts a combination structure of waterproof and corrosion-resistant engineering plastic shell and copper conductive terminals. The inner side of the engineering plastic shell is provided with an elastic sealing ring, and the surface of the conductive terminal is gold-plated. Both ends of the connecting clamp are provided with anti-misinsertion interfaces. The matrix electrodes are arranged in a row matrix in the top and bottom plates of the mine roadway. Each row of the matrix electrodes contains 16-32 electrode units. The row spacing and column spacing of the matrix electrodes are automatically calibrated by an automated control module. Each electrode unit of the matrix electrodes is equipped with a pressure sensor and an inclinometer. Each row of the matrix electrodes is equipped with an independent power supply circuit. The field unit integrates an intelligent data analysis module, a data visualization module, a wireless communication module, and an automation control module. The intelligent data analysis module is used for real-time preprocessing, 3D geological modeling, low-resistivity anomaly identification, and multi-source data fusion analysis of mine electrical resistivity exploration data. Internally, the intelligent data analysis module includes machine learning algorithms and a fault diagnosis submodule. The fault diagnosis submodule uses machine learning algorithms to automatically detect potential water hazard risks and automatically locate faults such as cable breaks and poor electrode contact, generating analysis reports with repair suggestions. The intelligent data analysis module preprocesses the raw electrical resistivity data using a multi-level filtering algorithm. It constructs a 3D resistivity inversion model using the finite difference method and combines geostatistical methods to spatially locate and characterize low-resistivity anomalies. The machine learning algorithm is based on a support vector machine classifier and trained using historical exploration data to classify water hazard risk levels. The fault diagnosis submodule integrates an expert knowledge base. The data visualization module is used to convert electrical resistivity exploration data into 3D geological models, contour maps, and pseudo-color profile maps. The data visualization module builds an interactive 3D geological model based on WebGL technology. The data visualization module adopts a double-buffered rendering mechanism to realize the dynamic updating of contour maps and pseudo-color profile maps, and can overlay and display the coordinates of measuring points, resistivity values, and boundaries of anomaly areas. The data visualization module has an internal visualization report generation engine, which is used to export PDF documents containing the 3D coordinates of anomaly areas, resistivity threshold ranges, and risk assessment conclusions, and automatically adds exploration timestamps and equipment serial numbers. The wireless communication module enables real-time data interaction between the field host and the remote terminal. The automation control module automatically adjusts the matrix electrode layout parameters and optimizes the data acquisition strategy according to the preset exploration plan. The automation control module has a built-in exploration plan parser based on a rule engine, an electrode layout parameter adjustment mechanism, and a data acquisition strategy optimization engine. The exploration plan parser parses the XML-formatted task script and automatically generates electrode layout instructions. The electrode layout parameter adjustment mechanism adopts a closed-loop control algorithm to dynamically optimize the electrode spacing, arrangement, and supply electrode spacing according to geological conditions. The data acquisition strategy optimization engine combines Shannon's sampling theorem and geological noise characteristics to automatically determine the optimal sampling frequency and integration time, and enables a multi-frequency superposition acquisition mode in high-resistivity areas.

