Coal mine fireproof early warning method based on intelligent monitoring

Abstract

The invention discloses a coal mine fireproof early warning method based on intelligent monitoring, and relates to the technical field of coal mine fireproof early warning, and the method comprises the following steps: obtaining coal mine environment information data parameters in real time through a coal mine fireproof early warning system, including humidity dynamic parameter information and air pressure parameter information; the environment is intelligently evaluated through the generated environment safety change index, when the environment is judged to be a high-risk environment, the system dynamically adjusts the sampling frequency based on the preset sampling frequency, fine changes such as methane diffusion or air static pressure fluctuation in the early stage of a fire disaster are captured in time, rapid early warning is ensured, and the safety of the fire disaster is improved. The accuracy and timeliness of signal capture are improved, and early intervention time is won for managers. And in a low-risk environment, the system collects data at a fixed sampling frequency, so that unnecessary calculation tasks and energy consumption are reduced, and long-term monitoring and trend analysis are supported. Through the dual regulation and control mechanism, the system not only ensures the data integrity, but also optimizes the resource utilization and operation efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine fire prevention and early warning technology, and in particular to a coal mine fire prevention and early warning method based on intelligent monitoring. Background Art

[0002] Coal mine fire warning based on intelligent monitoring is a systematic solution that uses modern information technology to monitor the coal mine environment in real time to prevent and reduce fire risks. It uses technologies such as sensor networks, the Internet of Things, big data, and artificial intelligence to monitor parameters such as temperature, gas concentration (such as carbon monoxide, methane), and ventilation status in the mine around the clock. The system automatically analyzes monitoring data, identifies potential fire risks, such as abnormal temperature rise or excessive harmful gases, and warns of possible fires through predictive models and algorithms.

[0003] The core value of this system is to improve the reaction speed and predictability of coal mine fire prevention and reduce the omissions of manual inspections. The early warning system can push real-time alarm information to mine managers and relevant departments according to different risk levels, and give corresponding emergency treatment suggestions. By discovering risks in advance and taking measures to avoid fire accidents, it not only protects the lives of coal miners, but also reduces the loss of equipment and mineral resources. At the same time, this intelligent monitoring system can also analyze historical data and summarize the laws of fire occurrence, so as to optimize the mine's fire prevention measures and management plans, and promote the continuous improvement of coal mine safety production.

[0004] The prior art has the following deficiencies:

[0005] Existing coal mine fire warning systems usually use fixed frequency to collect environmental data, which is effective under normal circumstances, but has limitations in capturing early fire signals. Fixed frequency acquisition cannot respond to small changes in a short period of time, and the temperature and gas concentration (such as carbon monoxide, methane, etc.) at the beginning of a fire may fluctuate violently in an instant. If the system fails to obtain these key data in time, mine managers will not be able to quickly perceive the risks and miss the golden time for early intervention in the fire. Once the fire gets out of control, a small-scale fire source will quickly evolve into an uncontrollable large fire, causing the mine environment to deteriorate and the smoke exhaust system to collapse, greatly increasing the difficulty of fire fighting and the risk of people being trapped, making evacuation and rescue work more complicated and difficult.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a coal mine fire prevention and early warning method based on intelligent monitoring to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above-mentioned purpose, the environment is intelligently evaluated through the generated environmental safety change index. When it is judged to be a high-risk environment, the system dynamically adjusts the sampling frequency based on the preset sampling frequency, and promptly captures subtle changes in the early stages of a fire, such as methane diffusion or air static pressure fluctuations, to ensure rapid early warning, improve the accuracy and timeliness of signal capture, and gain early intervention time for managers. In a low-risk environment, the system collects data at a fixed sampling frequency, reduces unnecessary computing tasks and energy consumption, and supports long-term monitoring and trend analysis. Through this dual control mechanism, the system not only ensures data integrity, but also optimizes resource utilization and operating efficiency. The present invention provides the following technical solutions: A coal mine fire prevention and early warning method based on intelligent monitoring, comprising the following steps:

[0009] Obtain coal mine environmental information data parameters in real time through the coal mine fire early warning system, including humidity dynamic parameter information and air pressure parameter information;

[0010] Perform an abnormal analysis on the collected humidity dynamic parameter information and air pressure parameter information, and input the humidity dynamic parameter information and air pressure parameter information after abnormal analysis into a pre-learned machine learning model, generate an environmental safety change index through the machine learning model, and use the environmental safety change index to intelligently evaluate the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system;

[0011] Based on the results of machine learning model evaluation, the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are divided into low-risk environmental information or high-risk environmental information, and the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are intelligently perceived;

[0012] When environmental information is judged to be low-risk, it will continue to collect data at the preset environmental sampling frequency. When environmental information is judged to be high-risk, the sampling frequency will be dynamically adjusted based on the preset environmental sampling frequency to capture subtle changes in the environment in a timely manner.

[0013] Preferably, the humidity dynamic parameter information includes changes in the diffusion rate of methane gas, and the air pressure parameter information includes fluctuations in air static pressure. The changes in the diffusion rate of methane gas refer to the dynamic changes in the diffusion rate of methane gas in different areas within a mine over time, and the fluctuations in air static pressure refer to the drastic changes in the static pressure of the air in the mine in a short period of time, which reflects the stability of the mine ventilation system and the equilibrium state of gas flow.

