A mine disaster early warning method based on the combination of field monitoring and numerical simulation

Abstract

The present invention discloses a mine disaster early warning method based on the combination of on-site monitoring and numerical simulation, which relates to the technical field of mine disaster early warning, and comprises the following steps: obtaining on-site monitoring data of mine geological conditions through a normal sensor group laid in the mine, obtaining the mine location and historical data of geological information through a GIS system, numerically optimizing the on-site monitoring data of mine geological conditions, and inputting the on-site monitoring data and the mine location and historical data of geological information into a FLAC 3D The simulation software is used for simulation to obtain the numerical simulation data of mine disasters. According to the numerical simulation data of mine disasters, risk sensor groups are densely laid in high-risk areas of mines. The feedback data of the risk sensor groups are input into the short-term prediction formula to obtain risk prediction data. The risk prediction data is compared with the numerical simulation data of mine disasters, and warning signals are issued according to the difference comparison results, thus realizing early warning of mine disasters.

Description

Technical Field

[0001] The present invention relates to the technical field of mine disaster early warning, and in particular to a mine disaster early warning method based on the combination of on-site monitoring and numerical simulation. Background Art

[0002] In recent years, with the continuous expansion of the world's industrial scale and the continuous growth of human resource consumption, the demand for mineral mining has continued to expand. However, the frequent disasters in the mining industry have had a huge impact on the development of the industry and the life safety of mining workers. With the continuous improvement of the country's industry control policies, research on real-time monitoring of mine geological conditions and early warning of disasters is of great significance to protecting the lives of mining workers and improving mining operation efficiency.

[0003] At present, surface monitoring is mostly used for mine disaster warning. Sensors are laid on the surface of the mine to monitor the geological conditions of the mine. This method still has the following problems:

[0004] (1) Due to the uneven distribution of minerals inside the mountain and the complex topography of the mine, surface data monitoring cannot fully reflect the actual geological conditions underground. In addition, as mining operations proceed, the geological conditions of the mine are in a state of continuous change, and surface data monitoring cannot dynamically reflect changes in the geological conditions.

[0005] (2) The environment inside the mine, especially at the mining site, is extremely complex. It is difficult to install a large number of sensors and build a data feedback network. The reliability and real-time performance of the feedback data are not high. Summary of the invention

[0006] The purpose of the present invention is to provide a mine disaster early warning method based on the combination of field monitoring and numerical simulation to solve the problems of incomplete monitoring of actual geological conditions in mines, lack of dynamics and difficulty in sensor data transmission in the prior art.

[0007] In order to solve the above technical problems, the present invention specifically provides the following technical solutions:

[0008] A mine disaster early warning method based on the combination of field monitoring and numerical simulation is provided, comprising the following steps:

[0009] The on-site monitoring data of the mine geological conditions is obtained through a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the mine surface, inside the mine and underground, and the historical data of the mine location and geological information is obtained through the GIS system;

[0010] Numerical optimization is performed on the field monitoring data of the mine geological conditions, and the optimized field monitoring data and the historical data of the mine location and geological information are input into FLAC3D Simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters;

[0011] According to the numerical simulation data of mine disasters, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in high-risk areas of mines to form a risk sensor group;

[0012] The return data from the risk sensor group obtained through real-time monitoring is numerically optimized, the optimized return data from the risk sensor group is input into the short-term prediction formula to obtain risk prediction data, the risk prediction data is compared with the numerical simulation data of mine disasters for difference, and an early warning signal is issued based on the difference comparison result.

[0013] As a preferred solution of the present invention, a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the surface, inside and underground of the mine is used to obtain on-site monitoring data of the mine geological conditions, including:

[0014] According to the actual terrain of the mine, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid in the center of the area with slow terrain changes, at the junction of the area with drastic terrain changes, and in the center and boundary of the mining operation area;

[0015] The vibration sensor, temperature sensor, pressure sensor and humidity sensor are organized into a normal sensor group and connected to a comprehensive data processing center via a wired network;

[0016] The normal sensor group data return cycle is set to 12 hours, and data monitoring and data return are performed at the beginning of each cycle.

