A mine construction safety monitoring method

Abstract

The application discloses a mine construction safety monitoring method, relates to the technical field of mine safety monitoring, and comprises the following steps: working condition identification and monitoring channel dynamic configuration, source monitoring data space-time alignment collection, self-adaptive dynamic threshold correction and preprocessing, multi-disaster coupling risk assessment, three-dimensional space risk mapping and early warning, monitoring-construction closed-loop regulation and control; the application focuses on the main risk of current construction by dynamically configuring the monitoring channel according to the working condition identification, adopts a self-adaptive double-threshold correction technology based on a dynamic baseline, can effectively filter out the construction transient interference, reduces false positives, accurately identifies real danger, simultaneously utilizes multi-source information fusion and a coupling risk entropy model, realizes advanced comprehensive early warning of multi-disaster chain risk, additionally maps the risk information to a three-dimensional digital twin model, completes spatial positioning and visual deduction of the risk, and finally constructs an intelligent closed loop of monitoring-evaluation-early warning-regulation-retest, and realizes automatic regulation and control of the risk.

Description

Technical Field

[0001] This invention relates to the field of mine safety monitoring technology, and in particular to a method for monitoring the safety of mine construction. Background Technology

[0002] A mine shaft is a general term encompassing the shafts, chambers, equipment, surface buildings, and structures that form the underground coal mine production system. Sometimes, inclined shafts, vertical shafts, and adits in underground mine development are also referred to as mine shafts. Mine construction is a comprehensive construction project primarily focused on mine construction, combined with civil engineering and electromechanical installation. It involves a large workload, complex technology, and a long construction period. The main engineering works of mine construction include the shaft, bottom roadways and chambers, main stone portals, main haulage roadways, and mining area roadways.

[0003] Mine construction safety monitoring refers to the systematic work of monitoring and managing risks of geological conditions, surrounding rock stability, hydrology, gas, ventilation, support structure, construction equipment, working environment and construction safety behavior throughout the entire process of mine construction, through technical and management means such as instrument monitoring, manual inspection, data collection, analysis and judgment, and early warning and forecasting.

[0004] Based on existing technological findings, current mine construction safety monitoring technologies generally suffer from problems such as isolated monitoring, rigid early warning systems, and delayed decision-making. Specifically, each sensor operates independently with fixed thresholds, failing to effectively identify the coupled risks of multiple hazards caused by construction disturbances; a large number of false alarms are easily generated under transient conditions such as blasting and tunneling, masking the true dangers; early warning information lacks three-dimensional spatial positioning and situational analysis, relying on manual judgment with delayed response; furthermore, the monitoring system is disconnected from construction control, making it difficult to achieve proactive intervention and prevention of risks. Therefore, this invention proposes a mine construction safety monitoring method to address the problems existing in the prior art. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to propose a method for monitoring safety during mine construction, which enables dynamic adjustment of monitoring strategies according to working conditions, adaptive correction of early warning thresholds to disturbances, fusion and coupling of multi-source information for risk assessment, and ultimately forms an intelligent closed loop with construction control.

[0006] To achieve the objectives of this invention, the present invention is implemented through the following technical solution: a method for monitoring safety during mine construction, comprising the following steps:

[0007] Step 1: Operating Condition Identification and Dynamic Configuration of Monitoring Channels

[0008] By deploying condition identification sensors on mine construction equipment and work areas, the current construction conditions at the mine construction site are identified in real time. Based on the identified current construction conditions, a pre-set condition-channel configuration mapping table is queried to dynamically activate the set of target monitoring channels associated with it in the monitoring sensor network. The condition-channel configuration mapping table defines the correspondence between construction conditions, sensor identifiers to be activated, and acquisition parameters.

[0009] The types of monitoring parameters collected by the target monitoring channel set include microseismic activity, surrounding rock strain, pore water pressure, gas concentration, water inflow, and wind speed and pressure. The types of monitoring channels activated and the collection frequency are different for different construction conditions.

[0010] Step 2: Spatiotemporal Alignment Acquisition of Multi-Source Monitoring Data

[0011] For the target monitoring channel set, the monitoring data sequence of each monitoring channel is collected synchronously with the same roadway cross-section coordinates as spatial constraints and the time synchronization signal as time reference. The data is then processed by timestamp calibration and spatial coordinate matching to form a spatiotemporally aligned multi-source monitoring dataset. The time synchronization signal is distributed by a high-precision clock source deployed underground.