[0014] Preferably, after performing an abnormal analysis on the local humidity fluctuation, a local humidity instantaneous fluctuation index is generated, and after performing an abnormal analysis on the air static pressure fluctuation, an air static pressure fluctuation abnormal index is generated. The local humidity instantaneous fluctuation index and the air static pressure fluctuation abnormal index generated after the abnormal analysis are input into a pre-learned machine learning model to generate an environmental safety change index through the machine learning model, and the coal mine environmental information data parameters acquired in real time by the coal mine fire warning system are intelligently evaluated through the environmental safety change index.

[0015] Preferably, the logic of performing abnormal analysis on the local humidity fluctuation and generating the local humidity instantaneous fluctuation index is as follows:

[0016] In the detection window, collect local humidity data and calibrate the collected local humidity data as H(t i ), H(t i ) indicates that at time t i The local humidity value obtained by the i-th sampling at time t i ={t1, t2, t3, ..., t n}, where n represents the total number of sampling times;

[0017] For every two consecutive sampling moments, the instantaneous humidity change rate is calculated, and the calculation expression is as follows:

[0018]

[0019] , where ΔH(t i ) is the instantaneous humidity change rate, at time point t i The instantaneous humidity value at the time, that is, the instantaneous humidity recorded at the i-th sampling time, H(t i-1 ) is at the previous time point t i-1 The instantaneous humidity value is used to calculate the change in humidity, and Δt is the time interval between two samplings;

[0020] Based on the instantaneous humidity change rate, the acceleration of humidity fluctuation is further calculated to capture the severity of humidity changes. The calculation expression is as follows:

[0021]

[0022] , where ΔA H (t i ) indicates that at time t i The acceleration of humidity fluctuation at any moment, ΔH(t i-1 ) is at the previous time point t i-1 The instantaneous rate of change of humidity;

[0023] The dynamic sensitivity function is introduced to capture the significant fluctuations of the acceleration of humidity fluctuations in a short period of time and amplify the instantaneous drastic changes. The calculation expression is as follows:

[0024]

[0025] , where S(t i ) represents the time point t i The dynamic sensitivity value at the moment, λ is the sensitivity adjustment parameter, and θ is the threshold of humidity fluctuation;

[0026] Then calculate the instantaneous anomaly coefficient, the calculation expression is as follows:

[0027] E H (t i )=A H (t i )·S(t i )

[0028] , where E H (t i ) represents the time point t i The instantaneous anomaly coefficient at the moment;

[0029] The absolute values ​​of all instantaneous anomaly coefficients are accumulated to generate the local humidity instantaneous fluctuation index. The calculation expression is as follows:

[0030]

[0031] , where I H It is the instantaneous fluctuation index of local humidity.

[0032] Preferably, the logic of performing abnormal analysis on the air static pressure fluctuation and generating the air static pressure fluctuation abnormal index is as follows:

[0033] Under the detection window, the air static pressure data of the mine is decomposed, and the time series data is divided into multiple small windows. The length of each sliding window is W, and the expression is as follows: {P(t)|t∈[t0, t0+W]), where P(t) is the air static pressure value at time point t, t0 is the start time of the sliding window, and t0+W is the end time of the sliding window;

[0034] In each sliding window, the change of the air static pressure value between consecutive time points is calculated by differential operation, that is, the instantaneous fluctuation amplitude of the air static pressure value. The calculation expression is as follows:

[0035] ΔP(t)=|P(t+Δt)-P(t)|

[0036] , where ΔP(t) is the change in air static pressure at time point t, and Δt is the time interval between two samplings;

[0037] The detected change in air static pressure value ΔP(t) is further subjected to nonlinear analysis to calculate the change rate. The calculation expression is as follows:

[0038]

[0039] , where R(t) is the nonlinear change rate of the air static pressure at time point t, α and β are both smoothing coefficients, and γ is the nonlinear adjustment coefficient;

[0040] In order to quantify the overall intensity of static pressure fluctuations, the nonlinear change rate R(t) at each time point is integrated to generate the cumulative fluctuation energy. The calculation expression is as follows:

[0041]

[0042] , where E W Accumulated wave energy;

[0043] The accumulated wave energy E W Input into the nonlinear mapping function to generate the final air static pressure fluctuation anomaly index. The calculation expression is as follows:

[0044]

[0045] , where I P is the air static pressure fluctuation abnormality index, is the amplification factor used to adjust the exponential range, τ is the energy threshold, and tanh(·) is the hyperbolic tangent function, which maps the energy value to the range of 0-1.

[0046] Preferably, the local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P Input into the pre-learned machine learning model, and generate the environmental safety change index ESCI through the machine learning model, based on the formula:

[0047]

[0048] , where f1 and f2 are the local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P The preset proportional coefficient, and f1 and f2 are both greater than 0.

[0049] Preferably, the generated environmental safety change index is compared and analyzed with a preset environmental safety change index reference threshold, and the analysis results are as follows:

[0050] If the environmental safety change index is greater than or equal to a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire early warning system are classified as high-risk environmental information;

[0051] If the environmental safety change index is less than a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire warning system are classified as low-risk environmental information.