[0019] Please see Figure 2 This invention provides a method for using a matrix electrode-based integrated electrical resistivity tomography (EPT) system for mine exploration. The method includes the following steps: S1 System Deployment and Self-Check: A matrix electrode array is arranged at a preset interval on the top and bottom plates of the mine roadway, and a multi-core cable is connected to the matrix electrode through a connecting clamp. After the on-site host is started, the automation control module automatically performs equipment self-test to verify cable conductivity, electrode contact resistance and inclinometer status. When poor contact and abnormal inclinometer are detected, the fault diagnosis submodule generates a repair suggestion report containing the coordinates of the fault point. The automation control module verifies the cable's conductivity by sending low-frequency pulse signals to the multi-core cable. When the signal attenuation exceeds a preset threshold, the fault diagnosis submodule locates the poorly contacting electrode based on the electrode potential difference distribution pattern. At the same time, it compares the inclinometer data with the preset burial angle. If the deviation exceeds the allowable range, it generates a visual repair report containing the electrode number, actual inclinometer angle, and calibration suggestions. The specific implementation process of the above method is as follows: When the system initiates its self-test program, the automated control module first generates a low-frequency electrical signal of a specific waveform using its built-in pulse generator. This signal, after power amplification, is injected into the starting end of the multi-core cable, forming a detection wave propagating longitudinally along the cable. The pulse parameter settings balance anti-interference capability and transmission loss characteristics, employing a stepped energy increment mode. Initial detection is performed using a low-amplitude signal first; if no abnormalities are detected, the transmission intensity is automatically increased. Simultaneously with signal injection, the module starts a high-precision timer to record the transmission timestamp, providing a benchmark for subsequent attenuation analysis. During the signal transmission phase, the control module monitors each channel of the cable in real time through distributed sampling nodes. Each sampling node is equipped with an independent analog-to-digital converter to capture voltage waveform changes at microsecond intervals. The system employs time-division multiplexing technology to rapidly switch monitoring channels during pulse transmission intervals, ensuring that the time-domain response characteristics of the entire cable path are completely recorded. When an abnormal attenuation of the echo amplitude of a certain channel compared to the transmission reference value is detected, the module immediately marks the channel and triggers a deep analysis program. During fault location, the fault diagnosis submodule invokes the potential difference analysis engine. This engine constructs a potential distribution topology map along the cable by analyzing the voltage gradient changes between adjacent healthy electrodes. When a poor connection occurs, characteristic potential jumps appear before and after the fault point. The module compares the theoretical potential decay curve with the measured curve to pinpoint the abnormal area. Combining the spatial coordinate data of the matrix electrode array, the system maps the potential abrupt change point to a specific electrode unit, generating a three-dimensional coordinate location result including lateral position and longitudinal depth. During the tilt calibration process, the module synchronously reads data from the MEMS tilt sensor built into the target electrode. The preset embedding angle parameters of the electrode are obtained via a wireless communication module; these parameters were calibrated during system deployment based on the roof and floor morphology of the tunnel. The diagnostic submodule employs a tiered comparison mechanism, first performing a coarse-grained angle range judgment; if the angle exceeds the normal operating range, a fine calibration process is initiated. The calibration suggestion generation engine, combined with electrode pressure sensor data, comprehensively assesses the current installation stability. For angle deviations caused by rock deformation, the system provides anchor bolt reinforcement suggestions; for installation errors, it generates fine-tuning angle values ​​and rotation direction commands. The final visualized repair report includes multi-dimensional information: a 3D geological model serves as the base map, highlighting the faulty cable segment and associated electrode locations with bright color blocks; data tables detail the faulty electrode number, measured tilt angle, recommended calibration angle, and pressure sensor status; for poor contact issues, the report includes instructions on the cable connector disassembly sequence and cleaning procedures for gold-plated terminals. All diagnostic conclusions and operational recommendations are rendered into an interactive interface using the WebGL engine of the data visualization module. On-site technicians can view the fault distribution from different perspectives via touch controls or export it as a standardized maintenance work order in PDF format. S2 parameter configuration and task loading: The XML format task script is loaded through a remote terminal and a local interface. After the exploration plan parser parses the script, it generates electrode deployment instructions. The user dynamically adjusts the electrode spacing, arrangement, and power supply electrode spacing according to geological conditions, and configures the power supply frequency combination and data quality threshold. At the same time, the data acquisition strategy optimization engine combines geological noise characteristics to automatically determine the multi-frequency superposition acquisition mode in high-resistivity areas. The specific implementation process of the above method is as follows: First, the user initiates the exploration plan configuration program via the on-site host touchscreen or the web interface of a remote terminal. The system then presents an interactive task template library. This library contains various preset geological scenario configuration schemes, such as fault fracture zone detection and aquifer tracking. Users can select a similar template for quick initialization or create a completely new task script. Once a template is selected, the system automatically generates an XML skeleton file containing the exploration area boundary, electrode baseline parameters, and power supply strategy framework. During the XML script editing phase, the system provides a visual programming interface. The left side displays a tree-like directory of functional modules, covering parameter categories such as electrode layout, power supply sequence, and data quality constraints; the middle section is the XML code editing area, supporting syntax highlighting and intelligent tag completion; the right side is a 3D geological model preview window, reflecting in real-time the impact of parameter modifications on the exploration area. Users can adjust the coverage area of ​​the electrode array by dragging and dropping; the system automatically calculates the required cable length and number of connection clamps, and dynamically updates equipment requirements in the bill of materials area. After the user completes the writing or modification of the XML script, clicking the "Parse and Verify" button triggers the exploration plan parser. This parser employs a layered verification mechanism: first, it performs a syntax check, confirming the legality of nested tags by traversing the DOM tree; then, it performs semantic verification, comparing whether the electrode coordinates exceed the tunnel cross-sectional dimensions, whether the power supply voltage exceeds safety thresholds, etc.; finally, it performs logical conflict detection, such as avoiding coupling interference caused by excessively small spacing between adjacent electrodes. If an anomaly is detected, the parser marks the error type and suggested corrections in the corresponding line of code. After verification, the parser maps the XML nodes into an executable instruction set. The electrode placement instructions include the absolute coordinates, rotation angle, and burial depth requirements for each row of electrodes. These instructions are sent to the microcontroller unit of the matrix electrodes via a wireless communication module. Upon receiving the instructions, the electrode unit drives the electric push rod mechanism to adjust the spacing, and the built-in laser ranging module provides real-time feedback on the actual position, forming a closed-loop control until millimeter-level positioning accuracy is achieved. During the dynamic parameter adjustment phase, the system provides an intelligent interface adapted to geological conditions. The interface overlays and displays multi-source data, including borehole core images and geophysical interpretation results. Users can delineate areas of lithological change through touch controls. For different lithologies such as sandstone and mudstone, the system has a built-in library of recommended parameter combinations. For example, it automatically reduces the electrode spacing to 0.5 meters in fractured zones and restores it to 2 meters in intact rock formations. The electrode spacing adjustment supports visual simulation; as the user drags the power supply electrode position, the system calculates a real-time thermal map of the current field distribution to assist in selecting the optimal electrode spacing combination. The data acquisition strategy optimization engine's workflow comprises three core modules: a geological noise monitoring unit that analyzes the signal-to-noise ratio fluctuation characteristics of historical data in real time and establishes a noise frequency distribution map; a spectrum optimizer that selects three non-interfering frequency bands to form a power supply sequence based on the noise map, ensuring that the main frequency band avoids strong noise regions; and a superposition mode decision-maker that automatically enables multi-frequency synchronous acquisition for high-impedance areas, achieving parallel transmission of multi-frequency signals through orthogonal frequency division multiplexing technology, and separating the data of each frequency band at the receiving end through a digital filter bank. The final multi-frequency acquisition scheme includes a frequency band switching timing table, integration time configuration, and data fusion rules. The system packages the configuration parameters into an encrypted instruction set and sends it to the waveform generator of each electrode unit via the fieldbus. During the