[0012] Step 3: Adaptive Dynamic Threshold Correction and Preprocessing

[0013] Based on the current construction conditions and lithological data obtained from geological information, adaptive dynamic threshold correction processing is performed on the monitoring parameter sequences in the multi-source monitoring dataset to filter out interference signals caused by transient disturbances during construction and output pre-processed multi-source effective risk data. The adaptive dynamic threshold correction processing establishes a dynamic baseline based on the typical disturbance characteristics of different construction stages.

[0014] Step 4: Multi-hazard Coupling Risk Assessment

[0015] Multi-source effective risk data is input into a multi-hazard coupling risk assessment model. A comprehensive coupling risk entropy value is calculated by integrating the anomaly probabilities of several monitoring parameters and their inter-parameter correlation. The multi-hazard coupling risk assessment model is used to quantify the spatiotemporal correlation risk between multiple disaster monitoring parameters.

[0016] Step 5: Three-dimensional spatial risk mapping and early warning

[0017] Based on the coupled risk entropy value, the corresponding risk level is determined. The risk level and risk location information are mapped to a three-dimensional mine geological-engineering digital twin model that is dynamically updated with the tunneling footage. Visual risk rendering and positioning are performed. Based on the geological structure and the spatial distribution of the current risk, the risk expansion path is simulated in the three-dimensional mine geological-engineering digital twin model, thereby generating multi-dimensional early warning information containing risk location, level and evolution direction.

[0018] Step Six: Monitoring - Closed-Loop Control of Construction

[0019] A mapping relationship between different risk levels and construction constraint rules is established in advance. When the risk level reaches the preset control level, the corresponding construction control instructions are automatically generated and issued to the on-site equipment. The construction control instructions include at least the control parameters of the tunneling, support or ventilation equipment. After the control is implemented, the process returns to steps two to five for risk retesting. If the risk level is reduced to the allowable range, the construction restrictions are lifted, forming a closed-loop control.

[0020] A further improvement is that, in step one, the current construction condition is identified in real time by collecting the status signals of the construction equipment or analyzing the image and sound features of the work surface. The condition-channel configuration mapping table defines the network address or logical identifier of the sensor to be activated corresponding to each construction condition.

[0021] Further improvements are made in the following aspects: In step two, the timestamp calibration is achieved through a unified clock synchronization module deployed underground, and the spatial coordinate matching is achieved by associating and binding the physical installation coordinates of each sensor in the monitoring sensor network with the roadway cross-section coordinates in the three-dimensional mine geological-engineering digital twin model.

[0022] A further improvement is made in step three, where the adaptive dynamic threshold correction process is performed by calling a dynamic baseline model that matches the current working conditions and lithology, and applying a transient disturbance allowance threshold for shielding short-term interference and a trend increment alarm threshold for capturing abnormal trends to each monitoring parameter.

[0023] A further improvement lies in the fact that the dynamic baseline model is established and applied in the following ways:

[0024] S1. A knowledge base is pre-established that includes the mapping relationship between lithology, construction method, disturbance intensity and benchmark threshold. The knowledge base is formed through historical data mining and field test calibration.

[0025] S2. Real-time acquisition of operating signals of construction equipment, including but not limited to tunneling motor parameters and blasting vibration wave trigger signals;

[0026] S3. Based on the short-term time window triggered by the working condition signal, and according to the established knowledge base and field lithology data, dynamically generate dynamic baseline models for each monitoring parameter to predict the normal data fluctuation range under construction disturbance.

[0027] S4. Use a dynamic baseline model for dual-threshold identification. First, allow the monitored data to fluctuate within the transient disturbance allowable threshold without triggering an alarm. Second, track the trend of the data after removing transient disturbances. If its rate of change or cumulative increment exceeds the trend increment alarm threshold, it is determined to be a valid risk signal.

[0028] A further improvement is made in step four, where the coupling risk entropy value is calculated using the following model:

[0029]

[0030] Where H is the coupling risk entropy value, P i The preprocessed probability of the i-th monitoring parameter is calculated based on the value of the i-th monitoring parameter in the multi-source effective risk data and its deviation from the trend increment alarm threshold corresponding to it in step three. Specifically, the deviation is mapped to a probability value in the interval [0,1] by a linear min-max normalization function or a nonlinear Sigmoid normalization function. n is the number of preprocessed monitoring parameters, and C is the coupling correlation coefficient (the larger the value, the stronger the synchronicity or causality of the anomalies of multiple parameters in time). It is determined by calculating the Deng's grey correlation degree or the mutual information entropy of continuous random variables within a set time window for any two parameter anomaly probability sequences, and is used to amplify the overall risk value when multiple parameters are coupled anomaly.