[0052] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0053] The present invention performs intelligent assessment of the environment through the generated environmental safety change index. When the system determines that it is high-risk environmental information, based on the preset sampling frequency, the system can dynamically adjust the sampling frequency so as to quickly respond and capture subtle changes in the early stages of a fire in a short period of time. For example, in the early stages of a fire, drastic fluctuations in the diffusion rate of methane gas or air static pressure may be extremely short-lived, and the traditional fixed sampling frequency cannot respond to these rapid fluctuations in time. Through this dynamic adjustment mechanism, the system can increase the sampling frequency, obtain key data more quickly, and ensure that early warnings are issued in a timely manner when risks increase, greatly improving the accuracy and timeliness of capturing signals in the early stages of a fire, thereby providing mine managers with sufficient time for early intervention.

[0054] For situations where environmental information is judged to be low-risk, the present invention uses a preset fixed sampling frequency to collect data and does not perform additional computing tasks. This design effectively saves the system's computing resources and energy consumption, because when the environment is stable and there is no obvious risk, the fixed sampling frequency can meet the data collection needs, and there is no need to dynamically adjust the frequency for over-sampling. This not only helps to monitor a stable environment for a long time and ensure continuous basic data collection, but also provides support for subsequent trend analysis and baseline establishment. This intelligent control mechanism enables the system to avoid unnecessary waste of resources while ensuring safety, thereby maximizing resource utilization. This dual control method maintains a fixed frequency when the risk is low and flexibly adjusts when the risk is high, which not only ensures the integrity of the data, but also improves the operating efficiency and sustainability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0056] Figure 1 The present invention is a method flow chart of a coal mine fire prevention and early warning method based on intelligent monitoring. DETAILED DESCRIPTION

[0057] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0058] The present invention provides Figure 1 A coal mine fire prevention early warning method based on intelligent monitoring is shown, comprising the following steps:

[0059] Obtain coal mine environmental information data parameters in real time through the coal mine fire early warning system, including humidity dynamic parameter information and air pressure parameter information;

[0060] The dynamic parameter information of humidity includes local humidity fluctuations, and the parameter information of air pressure includes air static pressure fluctuations. Local humidity fluctuations refer to the rapid increase or decrease of humidity in a specific area in a mine in a short period of time, rather than a smooth or slow change. Air static pressure fluctuations refer to the drastic change of the static pressure of the air in the mine in a short period of time, which reflects the stability of the mine ventilation system and the balance state of gas flow.

[0061] Perform an abnormal analysis on the collected humidity dynamic parameter information and air pressure parameter information, and input the humidity dynamic parameter information and air pressure parameter information after abnormal analysis into a pre-learned machine learning model, generate an environmental safety change index through the machine learning model, and use the environmental safety change index to intelligently evaluate the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system;

[0062] After performing an abnormal analysis on the local humidity fluctuation, a local humidity instantaneous fluctuation index is generated; after performing an abnormal analysis on the air static pressure fluctuation, an air static pressure fluctuation abnormal index is generated; the local humidity instantaneous fluctuation index and the air static pressure fluctuation abnormal index generated after the abnormal analysis are input into a pre-learned machine learning model, and an environmental safety change index is generated through the machine learning model; the coal mine environmental information data parameters acquired in real time by the coal mine fire warning system are intelligently evaluated through the environmental safety change index;

[0063] Local humidity fluctuations in coal mine environments may be important early signals for fires. During the spontaneous combustion or initial combustion stage of coal mines, the increase in temperature causes the water in the surrounding materials to evaporate rapidly, releasing water vapor, resulting in abnormal fluctuations in local humidity in a short period of time. In addition, poor ventilation or gas accumulation may also cause humidity changes. Therefore, monitoring instantaneous fluctuations in humidity can detect potential risks of fire or ventilation system abnormalities in advance. However, monitoring methods with fixed sampling frequencies may miss these small changes in a short period of time. Humidity fluctuations often occur instantaneously, and if the system collects data at fixed time intervals (such as every 10 minutes), it may not be able to capture key changes that occur within the sampling interval in a timely manner. For example, if the humidity rises or falls rapidly within a few seconds, but the system has a long sampling period, these important early warning signals will be ignored, resulting in the risk of the early stage of the fire not being perceived and handled in time. Once the fire expands, the difficulty of extinguishing and the risk of personnel evacuation will increase significantly.

[0064] The logic of abnormal analysis of local humidity fluctuation and generation of local humidity instantaneous fluctuation index is as follows:

[0065] In the detection window, collect local humidity data and calibrate the collected local humidity data as H(t i ), H(t i ) indicates that at time t i The local humidity value obtained by the i-th sampling at time t i ={t1, t2, t3, ..., t n}, where n represents the total number of sampling times;

[0066] The local humidity data collected usually includes the following key indicators, which help to fully understand the humidity conditions in different areas of the coal mine so as to monitor the early signs of fire or ventilation abnormalities in a timely manner:

[0067] 1. Instantaneous humidity value

[0068] This is the humidity value collected at a specific moment, indicating the humidity level of the environment at a certain point in time. This is the most basic humidity data, usually collected in real time by sensors.