acquisition process, the data quality monitoring module continuously evaluates the validity of the data in each frequency band. When a frequency band is affected by sudden interference, it automatically extends the integration time of that frequency band or switches to a backup frequency band to ensure the signal acquisition rate in the high-impedance abnormal area. S3 Multi-Frequency Data Acquisition: The automated control module sends electrical signals to the matrix electrodes in a preset frequency band sequence, synchronously triggering the data acquisition process. At the same time, the intelligent data analysis module uses a multi-level filtering algorithm to preprocess the raw electrical resistivity data. The automated control module supplies power to the matrix electrodes according to a preset sequence from low frequency to high frequency. When each frequency band is powered, multi-channel data acquisition is triggered synchronously. The intelligent data analysis module performs multi-level filtering on the raw data. That is, first, the power frequency interference is removed by bandpass filtering, then random noise is removed by wavelet transform, and finally the data curve is smoothed by sliding window algorithm. The specific implementation process of the above method is as follows: When the exploration mission enters the formal data acquisition phase, the automated control module first initiates the frequency band management program. Based on the geological noise characteristic analysis results, this program selects an initial working frequency band from a pre-set frequency band library, typically starting with a low-frequency signal. Frequency band switching employs a gradual transition mode, setting cross-fading intervals between adjacent frequency bands to avoid data distortion caused by sudden frequency changes. The power supply duration for each frequency band is dynamically adjusted using a geological conductivity prediction model, appropriately extending the signal transmission period in areas with high resistivity rock strata. Simultaneously with signal transmission, the multi-channel acquisition system enters synchronous operation. Each electrode unit is equipped with an independent high-precision analog-to-digital conversion channel, employing time-division multiplexing technology to achieve nanosecond-level synchronous sampling. The acquisition controller sends sampling clock signals to each channel via a fiber optic bus, ensuring that all channels initiate data recording at the precise moment of signal transmission. Addressing the unique electromagnetic interference characteristics of the mining environment, the system integrates a programmable gain amplifier at the data acquisition front end, automatically adjusting the measurement range based on the real-time monitored ambient noise level. The raw data first enters the first stage of the preprocessing pipeline—the bandpass filter unit. This unit employs a dynamic bandwidth adjustment mechanism, automatically setting the passband range according to the current operating frequency band, effectively suppressing 50Hz power frequency harmonics and their odd harmonics. The filter achieves a steep roll-off characteristic through cascading multiple second-order Butterworth filter stages, maintaining a flat amplitude response within the passband while providing 40dB suppression of out-of-band noise at tenth harmonics. After bandpass filtering, periodic interference components in the data are eliminated, and signal purity is significantly improved. The second-level processing employs an improved wavelet thresholding denoising algorithm. Based on the current geological structural characteristics, the system selects the best-matching mother wavelet from the discrete wavelet basis function library. For example, the symmetrical Coiflet wavelet is used in layered limestone regions, while the tightly supported DB wavelet is switched to in fractured zones. The wavelet transform uses a non-sampling tower decomposition structure, achieving multi-scale spatial frequency analysis while maintaining temporal resolution. Through an adaptive threshold calculation strategy, the system selectively removes random impulse noise caused by seepage from rock fissures, while preserving effective signal abrupt changes. The third-level processing introduces a sliding window mean filtering mechanism. The window length is dynamically set according to the complexity of the geological structure. Short windows are used in gently sloping rock areas to preserve detailed features, while the window is automatically extended in fault fracture zones to suppress local anomalies. Data points within each window are smoothed using a median-mean hybrid algorithm, eliminating data spikes while avoiding over-smoothing that blurs geological boundaries. The window sliding step size employs an adaptive strategy, decreasing the step size in areas of rapid signal change and increasing it in stable areas, achieving a balance between processing efficiency and detail preservation. The data, after undergoing three levels of filtering, is encapsulated into standardized data packets containing information such as the original sampling sequence number, processing timestamp, and quality assessment metrics. These data packets are transmitted in real-time to the 3D modeling engine via memory-mapped file technology, and simultaneously backed up to the solid-state storage array on the on-site host. During data transmission, the system continuously runs an integrity verification program, ensuring zero data loss through a cyclic redundancy check mechanism. When data anomalies are detected, a resampling mechanism is automatically triggered, supplementing data only for the problematic frequency band, avoiding the efficiency loss caused by retesting the entire frequency band. S4 Real-time Analysis and Anomaly Detection: The intelligent data analysis module constructs a three-dimensional resistivity inversion model based on the finite difference method, and combines geostatistical methods to locate low resistivity anomaly areas. At the same time, the machine learning algorithm classifies water hazard risk levels through a support vector machine classifier. If the resistivity value is lower than the preset threshold and the morphology is consistent with the characteristics of a water-bearing body, an audible and visual alarm is triggered and the abnormal coordinates are pushed to a remote terminal through the wireless communication module. When the intelligent data analysis module uses the finite difference method to perform three-dimensional resistivity inversion on the preprocessed data, it first discretizes the exploration area into a grid model, optimizes the resistivity value of each grid point through iterative calculation, and then combines the Kriging interpolation method of geostatistics to characterize the spatial morphology of the low resistivity anomaly area. At the same time, the machine learning algorithm, based on the trained support vector machine model, divides the anomaly area into three risk levels: safe, warning, and dangerous. When it is determined to be dangerous, the on-site host immediately activates the audible and visual alarm and marks the three-dimensional coordinates of the anomaly area. The specific implementation process of the above method is as follows: When the preprocessed electrical resistivity data enters the intelligent analysis engine, the system first initiates a 3D modeling program. This program automatically generates a 3D mesh framework based on the exploration area, employing a dynamic hierarchical strategy for mesh density: high-precision fine meshes are used in densely populated electrode areas, gradually thinning out in transitional edge zones. Each mesh cell is assigned an initial resistivity value, obtained through interpolation of data from similar geological conditions in a historical exploration database, forming an initial geological model. The finite difference method inversion engine then initiates an iterative optimization process. The system compares the observed electric field data with the initial model in a forward simulation, calculating the residual distribution between the theoretical and measured potentials. Based on the residual gradient information, the inversion engine employs an improved conjugate gradient algorithm to dynamically adjust the resistivity value of each grid cell. The iteration process is set with dual termination conditions: it automatically terminates when the residual decrease is less than a set threshold three times consecutively, or when the maximum number of iterations is reached. The optimized resistivity model needs to pass a smoothness check to avoid unreasonable local abrupt changes. During the characterization of low-resistivity anomaly zones, the Kriging interpolation module initiates spatial variogram analysis. This module first calculates the experimental variogram and fits a spherical, exponential, or Gaussian model using the least squares method. Based on the variogram parameters, the system performs structured interpolation on the finite-difference inversion results, particularly enhancing the boundary identification capability of low-resistivity anomaly zones. The interpolation process incorporates geological structural constraints, employing anisotropic variograms in fault fracture zones and switching to isotropic modes in intact rock strata regions to ensure that the anomaly morphology conforms to geological patterns. The risk level assessment engine simultaneously initiates a feature extraction process. This engine extracts three key indicators from the resistivity model: anomaly zone volume, resistivity contrast, and morphological complexity. These indicators, after normalization, are input into a pre-trained support vector machine classifier. The classifier employs a radial basis function kernel, and its decision boundary is learned from historical flood disaster case data, including sample features of various water-bearing structures such as water-rich sandstone and mudstone fracture zones. When an abnormal area is determined to be at a dangerous level, the emergency response procedure is immediately activated. The audible and visual alarm unit of the on-site host activates the red warning mode, and simultaneously triggers the camera to record on-site images. The three-dimensional coordinate information of the abnormal area is pushed to the remote terminal via the wireless communication module and highlighted with a flashing icon in the data visualization module. The alarm information includes the boundary range of the abnormal area, the minimum resistivity, and recommended handling measures, such as immediately stopping drilling and evacuating personnel, and strengthening drainage. S5 Remote Collaboration and Decision Support: The operating status of the on-site host, data quality heatmap, and exploration progress bar can be displayed in real time via a remote terminal.