[0031] A further improvement is made in step five, where the risk level is determined by pre-setting several consecutive numerical intervals for the coupled risk entropy value, with each interval corresponding to a risk level. The specific path of the simulated risk extension is as follows: starting from the current risk location, the spatial transfer calculation of the risk value is performed along the weak surface direction of the geological structure. The spatial transfer calculation is based on the seepage theory.

[0032] A further improvement is that, in step five, the three-dimensional mine geological-engineering digital twin model is dynamically updated with geometric shape and attributes based on geological sketches, advanced drilling and support data, using tunneling cycles or preset advances as units. The updates are incremental and synchronized with the project schedule.

[0033] A further improvement is that, in step six, the construction control instructions include one or more of the following: reducing the tunneling speed, adjusting blasting parameters, increasing support density, initiating emergency grouting, or issuing personnel evacuation instructions; the removal of construction restrictions includes restoring the original tunneling speed, blasting scheme, or support parameters; and the decision logic of the closed-loop control is driven by a pre-set expert rule base.

[0034] The beneficial effects of this invention are as follows: By dynamically configuring monitoring channels that adapt to different working conditions, this invention enables monitoring resources to precisely focus on the types of disasters most likely to occur during current construction activities, significantly improving monitoring efficiency. By employing adaptive dual-threshold correction technology and intelligently filtering out transient interference from construction based on dynamic baselines, it fundamentally reduces the false alarm rate, ensuring reliable identification of true risk trends. Furthermore, by introducing multi-source information fusion and a coupled risk entropy assessment model, it achieves advanced early warning of multi-parameter correlated abnormal chain risks, completing the leap from isolated alarms to comprehensive coupled analysis. In addition, by mapping risk information to a real-time updated three-dimensional digital twin model, it realizes the visualized spatial positioning and extended path deduction of risks, greatly improving the intuitiveness and accuracy of decision-making. Finally, by constructing an intelligent closed loop of monitoring-assessment-early warning-control-retesting, it achieves automatic intervention in the construction process when risks reach thresholds, promoting a paradigm shift in mine safety management from passive alarms to proactive control, and comprehensively enhancing the proactive safety prevention and control capabilities during the construction period. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating the mine construction safety monitoring method of the present invention. Detailed Implementation

[0036] 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.

[0037] It should be noted that the technical means not described in detail in the following embodiments are all conventional means in the art, are not the key points of the invention, and will not be elaborated upon.

[0038] Example 1

[0039] See Figure 1 This embodiment provides a method for monitoring safety during mine construction, taking a mine roadway excavation project as an application scenario. The implementation process includes the following stages:

[0040] System initialization and model building

[0041] Before construction begins, complete the following preparatory work:

[0042] Monitoring sensor network deployment: Monitoring sections are arranged along the tunnel at designed intervals (e.g., 20 meters). Multiple sensors are deployed at each section, such as: microseismic monitoring probes (for detecting rock fractures), fiber optic strain sensors (for monitoring deep strain in the surrounding rock), pore water pressure gauges, gas concentration sensors, ultrasonic flow meters (for monitoring water inflow), and wind speed and pressure sensors. All sensors are connected to a unified underground industrial ring network.

[0043] 3D Digital Twin Model Construction: Based on geological survey reports and tunnel design drawings, an initial 3D mine geological-engineering digital twin model is constructed using 3D modeling software. The model includes information such as strata, faults, and tunnel design outlines. A corresponding "digital twin" is created for each physical sensor in the model, and its precise spatial coordinates (X, Y, Z) are recorded.

[0044] Pre-built knowledge base and rule base:

[0045] Working Condition-Channel Mapping Table: Predefined working conditions such as "blasting", "tunneling", "support", and "stopping mining". For example, under the "blasting" working condition, activate the micro-vibration, gas, and wind speed (for monitoring ventilation anomalies) channels and set them to 1kHz high-frequency acquisition (for 2 minutes); under the "support" working condition, activate the surrounding rock strain and roof delamination monitoring channels and set them to 1Hz low-frequency acquisition.

[0046] Dynamic baseline knowledge base: A knowledge base is established through historical data analysis and field tests. For example, under "hard rock-blasting" conditions, the normal disturbance limit of the microseismic event rate is recorded as 10 times / minute, and the allowable range of instantaneous fluctuation of surrounding rock strain is ±50 microstrain.

[0047] Risk-Control Rule Base: Defines the mapping between risk levels (e.g., low, medium, high, extremely high) and control measures. For example: "medium risk" triggers the instruction to "reduce tunneling speed to 50%"; "high risk" triggers the instruction to "stop tunneling and start reinforced support near the working face".