[0069] 2. Humidity change rate

[0070] Calculate the humidity change amplitude between two consecutive sampling moments to determine whether the humidity is rising steadily, falling, or fluctuating suddenly. This is very important for capturing subtle changes in the early stages of a fire.

[0071] 3. Humidity fluctuation range

[0072] Indicates the range of humidity fluctuations over a specific period of time. By comparing the highest and lowest humidity values ​​over a period of time, you can understand whether the humidity has experienced drastic fluctuations, especially changes in a short period of time.

[0073] 4. Differences in local humidity distribution

[0074] The humidity distribution at different locations, especially whether the humidity is uniform at different depths and in different areas within the mine. This data can help identify poor ventilation or local anomalies.

[0075] 5. Time series trend of humidity

[0076] The trend of humidity values ​​over a period of time is often used to predict the direction of humidity changes. A continuous upward or downward trend may be a sign of environmental changes or a precursor to fire.

[0077] 6. Humidity gradient

[0078] This refers to the difference in humidity between different areas. Humidity gradients can indicate whether humidity is evenly distributed in the mine, and poorly ventilated areas may have higher humidity gradients, indicating potential risks.

[0079] Local humidity data includes indicators such as instantaneous humidity value, humidity change rate, fluctuation amplitude, humidity distribution difference, time trend, humidity gradient, etc. These data can not only provide current environmental humidity information, but also help to warn potential dangers such as fire or ventilation failure in advance by analyzing the dynamic changes of humidity.

[0080] For every two consecutive sampling moments, the instantaneous humidity change rate is calculated, and the calculation expression is as follows:

[0081]

[0082] , where ΔH(t i ) is the instantaneous humidity change rate, at time point t i The instantaneous humidity value at the time, that is, the instantaneous humidity recorded at the i-th sampling time, H(t i-1 ) is at the previous time point t i-1 The instantaneous humidity value, i.e. the result of the i-1th sampling, H(t i-1 ) and H(t i ) is the data of two consecutive samplings, which is used to calculate the change of humidity, and Δt is the time interval between two samplings, that is, the time from the i-1th sampling to the i-th sampling;

[0083] Based on the instantaneous humidity change rate, the acceleration of humidity fluctuation is further calculated to capture the severity of humidity changes. The calculation expression is as follows:

[0084]

[0085] , where ΔA H (t i ) indicates that at time t i The acceleration of humidity fluctuation at any moment, ΔH(t i-1 ) is at the previous time point t i-1 The instantaneous rate of change of humidity;

[0086] The dynamic sensitivity function is introduced to capture the significant fluctuations of the acceleration of humidity fluctuations in a short period of time and amplify the instantaneous drastic changes. The calculation expression is as follows:

[0087]

[0088] , where S(t i ) represents the time point t i The dynamic sensitivity value at the moment, λ is the sensitivity adjustment parameter used to control the steepness of the curve, and θ is the threshold of humidity fluctuation, which is used to determine whether it reaches an abnormal level;

[0089] Significant fluctuations in humidity acceleration refer to dramatic changes in humidity over a short period of time. These changes not only reflect the speed of change in humidity levels, but also reveal abnormal dynamics in the environment. The following are several common significant humidity acceleration fluctuations and their potential meanings:

[0090] 1. Humidity rises faster

[0091] Phenomenon: Humidity increases rapidly in a short period of time and rises at an accelerated rate (the rate of change continues to increase).

[0092] Possible causes:

[0093] Early stage of coal seam spontaneous combustion: The spontaneous combustion process will release water vapor, causing the humidity to rise rapidly in local areas.

[0094] Equipment or pipe rupture: Moisture suddenly leaks into a specific area of ​​the mine.

[0095] Restricted ventilation: causes humidity to accumulate in certain areas and not escape.

[0096] risk:

[0097] An accelerated rise in humidity over a short period of time may indicate the early signs of spontaneous combustion, and timely detection can help prevent and control fires in advance.

[0098] Poor ventilation can lead to humidity accumulation, increasing the risk of exceeding other parameters (such as methane concentrations).

[0099] 2. Humidity drops rapidly

[0100] Phenomenon: The humidity level drops rapidly in a short period of time, and the rate of decline continues to accelerate.

[0101] Possible causes:

[0102] Sudden strong wind inflow: The ventilation system is not functioning properly, causing humidity to be expelled quickly.

[0103] A sudden rise in temperature, such as in the early stages of a fire or when equipment overheats, causes moisture in the air to evaporate and the air becomes dry.

[0104] Excessive fan operation: The exhaust system operates abnormally, excessive moisture is discharged, and humidity imbalance occurs.

[0105] risk:

[0106] A sharp drop in humidity will make the environment dry and increase the risk of spontaneous combustion of combustibles.

[0107] Abnormalities in the ventilation system can cause air pressure imbalance, uneven gas concentration distribution, and increase fire hazards.

[0108] 3. Humidity fluctuations are frequent and the amplitude is increasing (oscillations are accelerated)

[0109] Phenomenon: Humidity fluctuates frequently up and down in a short period of time, and the changes are accelerated, that is, the humidity fluctuations in different directions are increasing.

[0110] Possible causes:

[0111] Unstable ventilation system: If the ventilation equipment starts and stops frequently, the humidity will be unstable.