[0017] As a preferred solution of the present invention, numerical optimization is performed on the field monitoring data of the mine geological conditions, including:

[0018] Perform fast Fourier transform and Butterworth filtering on the data sent back by the vibration sensor and the pressure sensor to filter out the frequency components generated by human interference;

[0019] The data missing rate is calculated for the data returned by the temperature sensor and the data returned by the humidity sensor. If the data missing rate is less than 5%, no data interpolation operation is performed and the original data is retained. If the data missing rate is greater than 5% and less than 30%, the returned data at two time points on the left and right of the missing data are selected, and a polynomial is constructed using the Lagrange interpolation method. The value of the polynomial at the time point corresponding to the missing data is inserted into the returned data. If the data missing rate is greater than 30%, the data is retransmitted.

[0020] As a preferred solution of the present invention, the optimized field monitoring data and the mine location and geological information history data are input into the FLAC 3D The simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters, including:

[0021] The length L, width W and height H of the simulation model are set according to the mine location and the historical data of geological information, the number of layers and the layer height of the simulation model are set according to the optimized field monitoring data, and the vibration data, pressure data, temperature data and humidity data of each layer are marked;

[0022] The current normal sensor group feedback data cycle is divided into twelve equal parts into one-hour simulation cycles, and simulation runs are performed within the simulation cycles. The simulation results generated by each simulation cycle are used as the initial conditions for the next simulation cycle. In this way, a complete simulation is performed within the normal sensor group feedback data cycle, and the time nodes and simulation results of each simulation cycle are recorded to form numerical simulation data of mine disasters.

[0023] As a preferred solution of the present invention, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in high-risk areas of mines according to the numerical simulation data of mine disasters to form a risk sensor group, including:

[0024] According to the numerical simulation data of mine disasters within a complete data cycle of the normal sensor group, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid at the layers with the risk of mine disasters, the vibration sensors, temperature sensors, pressure sensors and humidity sensors are organized into a risk sensor group and connected to the edge data processing center through a wired network and a wireless network for edge data processing, and the edge data processing center is connected to the comprehensive data processing center through a wired network and a wireless network;

[0025] The risk sensor group is set to have a data return cycle of one minute, and data monitoring and data return are performed at the beginning of each cycle.

[0026] As a preferred solution of the present invention, numerical optimization is performed on the data returned by the risk sensor group obtained by real-time monitoring, including:

[0027] The data returned by the risk sensor group transmitted through two independent wired and wireless network paths are compared. If the data of the corresponding periods are the same, they are directly retained. If the data of the corresponding periods are different, the forward period time average value is used for data correction. The forward period time average value is expressed as:

[0028]

[0029] Among them, Y t-1 Indicates the result of data comparison and retention or data correction in the previous cycle, Y t-2 Indicates the result of data comparison and retention or data correction in the previous two cycles, Y t-3 Indicates the results of data comparison and retention or data correction in the previous three cycles;

[0030] The absolute data difference is calculated according to the forward period time average value, and the absolute data difference expression is:

[0031]

[0032] in, Indicates the absolute data difference between the wireless network return data of the current period and the average value of the forward period time, Indicates the absolute data difference between the wired network return data of the current cycle and the average value of the forward cycle time, U t Indicates the wireless network backhaul data in the current cycle, V t Indicates the wired network return data in the current cycle;

[0033] like Less than Then U t The value of Y t ,like Greater than Then V t The value of Y t .

[0034] As a preferred solution of the present invention, the optimized risk sensor group feedback data is input into the short-term prediction formula to obtain risk prediction data, including:

[0035] The optimized risk sensor group feedback data is processed using a short-term prediction formula, and the short-term prediction formula is expressed as:

[0036]

[0037] Among them, Y * (t N+1 ) represents the short-term forecast value of the next period, t i Represents a cycle sequence, and N represents the number of the current cycle.

[0038] As a preferred solution of the present invention, the risk prediction data is compared with the mine disaster numerical simulation data, and a warning signal is issued according to the difference comparison result, including:

[0039] The risk prediction data is numerically compared with the numerical simulation data of mine disasters. If the comparison result is that the numerical difference between the two is less than 20%, the warning state is a red warning. If the comparison result is that the numerical difference between the two is greater than 20% and less than 40%, the warning state is an orange warning. If the comparison result is that the numerical difference between the two is greater than 40%, the warning state is a green warning.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention distributes multiple types of data sensors on the mine surface, inside the mine and underground the mine, obtains historical data of the mine location and geological information through the GIS system, completely monitors the ore geological conditions through multi-source data, uses short-term prediction functions, conducts real-time monitoring and dynamic prediction of mine disasters, and improves the integrity and real-time performance of on-site monitoring data.