[0020] This solution, by integrating an intelligent data analysis module into the on-site host, changes the inefficient mode of traditional mine electrical resistivity exploration systems that require data transmission to external computers for analysis. Furthermore, through the machine learning algorithms and fault diagnosis sub-modules integrated within the intelligent data analysis module, real-time multi-level filtering preprocessing, three-dimensional resistivity inversion, and low-resistivity anomaly identification of electrical resistivity data can be performed on-site. Compared to traditional offline analysis, this significantly improves data processing speed. Moreover, when the intelligent algorithm determines, through a support vector machine classifier, that the resistivity value is below a preset threshold and the morphology matches the characteristics of a water-bearing body, the system immediately triggers an audible and visual alarm and pushes the abnormal coordinates to a remote terminal via a wireless communication module. This achieves minute-level response to mine water hazard risks, providing real-time decision support for on-site construction personnel to suspend excavation and investigate potential hazards.

[0021] This solution utilizes a data visualization module and an automated control module. The data visualization module transforms electrical resistivity data into an interactive 3D geological model, allowing users to rotate, section, and query attributes of the model. Compared to traditional 2D displays, it more intuitively presents the spatial distribution of aquifers, faults, and other structures, making water hazard prevention more precise. Furthermore, the automated control module automatically analyzes geological parameters and dynamically adjusts the electrode spacing, arrangement, and power supply frequency sequence of the matrix electrodes, achieving intelligent optimization of exploration parameters without manual intervention. Combined with the real-time monitoring function of the remote terminal, it enables remote viewing of equipment operating status and data quality heatmaps, and remote diagnosis and repair of faults such as cable breaks or poor electrode contact, improving system reliability and construction safety.