[0112] Multiple sources of humidity interference: If there are multiple humidity sources at the same time (such as pipeline leakage and spontaneous combustion), it will cause humidity oscillation.

[0113] Frequent changes in air flow direction: The ventilation paths are variable, which prevents moisture from dispersing stably.

[0114] risk:

[0115] Humidity oscillations may cause sensor data anomalies and increase the probability of false positives or negatives.

[0116] Drastic changes in humidity fluctuations can mask other potential fire precursors and affect the system's early warning effect.

[0117] 4. Humidity changes and other parameters synchronous acceleration fluctuations

[0118] Phenomenon: The accelerated fluctuation of humidity changes synchronously with parameters such as temperature and gas concentration (such as methane and carbon monoxide).

[0119] Possible causes:

[0120] Coal seam spontaneous combustion process: Humidity, temperature and gas concentration may change dramatically at the same time in the early stages of a fire.

[0121] Overall failure of the ventilation system: causing humidity and gas concentration to rise or fall simultaneously.

[0122] risk:

[0123] The synchronous accelerated changes of multiple parameters indicate that there may be a pre-fire situation or a serious failure in the ventilation system, which requires urgent investigation.

[0124] Significant fluctuations in humidity fluctuation acceleration mainly include accelerated humidity rise, accelerated humidity fall, frequent oscillations, and synchronous changes with other parameters. These fluctuations not only reflect the dramatic changes in humidity itself, but may also reveal potential fires, ventilation anomalies, or equipment failures inside coal mines. By accurately capturing these significant fluctuations in humidity acceleration, the fire warning system can identify risks in advance and avoid accidents.

[0125] Then calculate the instantaneous anomaly coefficient, the calculation expression is as follows:

[0126] E H (t i )=A H (t i )·S(t i )

[0127] , where E H (t i ) represents the time point t i The instantaneous anomaly coefficient at the moment;

[0128] Instantaneous abnormal coefficient E H (t i ) is to quantitatively evaluate the degree of dramatic fluctuation of humidity at a specific moment, and by combining the results of the humidity change rate and the sensitivity function, amplify the humidity fluctuation signals that may indicate environmental abnormalities. It can capture sudden anomalies in humidity changes, such as water vapor released before a fire, and sudden changes in humidity caused by the failure of the ventilation system. This coefficient can not only reflect the significance of the current humidity fluctuation, but also screen out those abnormal fluctuations that are critical to mine safety through real-time analysis. Its main value lies in avoiding underreporting caused by fixed-frequency sampling, improving the system's response speed to fire precursors and ventilation failures, and providing timely warnings for managers. In other words, the instantaneous anomaly coefficient amplifies potential risk signals that are not easily detected in the environment, enabling the system to quickly detect anomalies in small changes and prevent them before they happen.

[0129] The absolute values ​​of all instantaneous anomaly coefficients are accumulated to generate the local humidity instantaneous fluctuation index. The calculation expression is as follows:

[0130]

[0131] , where I H is the instantaneous fluctuation index of local humidity;

[0132] It can be seen from the calculation expression of the local humidity instantaneous fluctuation index that the larger the local humidity instantaneous fluctuation index performance value generated after the abnormal analysis of the local humidity fluctuation, the more it means that the humidity has fluctuated violently in a short period of time. This indicates that there are unstable factors in the mine environment, such as poor ventilation or the release of water vapor before the spontaneous combustion of the coal seam. At this time, the early warning system with a fixed sampling frequency is more likely to miss these rapid and short-term change signals, increasing the probability that the system cannot respond in time. The larger the performance value of the local humidity instantaneous fluctuation index, the higher the risk of system omissions, because the key information within the sampling interval may not be captured. When the performance value of the local humidity instantaneous fluctuation is small, it means that the humidity change tends to be stable. At this time, the effect of fixed frequency sampling is relatively more reliable, and the probability of system omissions is also reduced.

[0133] In coal mine environments, changes in air static pressure fluctuations can indeed be an important signal in the early stages of a fire. When combustion occurs in a local area or coal seams spontaneously combust, the rise in air temperature will cause air expansion, resulting in dramatic fluctuations in static pressure in a short period of time. In addition, gas release, ventilation system failure, or airflow turbulence can also change the air static pressure, and these phenomena are often closely related to the initial signs of a fire. However, if the fire warning system uses a fixed sampling frequency, these instantaneous pressure fluctuation signals may be missed. This is because fixed-frequency sampling can only obtain data within a specific interval and cannot continuously capture rapid and short-term abnormal changes. In the early stages of a fire, static pressure fluctuations and other environmental parameters such as temperature or harmful gas concentrations often change dramatically within a few seconds. If the system happens not to sample during this period, the risk signal will be ignored. This will result in mine managers being unable to perceive the precursors of a fire in a timely manner, missing the golden time for emergency intervention, and increasing the possibility of a worsening fire.