[0042] By combining on-site monitoring and numerical simulation, the areas where mine disasters occur are predicted, and sensors are selectively laid to reduce the number of sensors laid. By combining numerical optimization, edge computing and two network connection methods, the reliability and real-time performance of sensor data feedback are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0044] Figure 1 A flow chart of a mine disaster early warning method based on the combination of on-site monitoring and numerical simulation provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, the present invention provides a mine disaster early warning method based on the combination of on-site monitoring and numerical simulation, comprising the following steps:

[0047] The on-site monitoring data of the mine geological conditions is obtained through a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the mine surface, inside the mine and underground, and the historical data of the mine location and geological information is obtained through the GIS system;

[0048] In this embodiment, the mine types include coal mines, metal mines, non-metal mines and building material mines.

[0049] In this embodiment, the vibration sensor is an eddy current sensor, which can obtain a larger linear working range and higher sensitivity; the temperature sensor is a thermocouple sensor, which can obtain a wider temperature range and adapt to various geological environments; the pressure sensor is a piezoelectric pressure sensor, which can obtain a wider frequency band and higher sensitivity; the humidity sensor is a linear voltage output integrated humidity sensor, which can obtain a faster response speed and stronger anti-pollution ability.

[0050] Numerical optimization of the field monitoring data of the mine geological conditions is performed, and the optimized field monitoring data and the historical data of the mine location and geological information are input into FLAC 3D Simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters;

[0051] In this embodiment, due to the complex geological conditions at the mine site and the interference caused by manual mining operations, the original data collected by the sensor contains a large amount of interference data. Numerical optimization can eliminate the impact of the interference data and improve the accuracy of disaster warning.

[0052] In this embodiment, FLAC 3D The explicit Lagrangian algorithm and hybrid discrete partitioning technology can accurately simulate the plastic failure and elastic deformation of mining geology over a large range.

[0053] According to the numerical simulation data of mine disasters, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in the high-risk areas of the mine to form a risk sensor group;

[0054] In this embodiment, numerical simulation is used to simulate and calculate high-risk areas, and risk sensor groups are specifically laid in high-risk areas shown in the mine disaster numerical simulation data, rather than laying risk sensor groups over a large area in the entire mining area. This greatly reduces the number of sensors laid, saving costs while ensuring the accuracy of the early warning.

[0055] Perform numerical optimization on the data sent back by the risk sensor group obtained through real-time monitoring, input the optimized data sent back by the risk sensor group into the short-term prediction formula, obtain risk prediction data, compare the risk prediction data with the numerical simulation data of mine disasters, and issue early warning signals based on the comparison results;

[0056] In this embodiment, numerical optimization is performed on the real-time monitoring data to ensure the accuracy of the risk sensor group in monitoring the geological conditions in the high-risk areas of the mine.

[0057] The on-site monitoring data of the mine geological conditions is obtained through a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the mine surface, inside the mine and underground, including:

[0058] According to the actual terrain of the mine, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid in the center of the area with slow terrain changes, at the junction of the area with drastic terrain changes, and in the center and boundary of the mining operation area;

[0059] The vibration sensor, temperature sensor, pressure sensor and humidity sensor are organized into a normal sensor group and connected to the integrated data processing center via a wired network;

[0060] Set the normal sensor group data transmission cycle to 12 hours, and perform data monitoring and data transmission at the beginning of each cycle;

[0061] In this embodiment, the sensors are laid unevenly, reducing the number of sensors in areas with simple geological conditions and increasing the number of sensors in areas with complex geological conditions. This not only ensures the effective detection of the overall geological conditions of the mine, provides a data basis for accurate mine simulation, but also saves costs to the greatest extent.

[0062] In this embodiment, a vibration sensor, a temperature sensor, a pressure sensor and a humidity sensor are set up to form a normal sensor group to monitor the mine, forming a multimodal data set. The four sensor data corroborate each other, reducing the randomness of a single data source.

[0063] In this embodiment, the location of the mine and historical data of geological information are obtained through the GIS system, including the longitude and latitude of the mine, altitude, historical geological activity records, and surface vegetation coverage.