[0022] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0023] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A mine electrical resistivity tomography integrated exploration system based on matrix electrodes, characterized in that, include: The multi-core cable includes a connecting clamp, a matrix electrode, a field host, and a remote terminal. One end of the multi-core cable is connected to the field host, and the other end is connected to the matrix electrode via the connecting clamp. The multi-core cable is used for the transmission of electrical signals. The connecting clamp is used to realize the physical and electrical connection between the multi-core cable and the matrix electrode to ensure the stable transmission of electrical signals. The matrix electrode is arranged in a matrix in the top and bottom plates of the mine roadway. The field host integrates an intelligent data analysis module, a data visualization module, a wireless communication module and an automatic control module. The intelligent data analysis module is used for real-time preprocessing, three-dimensional geological modeling, low-resistivity anomaly identification, and multi-source data fusion analysis of mine electrical resistivity exploration data. The intelligent data analysis module is equipped with machine learning algorithms and a fault diagnosis sub-module. The fault diagnosis sub-module is used to automatically detect potential water hazard risks through machine learning algorithms, and to automatically locate faults such as cable breakage and poor electrode contact and generate analysis reports with repair suggestions. The data visualization module is used to convert electrical exploration data into three-dimensional geological models, contour maps, and pseudo-color profile maps. The wireless communication module is used to realize real-time data interaction between the field host and the remote terminal. The automation control module is used to automatically adjust the matrix electrode arrangement parameters and optimize the data acquisition strategy according to the preset exploration plan.