[0134] The logic of abnormal analysis of air static pressure fluctuation and generation of air static pressure fluctuation abnormal index is as follows:

[0135] Under the detection window, the air static pressure data of the mine is decomposed, and the time series data is divided into multiple small windows. The length of each sliding window is W, which is used to capture fluctuations in a short period of time. The expression is as follows: {P(t)|t∈[t0, t0+W]}, where P(t) is the air static pressure value at time point t, t0 is the start time of the sliding window, and t0+W is the end time of the sliding window;

[0136] In each sliding window, the change of the air static pressure value between consecutive time points is calculated by differential operation, that is, the instantaneous fluctuation amplitude of the air static pressure value. The calculation expression is as follows:

[0137] ΔP(t)=|P(t+Δt)−P(t)|

[0138] , where ΔP(t) is the change in air static pressure at time point t, and Δt is the time interval between two samplings;

[0139] The purpose of calculating the change in air static pressure between consecutive time points is to capture the instantaneous fluctuations in air pressure in the mine, which is an important signal for identifying ventilation system anomalies, fire precursors or gas leaks. When a fire or spontaneous combustion occurs, it will cause rapid expansion of local air or release of gas, resulting in drastic changes in air static pressure in a short period of time. If the system can detect these pressure changes in time, it can issue an early warning before the risk of fire or explosion deteriorates further to avoid the spread of disasters. On the contrary, if the pressure fluctuations are not captured, the danger signal will be ignored, which may cause managers to miss critical emergency intervention time. Therefore, the change in static pressure between consecutive time points can reflect the dynamic state of airflow in the mine and provide a key basis for fire prevention and early warning.

[0140] The detected change in air static pressure value ΔP(t) is further subjected to nonlinear analysis to calculate the change rate. The calculation expression is as follows:

[0141]

[0142] , where R(t) is the nonlinear change rate of the air static pressure value at time point t, α and β are both smoothing coefficients, α is mainly used to avoid zero division errors or extreme results in the calculation results when the change in the air static pressure value |ΔP(t)| is close to 0, β is mainly used to adjust the fluctuation intensity response of the change in the air static pressure value, especially to control the growth rate of the result when the nonlinear change is large, and γ is the nonlinear adjustment coefficient, which is used to control the sensitivity to drastic changes;

[0143] In order to quantify the overall intensity of static pressure fluctuations, the nonlinear change rate R(t) at each time point is integrated to generate the cumulative fluctuation energy. The calculation expression is as follows:

[0144]

[0145] , where E W Accumulated wave energy;

[0146] The accumulated wave energy E W Input into the nonlinear mapping function to generate the final air static pressure fluctuation anomaly index. The calculation expression is as follows:

[0147]

[0148] , where I P is the air static pressure fluctuation abnormality index, is the amplification factor, which is used to adjust the index range, τ is the energy threshold, which is used to control the sensitivity of the system to abnormal fluctuations, and tanh(·) is the hyperbolic tangent function, which maps the energy value to the range of 0-1;

[0149] The main purpose of mapping the accumulated energy values ​​to the range of 0-1 is to standardize and uniformly handle fluctuating energies of different scales, making the early warning system more flexible and easier to understand. In a coal mine environment, different detection windows may accumulate fluctuating energies of different magnitudes. For example, the energy generated by a small fluctuation in static pressure may differ greatly from that generated by a severe ventilation anomaly. Using the hyperbolic tangent function to map these energy values ​​to the range of [0,1] ensures that the output of all abnormal indicators is in the same standardized interval, which helps managers more intuitively assess the severity of risks and avoid judgments that are affected by excessively large or complex values.

[0150] In addition, compressing the energy value to the range of 0-1 can enhance the system's ability to distinguish between minor anomalies and violent fluctuations. The characteristic of the hyperbolic tangent function is that it changes slowly when the value is small, and gradually tends to saturation when the value is large, which makes the system more tolerant to low energy fluctuations and less likely to give false alarms; when the energy value is large, the index will quickly approach 1, clearly indicating that a serious anomaly has occurred. In this way, the system can filter out minor fluctuations to avoid false alarms, and quickly issue alarms for high-risk events, ensuring the accuracy and timeliness of early warnings.

[0151] It can be seen from the calculation expression of the air static pressure fluctuation anomaly index that the larger the performance value of the air static pressure fluctuation anomaly index generated after the abnormal analysis of the air static pressure fluctuation, the more violent the fluctuation of the air static pressure in the detection window, which means that the mine environment may have sudden or instantaneous changes, such as ventilation system disorder, gas release or local combustion. In this case, the monitoring method using a fixed sampling frequency is likely to miss these important changes in a short period of time, resulting in the system being unable to respond and warn of fire in a timely manner, increasing the risk of missed reports. Therefore, the larger the performance value of the air static pressure fluctuation anomaly index, the higher the probability that fixed frequency sampling cannot capture instantaneous changes. On the contrary, the smaller the performance value, the relatively stable fluctuation of the air static pressure and the less changes in a short period of time. At this time, although the fixed frequency sampling method still has limitations, the probability of missing key signals is relatively low.

[0152] The local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P Input into the pre-learned machine learning model, and generate the environmental safety change index ESCI through the machine learning model, based on the formula:

[0153]

[0154] , where f1 and f2 are the local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P The preset proportional coefficient, and f1 and f2 are both greater than 0.