[0048] Real-time monitoring and data processing cycle

[0049] During construction, the following steps are performed cyclically:

[0050] Step 1: Operating Condition Identification and Dynamic Channel Configuration

[0051] Condition identification sensors (such as current transformers installed on the tunneling machine and microphones installed near the working face) operate in real time. For example, upon detecting the detonation signal of a blasting detonator, the system immediately determines the current condition as "blasting." Subsequently, it queries the condition-channel mapping table and automatically switches the monitoring network's acquisition focus to the preset "blasting monitoring mode," activating the high-frequency acquisition tasks of the relevant sensors.

[0052] Step 2: Spatiotemporal Alignment Acquisition of Multi-Source Data

[0053] In "blasting monitoring mode," activated microseismic and gas sensors begin collecting data using a unified millisecond-level timestamp provided by the downhole clock synchronization module. Based on the coordinates of each sensor, the system automatically categorizes data packets collected at the same cross-section and time, forming a "data cube." For example, at 12:00:00.000, the number of microseismic events, the percentage of gas concentration, and the wind speed value at the "K5+120" cross-section are packaged into a single data unit.

[0054] Step 3: Adaptive Dynamic Threshold Correction and Preprocessing

[0055] Assume the current working condition is "hard rock tunneling". The system calls the dynamic baseline model corresponding to "hard rock tunneling".

[0056] Dynamic baseline generation: A 30-second "strong disturbance window" is defined, triggered by the rising edge of the current signal from the high-power cutting of the tunnel boring machine. Within this window, based on the knowledge base, the transient disturbance allowable threshold for microseismic events is temporarily relaxed to 15 times / minute (normally 5 times / minute), and the transient disturbance allowable threshold for fiber optic strain sensors is relaxed to ±80 microstrains (normally ±30 microstrains). Simultaneously, the trend increment alarm threshold remains unchanged (e.g., the microseismic event rate continuously increases by more than 5% within a 10-minute sliding window).

[0057] Dual-threshold identification: The collected data is first compared with the "transient disturbance allowable threshold". As long as the data does not exceed this relaxed threshold within the window period, no alarm will be triggered even if the data spikes. After the window period ends, the system performs trend analysis on the data sequence. If it is found that although transient fluctuations are filtered out, the microseismic event rate slowly and continuously increases from 5 times / minute to 8 times / minute in the following 10 minutes, exceeding the 5% trend increment threshold, it is judged as a "valid microseismic anomaly risk signal" and output to the next stage.

[0058] Step 4: Multi-hazard Coupling Risk Assessment

[0059] Assuming that after the preprocessing in step three, the system simultaneously outputs two valid risk signals: “increased microseismic activity (P1=0.7)” and “gradual increase in gas concentration (P2=0.6)” (P values ​​are calculated by min-max normalization, 0 is normal, and 1 is extremely abnormal).

[0060] The system calculates the Deng's grey relational degree of these two anomalous sequences within the past 30-minute time window, and obtains a coupling relational degree coefficient C=1.8 (strong correlation, greater than 1).

[0061] Substitute into the coupling risk entropy formula to calculate:

[0062]

[0063] H=-[0.7*log(0.7)+ 0.6*log(0.6)]*1.8

[0064] The calculated H value is higher than the risk quantification value of any single-parameter anomaly, reflecting the coupled amplification effect of "microseismic fractures may conduct gas".

[0065] Step 5: Three-dimensional spatial risk mapping and early warning

[0066] Risk rating and location: Assuming the calculated H value falls within the preset "high risk" range, the system immediately locates the "K5+120" roadway cross-section, the source of the risk signal, in the three-dimensional mine geological-engineering digital twin model, and highlights this cross-section and a 20-meter area before and after it in red in the model.

[0067] Risk simulation: The model combines known geological information of small faults in the area to automatically simulate possible paths for risk to develop along the weak face of the fault into the roadway or working face above (indicated by red arrows or spreading animations).

[0068] Early Warning Generation: The system generates a structured early warning message: "High-risk alert! Location: Near section K5+120. Risk type: Risk of combined rockburst and gas outburst. Trend: May extend towards the working face. Recommendation: Immediately stop tunneling, strengthen support and gas extraction." This message is simultaneously pushed to the large screen at the surface monitoring center and the explosion-proof smartphone of the shift leader underground.

[0069] Step Six: Monitoring - Closed-Loop Control of Construction

[0070] The ground monitoring center confirms or the system automatically triggers (according to preset rules) the closed-loop control process.

[0071] Automatic control: Based on the "high risk" level, the system automatically issues a "stop advancing" command to the tunneling machine's electrical control system, and at the same time issues a command to the intelligent grouting system to "perform high-pressure grouting reinforcement on the section from K5+100 to K5+140".