2. The integrated mine electrical resistivity exploration system based on matrix electrodes according to claim 1, characterized in that: The connecting clamp adopts a combination structure of waterproof and corrosion-resistant engineering plastic shell and copper conductive terminals. The inner side of the engineering plastic shell is provided with an elastic sealing ring, the surface of the conductive terminals is gold-plated, and the two ends of the connecting clamp are respectively provided with anti-misinsertion interfaces.

3. The integrated mine electrical resistivity exploration system based on matrix electrodes according to claim 1, characterized in that: Each row of the matrix electrode contains 16-32 electrode units. The row spacing and column spacing of the matrix electrode are automatically calibrated by an automated control module. Each electrode unit of the matrix electrode is equipped with a pressure sensor and an inclinometer. Each row of electrodes in the matrix electrode is equipped with an independent power supply circuit.

4. The integrated mine electrical resistivity exploration system based on matrix electrodes according to claim 1, characterized in that: The intelligent data analysis module preprocesses the raw electrical resistivity data using a multi-level filtering algorithm. It also constructs a three-dimensional resistivity inversion model using the finite difference method and combines geostatistical methods to spatially locate and characterize low-resistivity anomaly areas. The machine learning algorithm is built based on a support vector machine classifier and is trained using historical exploration data to classify water hazard risk levels. The fault diagnosis submodule integrates an expert knowledge base.

5. The integrated mine electrical resistivity exploration system based on matrix electrodes according to claim 1, characterized in that: The data visualization module constructs an interactive three-dimensional geological model based on WebGL technology. The data visualization module adopts a double-buffered rendering mechanism to realize the dynamic updating of contour maps and pseudo-color profile maps, and overlays and displays the coordinates of measuring points, resistivity values ​​and the boundaries of anomaly areas. The data visualization module has an internal visualization report generation engine, which is used to export a PDF document containing the three-dimensional coordinates of the anomaly area, the resistivity threshold range and the risk assessment conclusion, and automatically adds the exploration timestamp and equipment serial number.

6. The integrated mine electrical resistivity exploration system based on matrix electrodes according to claim 1, characterized in that: The automated control module incorporates a rule-based exploration plan parser, an electrode placement parameter adjustment mechanism, and a data acquisition strategy optimization engine. The exploration plan parser parses XML-formatted task scripts and automatically generates electrode placement instructions. The electrode placement parameter adjustment mechanism employs a closed-loop control algorithm to dynamically optimize electrode spacing, arrangement, and supply electrode spacing based on geological conditions. The data acquisition strategy optimization engine combines Shannon's sampling theorem with geological noise characteristics to automatically determine the optimal sampling frequency and integration time, and enables a multi-frequency overlay acquisition mode in high-resistivity areas.