[0155] It can be seen from the calculation expression of the environmental safety change index that the greater the performance value of the local humidity instantaneous fluctuation index generated after the abnormal analysis of the local humidity fluctuation, the greater the performance value of the air static pressure fluctuation abnormal index generated after the abnormal analysis of the air static pressure fluctuation, and the greater the performance value of the environmental safety change index, the greater the probability of sudden abnormalities in the mine environment, and vice versa, it means that the mine environment is within a safe and controllable range;

[0156] Based on the results of machine learning model evaluation, the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are divided into low-risk environmental information or high-risk environmental information, and the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are intelligently perceived;

[0157] The generated environmental safety change index is compared with the preset environmental safety change index reference threshold, and the analysis results are as follows:

[0158] If the environmental safety change index is greater than or equal to a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire early warning system are classified as high-risk environmental information;

[0159] If the environmental safety change index is less than a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire early warning system are classified as low-risk environmental information;

[0160] After the environmental information is judged as low-risk, it will continue to collect information at the preset environmental sampling frequency. After the environmental information is judged as high-risk, the sampling frequency will be dynamically adjusted based on the preset environmental sampling frequency to capture subtle changes in the environment in a timely manner.

[0161] For low-risk environment information, the system will collect data at a pre-set fixed sampling frequency, which means that the system will not increase the sampling frequency or perform additional data analysis operations in this case, thereby saving computing resources and energy consumption. The fixed sampling frequency is suitable for monitoring stable environments, ensuring the continuous collection of basic data, and providing support for long-term trend analysis and the establishment of a baseline for normal conditions.

[0162] For high-risk environmental information, the specific steps for dynamically adjusting the sampling frequency based on the preset environmental sampling frequency and timely capturing subtle changes in the environment are as follows:

[0163] According to the risk changes in the environment, the sampling frequency is dynamically adjusted through a nonlinear adjustment formula. The calculation expression for sampling frequency adjustment is as follows:

[0164]

[0165] In the formula, Sampling Fr represents the adjusted sampling frequency, Sampling Fr0 represents the preset sampling frequency, k is the adjustment coefficient used to control the sampling frequency increase, ESCI ref is the reference threshold of the environmental safety change index, and π is the sensitivity coefficient, which is used to adjust the degree of response to environmental changes;

[0166] The adjusted sampling frequency Smapling Fr will be applied to data collection in the next period, and environmental changes will be continuously monitored to form a feedback loop. If the risk level continues to increase, the sampling frequency will be further increased; if the environment returns to stability, the sampling frequency will gradually drop back to the preset level Sampling Fr0.

[0167] The present invention performs intelligent assessment of the environment through the generated environmental safety change index. When the system determines that it is high-risk environmental information, based on the preset sampling frequency, the system can dynamically adjust the sampling frequency so as to quickly respond and capture subtle changes in the early stages of a fire in a short period of time. For example, in the early stages of a fire, drastic fluctuations in the diffusion rate of methane gas or air static pressure may be extremely short-lived, and the traditional fixed sampling frequency cannot respond to these rapid fluctuations in time. Through this dynamic adjustment mechanism, the system can increase the sampling frequency, obtain key data more quickly, and ensure that early warnings are issued in a timely manner when risks increase, greatly improving the accuracy and timeliness of capturing signals in the early stages of a fire, thereby providing mine managers with sufficient time for early intervention.

[0168] For situations where environmental information is judged to be low-risk, the present invention uses a preset fixed sampling frequency to collect data and does not perform additional computing tasks. This design effectively saves the system's computing resources and energy consumption, because when the environment is stable and there is no obvious risk, the fixed sampling frequency can meet the data collection needs, and there is no need to dynamically adjust the frequency for over-sampling. This not only helps to monitor a stable environment for a long time and ensure continuous basic data collection, but also provides support for subsequent trend analysis and baseline establishment. This intelligent control mechanism enables the system to avoid unnecessary waste of resources while ensuring safety, thereby maximizing resource utilization. This dual control method maintains a fixed frequency when the risk is low and flexibly adjusts when the risk is high, which not only ensures the integrity of the data, but also improves the operating efficiency and sustainability of the system.

[0169] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A coal mine fire prevention early warning method based on intelligent monitoring, characterized in that: The following steps are involved: Obtain coal mine environmental information data parameters in real time through the coal mine fire early warning system, including humidity dynamic parameter information and air pressure parameter information; Perform an abnormal analysis on the collected humidity dynamic parameter information and air pressure parameter information, and input the humidity dynamic parameter information and air pressure parameter information after abnormal analysis into a pre-learned machine learning model, generate an environmental safety change index through the machine learning model, and use the environmental safety change index to intelligently evaluate the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system; Based on the results of machine learning model evaluation, the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are divided into low-risk environmental information or high-risk environmental information, and the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system are intelligently perceived; When environmental information is judged to be low-risk, it will continue to collect data at the preset environmental sampling frequency. When environmental information is judged to be high-risk, the sampling frequency will be dynamically adjusted based on the preset environmental sampling frequency to capture subtle changes in the environment in a timely manner.

2. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 1, characterized in that: The dynamic parameter information of humidity includes the change of methane gas diffusion rate, and the air pressure parameter information includes the fluctuation of air static pressure. The change of methane gas diffusion rate refers to the dynamic change of the diffusion rate of methane gas in different areas of the mine over time. The fluctuation of air static pressure refers to the drastic change of the static pressure of the air in the mine in a short period of time, which reflects the stability of the mine ventilation system and the equilibrium state of gas flow.

3. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 2, characterized in that: After performing an abnormal analysis on the local humidity fluctuation, a local humidity instantaneous fluctuation index is generated, and after performing an abnormal analysis on the air static pressure fluctuation, an air static pressure fluctuation abnormal index is generated. The local humidity instantaneous fluctuation index and the air static pressure fluctuation abnormal index generated after the abnormal analysis are input into a pre-learned machine learning model, and an environmental safety change index is generated through the machine learning model. The environmental safety change index is used to perform an intelligent evaluation of the coal mine environmental information data parameters obtained in real time by the coal mine fire warning system.

4. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 3, characterized in that: The logic of abnormal analysis of local humidity fluctuation and generation of local humidity instantaneous fluctuation index is as follows: In the detection window, collect local humidity data and calibrate the collected local humidity data as H(t i ), H(t i ) indicates that at time t i The local humidity value obtained by the i-th sampling at time t i ={t1, t2, t3, ..., t n }, where n represents the total number of sampling times; For every two consecutive sampling moments, the instantaneous humidity change rate is calculated, and the calculation expression is as follows: ， In the formula, ΔH(t i ) is the instantaneous humidity change rate, at time point t i The instantaneous humidity value at the time, that is, the instantaneous humidity recorded at the i-th sampling time, H(t i-1 ) is at the previous time point t i-1 The instantaneous humidity value is used to calculate the change in humidity, and Δt is the time interval between two samplings; Based on the instantaneous humidity change rate, the acceleration of humidity fluctuation is further calculated to capture the severity of humidity changes. The calculation expression is as follows: ， In the formula, ΔA H (t i ) indicates that at time t i The acceleration of humidity fluctuation at any moment, ΔH(t i-1 ) is at the previous time point t i-1 The instantaneous rate of change of humidity; The dynamic sensitivity function is introduced to capture the significant fluctuations of the acceleration of humidity fluctuations in a short period of time and amplify the instantaneous drastic changes. The calculation expression is as follows: ， In the formula, S(t i ) represents the time point t i The dynamic sensitivity value at the moment, λ is the sensitivity adjustment parameter, and θ is the threshold of humidity fluctuation; Then calculate the instantaneous anomaly coefficient, the calculation expression is as follows: R H (t i )=A H (t i )·S(t i ), In the formula, E H (t i ) represents the time point t i The instantaneous anomaly coefficient at the moment; The absolute values ​​of all instantaneous anomaly coefficients are accumulated to generate the local humidity instantaneous fluctuation index. The calculation expression is as follows: ， In the formula, I H It is the instantaneous fluctuation index of local humidity.

5. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 3, characterized in that: The logic of abnormal analysis of air static pressure fluctuation and generating air static pressure fluctuation abnormal index is as follows: Under the detection window, the air static pressure data of the mine is decomposed, and the time series data is divided into multiple small windows. The length of each sliding window is W. The expression is as follows: {P(t)|t∈[t0,t0+W]}, where P(t) is the air static pressure value at time point t, t0 is the start time of the sliding window, and t0+W is the end time of the sliding window; In each sliding window, the change of the air static pressure value between consecutive time points is calculated by differential operation, that is, the instantaneous fluctuation amplitude of the air static pressure value. The calculation expression is as follows: ΔP(t)=|P(t+Δt)-P(t)|, In the formula, ΔP(t) is the change in the static air pressure at time point t, and Δt is the time interval between two samplings; The detected change in air static pressure value ΔP(t) is further subjected to nonlinear analysis to calculate the change rate. The calculation expression is as follows: ， Where R(t) is the nonlinear change rate of the static air pressure at time point t, α and β are smoothing coefficients, and γ is the nonlinear adjustment coefficient; In order to quantify the overall intensity of static pressure fluctuations, the nonlinear change rate R(t) at each time point is integrated to generate the cumulative fluctuation energy. The calculation expression is as follows: ， In the formula, E W Accumulated wave energy; The accumulated wave energy E W Input into the nonlinear mapping function to generate the final air static pressure fluctuation anomaly index. The calculation expression is as follows: ， In the formula, I P is the air static pressure fluctuation abnormality index, is the amplification factor used to adjust the exponential range, τ is the energy threshold, and tanh(·) is the hyperbolic tangent function, which maps the energy value to the range of 0-1.

6. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 3, characterized in that: The local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P Input into the pre-learned machine learning model, and generate the environmental safety change index ESCI through the machine learning model, based on the formula: ， Where f1 and f2 are the local humidity instantaneous fluctuation index I H and air static pressure fluctuation anomaly index I P The preset proportional coefficient, and f1 and f2 are both greater than 0.

7. A coal mine fire prevention early warning method based on intelligent monitoring according to claim 6, characterized in that: The generated environmental safety change index is compared with the preset environmental safety change index reference threshold, and the analysis results are as follows: If the environmental safety change index is greater than or equal to a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire early warning system are classified as high-risk environmental information; If the environmental safety change index is less than a preset environmental safety change index reference threshold, the coal mine environmental information data parameters acquired in real time by the coal mine fire warning system are classified as low-risk environmental information.

Images