[0064] In this embodiment, the GIS system combines geography, cartography and satellite remote sensing data to form long-term geological observation historical data. The historical data is used to impose boundary constraints on numerical simulations, ensuring that the numerical simulations accurately describe the actual geological conditions of the mine.

[0065] Numerical optimization of field monitoring data of mine geology, including:

[0066] Perform fast Fourier transform and Butterworth filtering on the data sent back by the vibration sensor and the pressure sensor to filter out the frequency components generated by human interference;

[0067] The data missing rate of the temperature sensor return data and the humidity sensor return data is calculated. If the data missing rate is less than 5%, no data interpolation operation is performed and the original data is retained. If the data missing rate is greater than 5% and less than 30%, the return data at two time points on the left and right of the missing data are selected, and the Lagrange interpolation method is used to construct a polynomial. The value of the polynomial at the time point corresponding to the missing data is inserted into the return data. If the data missing rate is greater than 30%, the data is retransmitted.

[0068] In this embodiment, the vibration sensor return data and the pressure sensor return data are transformed from the time domain to the frequency domain through fast Fourier transform, and the multiple frequency components are separated. Then, a Butterworth filter is used to filter out the frequency components corresponding to the interference caused by artificial mining operations. The remaining frequency components are the frequency components generated by the geological condition transformation.

[0069] In this embodiment, since data retransmission takes a long time, the data transmission delay increases. In the part with low data loss rate, the sensor return data retains the data trend. Data interpolation can quickly restore the original data, reduce the overall data transmission delay and do not affect the data accuracy.

[0070] Input the optimized field monitoring data and historical data of mine location and geological information into FLAC 3D The simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters, including:

[0071] The length L, width W, and height H of the simulation model are set according to the mine location and historical data of geological information. The number of layers and layer height of the simulation model are set according to the optimized field monitoring data, and the vibration data, pressure data, temperature data, and humidity data of each layer are marked.

[0072] The current normal sensor group return data cycle is divided into twelve equal parts into one-hour simulation cycles, and simulation runs are performed within the simulation cycle. The simulation results generated by each simulation cycle are used as the initial conditions of the next simulation cycle. A complete simulation within the normal sensor group return data cycle is performed accordingly, and the time nodes and simulation results of each simulation cycle are recorded to form numerical simulation data of mine disasters;

[0073] In this embodiment, the simulation first needs to build a block elastic model, then set the overall volume and coordinate position of the mine, input the number of geological layers, layer height and layer density, bulk modulus, shear modulus of the mine, add gravity constraints and boundary constraints, and finally set the manual mining operation mode and start simulating the mining operation, and record the numerical simulation results.

[0074] According to the numerical simulation data of mine disasters, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in high-risk areas of mines to form a risk sensor group, including:

[0075] According to the numerical simulation data of mine disasters within the data cycle returned by a complete normal sensor group, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid at the layers with the risk of mine disasters, and the vibration sensors, temperature sensors, pressure sensors and humidity sensors are organized into a risk sensor group and connected to the edge data processing center through wired and wireless networks for edge data processing, and the edge data processing center is connected to the comprehensive data processing center through wired and wireless networks;

[0076] Set the risk sensor group to send back data for one minute, and perform data monitoring and data transmission at the beginning of each cycle;

[0077] In this embodiment, since the risk sensor group is laid at a high density and the sensor feedback cycle is extremely short, the amount of data that needs to be transmitted back is extremely large. A large amount of sensor feedback data is processed at the edge data processing center close to the sensor side, and then the simplified data is transmitted to the comprehensive data processing center. This greatly reduces the data processing task volume of the comprehensive data processing center and realizes effective data transmission inside mines with poor network conditions.