7. A method of using a matrix electrode-based integrated electrical resistivity tomography (EPT) system for mine exploration, based on any one of claims 1-6, characterized in that, Includes the following steps: S1 System Deployment and Self-Check: A matrix electrode array is arranged at a preset interval on the top and bottom plates of the mine roadway, and a multi-core cable is connected to the matrix electrode through a connecting clamp. After the on-site host is started, the automation control module automatically performs equipment self-test to verify cable conductivity, electrode contact resistance and inclinometer status. When poor contact and abnormal inclinometer are detected, the fault diagnosis submodule generates a repair suggestion report containing the coordinates of the fault point. S2 parameter configuration and task loading: The XML format task script is loaded through remote terminal and local interface. After the exploration plan parser parses the script, it generates electrode deployment instructions. The user can dynamically adjust the electrode spacing, arrangement and power supply electrode spacing according to geological conditions, and configure the power supply frequency combination and data quality threshold. At the same time, the data acquisition strategy optimization engine combines geological noise characteristics to automatically determine the multi-frequency superposition acquisition mode in high-resistivity areas. S3 Multi-Frequency Data Acquisition: The automated control module sends electrical signals to the matrix electrodes in a preset frequency band sequence, synchronously triggering the data acquisition process. At the same time, the intelligent data analysis module uses a multi-level filtering algorithm to preprocess the raw electrical resistivity data. S4 Real-time Analysis and Anomaly Detection: The intelligent data analysis module constructs a three-dimensional resistivity inversion model based on the finite difference method, and combines geostatistical methods to locate low resistivity anomaly areas. At the same time, the machine learning algorithm classifies water hazard risk levels through a support vector machine classifier. If the resistivity value is lower than the preset threshold and the morphology is consistent with the characteristics of a water-bearing body, an audible and visual alarm is triggered and the abnormal coordinates are pushed to a remote terminal through the wireless communication module. S5 Remote Collaboration and Decision Support: The operating status of the on-site host, data quality heatmap, and exploration progress bar can be displayed in real time via a remote terminal.

8. The method of using a matrix electrode-based integrated electrical resistivity tomography (EPT) exploration system for mines according to claim 7, characterized in that: In step S1, the automated control module verifies the cable's conductivity by sending a low-frequency pulse signal to the multi-core cable. When the signal attenuation exceeds a preset threshold, the fault diagnosis submodule locates the poorly contacting electrode based on the electrode potential difference distribution pattern. At the same time, it compares the inclinometer data with the preset burial angle. If the deviation exceeds the allowable range, it generates a visual repair report containing the electrode number, actual inclinometer angle, and calibration suggestions.

9. The method of using a matrix electrode-based integrated electrical resistivity tomography (EPT) exploration system for mines according to claim 7, characterized in that: In step S3, the automated control module supplies power to the matrix electrodes according to a preset sequence from low frequency to high frequency. When power is supplied to each frequency band, multi-channel data acquisition is triggered synchronously. The intelligent data analysis module performs multi-level filtering on the raw data, that is, firstly, the power frequency interference is removed by bandpass filtering, then random noise is removed by wavelet transform, and finally the data curve is smoothed by sliding window algorithm.

10. The method of using a matrix electrode-based integrated electrical resistivity tomography (EPT) exploration system for mines according to claim 7, characterized in that: In step S4, when the intelligent data analysis module performs three-dimensional resistivity inversion on the preprocessed data using the finite difference method, it first discretizes the exploration area into a grid model, optimizes the resistivity value of each grid point through iterative calculation, and then combines the Kriging interpolation method of geostatistics to characterize the spatial morphology of the low resistivity anomaly area. At the same time, the machine learning algorithm, based on the trained support vector machine model, divides the anomaly area into three risk levels: safe, warning, and dangerous. When it is determined to be dangerous, the on-site host immediately activates the audible and visual alarm and marks the three-dimensional coordinates of the anomaly area.

Images