[0072] Retesting and Unlocking: After the grouting reinforcement measures were implemented, the system continuously monitored the microseismic and gas data in the area. Approximately one hour later, the coupled risk entropy H value calculated from the retesting decreased to the "low risk" range. The system automatically unlocked the "stop advancing" lock and issued a command to the tunneling machine to "resume slow tunneling (70% of the original speed)". This completed a full "monitoring-early warning-control-unlocking" intelligent closed loop.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring safety during mine construction, characterized in that, Includes the following steps: Step 1: By deploying condition identification sensors on mine construction equipment and work areas, the current construction conditions at the mine construction site are identified in real time. Based on the current construction conditions, the pre-set condition-channel configuration mapping table is queried, and the set of target monitoring channels associated with it in the monitoring sensor network is dynamically activated. Step 2: For the target monitoring channel set, using the same roadway cross-section coordinates as spatial constraints and the time synchronization signal as time reference, synchronously collect the monitoring data sequence of each monitoring channel, and perform timestamp calibration and spatial coordinate matching processing on the data to form a spatiotemporally aligned multi-source monitoring dataset; Step 3: Based on the current construction conditions and the lithological data obtained from geological information, adaptive dynamic threshold correction processing is performed on the monitoring parameter sequences in the multi-source monitoring dataset to output the preprocessed multi-source effective risk data; Step 4: Input the multi-source effective risk data into the multi-hazard coupled risk assessment model to calculate a comprehensive coupled risk entropy value; Step 5: Based on the coupled risk entropy value, determine the risk level, map the risk level and risk location information to the 3D mine geological-engineering digital twin model, perform visual risk rendering and positioning, and simulate the risk expansion path in the 3D mine geological-engineering digital twin model to generate multi-dimensional early warning information; Step Six: Establish a mapping relationship between different risk levels and construction constraint rules in advance. When the risk level reaches the preset control level, the corresponding construction control instructions are automatically generated and issued to the on-site equipment. After the control is implemented, the risk is retested. If the risk level is reduced to the allowable range, the construction restrictions are lifted, forming a closed-loop control.

2. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step one, the current construction condition is identified in real time by collecting the status signals of the construction equipment or analyzing the image and sound features of the work surface. The condition-channel configuration mapping table defines the network address or logical identifier of the sensor to be activated corresponding to each construction condition.

3. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step two, the timestamp calibration is achieved through a unified clock synchronization module deployed underground, and the spatial coordinate matching is achieved by associating and binding the physical installation coordinates of each sensor in the monitoring sensor network with the roadway cross-section coordinates in the three-dimensional mine geological-engineering digital twin model.

4. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step three, the adaptive dynamic threshold correction process is performed by calling a dynamic baseline model that matches the current working conditions and lithology, and applying a transient disturbance allowance threshold for shielding short-term interference and a trend increment alarm threshold for capturing abnormal trends to each monitoring parameter.

5. The method for monitoring safety during mine construction according to claim 4, characterized in that: The steps for establishing the dynamic baseline model are as follows: S1. Establish a knowledge base in advance that includes the mapping relationship between lithology, construction method, disturbance intensity and benchmark threshold; S2. Real-time acquisition of operating signals from construction equipment; S3. Based on the short time window triggered by the working condition signal, and according to the established knowledge base and field lithology data, dynamically generate dynamic baseline models for each monitoring parameter. S4. Use a dynamic baseline model for dual-threshold identification. Allow data to fluctuate within the transient disturbance allowable threshold without triggering an alarm. Track the trend of the data after removing transient disturbances. If its rate of change or cumulative increment exceeds the trend increment alarm threshold, it is determined to be a valid risk signal.

6. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step four, the coupling risk entropy value is calculated using the following model: Where H is the coupling risk entropy value, P i Let represent the abnormal probability of the i-th monitoring parameter after preprocessing, n be the number of preprocessed monitoring parameters, and C be the coupling correlation coefficient.

7. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step five, the risk level is determined by pre-setting several continuous numerical intervals for the coupled risk entropy value. Each interval corresponds to a risk level. The specific path of the simulated risk extension is as follows: starting from the current risk location, the spatial transfer calculation of the risk value is performed along the weak surface direction of the geological structure.

8. The method for monitoring safety during mine construction according to claim 1, characterized in that: In step five, the three-dimensional mine geological-engineering digital twin model is based on tunneling cycles or preset advances as units, and its geometric shape and attributes are dynamically updated according to geological sketches, advanced drilling and support data.

Images