[0078] Numerical optimization of the data returned by the risk sensor group obtained through real-time monitoring, including:

[0079] Compare the data sent back by the risk sensor group through two independent wired and wireless network paths. If the data of the corresponding periods are the same, they are directly retained. If the data of the corresponding periods are different, the forward period time average value is used to correct the data. The forward period time average value expression is:

[0080]

[0081] Among them, Y t-1 Indicates the result of data comparison and retention or data correction in the previous cycle, Y t-2 Indicates the result of data comparison and retention or data correction in the previous two cycles, Y t-3 Indicates the results of data comparison and retention or data correction in the previous three cycles;

[0082] The absolute data difference is calculated according to the forward period time average value, and the absolute data difference expression is:

[0083]

[0084] in, Indicates the absolute data difference between the wireless network return data of the current period and the average value of the forward period time, Indicates the absolute data difference between the wired network return data of the current cycle and the average value of the forward cycle time, U t Indicates the wireless network backhaul data in the current cycle, V t Indicates the wired network return data in the current cycle;

[0085] like Less than Then U t The value of Y t ,like Greater than Then V t The value of Y t ;

[0086] In this embodiment, due to the poor network conditions inside the mine, a single network connection cannot guarantee that a large amount of feedback data from the risk sensor group can be transmitted completely and correctly. Two network connection methods are set up to back up each other, and multivariate data are used for mutual correction during numerical optimization, thereby improving the reliability and integrity of the feedback data.

[0087] Input the optimized risk sensor group feedback data into the short-term prediction formula to obtain risk prediction data, including:

[0088] Use the short-term prediction formula to process the optimized risk sensor group feedback data. The short-term prediction formula expression is:

[0089]

[0090] Among them, Y * (t N+1 ) represents the short-term forecast value of the next period, t i Represents a cycle sequence, and N represents the number of current cycles;

[0091] In this embodiment, the data return period of the risk sensor group is introduced as the minimum interval for short-term prediction, which realizes the precise prediction and real-time monitoring of mine disaster data. As the artificial mining operation proceeds, the mine geological conditions will change accordingly. The prediction results obtained by the short-term prediction formula can provide time margin for mine disaster warning for the prediction of mine geological conditions in future periods.

[0092] The risk prediction data is compared with the numerical simulation data of mine disasters, and early warning signals are issued according to the difference comparison results, including:

[0093] The risk prediction data is numerically compared with the numerical simulation data of mine disasters. If the comparison result shows that the difference between the two values ​​is less than 20%, the warning state is red warning. If the comparison result shows that the difference between the two values ​​is greater than 20% and less than 40%, the warning state is orange warning. If the comparison result shows that the difference between the two values ​​is greater than 40%, the warning state is green warning.

[0094] In this embodiment, the predicted data obtained by on-site monitoring and the simulated data obtained by numerical simulation are compared for difference. If the comparison result is that the numerical difference between the two is less than 20%, it indicates that the on-site monitoring data and the data of mining disasters obtained by numerical simulation are extremely similar, and the probability of mining disasters occurring in the future period is relatively high. The warning state is red warning, and the manual mining operations in the corresponding mining area should be stopped and withdrawn immediately. If the comparison result is that the numerical difference between the two is greater than 20% and less than 40%, it indicates that the similarity between the on-site monitoring data and the data of mining disasters occurring obtained by numerical simulation is general, but the probability of the data difference decreasing in the future period is relatively high. The warning state is orange warning, and the manual mining operations in the corresponding area can continue, but the priority of the feedback and data processing of the on-site monitoring data in the corresponding area needs to be increased. If the comparison result is that the numerical difference between the two is greater than 40%, it indicates that the similarity between the on-site monitoring data and the data of mining disasters occurring obtained by numerical simulation is relatively small, and the probability of mining disasters occurring in the future period is relatively small. The warning state is green warning.

[0095] The present invention distributes multiple types of data sensors on the mine surface, inside the mine and underground the mine, obtains historical data of the mine location and geological information through the GIS system, completely monitors the ore geological conditions through multi-source data, uses short-term prediction functions, conducts real-time monitoring and dynamic prediction of mine disasters, and improves the integrity and real-time performance of on-site monitoring data.

[0096] By combining on-site monitoring and numerical simulation, the areas where mine disasters occur are predicted, and sensors are selectively laid to reduce the number of sensors laid. By combining numerical optimization, edge computing and two network connection methods, the reliability and real-time performance of sensor data feedback are improved.

[0097] The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application. The protection scope of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and protection scope of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present application.

Claims

1. A mine disaster early warning method based on the combination of field monitoring and numerical simulation, characterized in that: The following steps are involved: The on-site monitoring data of the mine geological conditions is obtained through a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the mine surface, inside the mine and underground, and the historical data of the mine location and geological information is obtained through the GIS system; Numerical optimization is performed on the field monitoring data of the mine geological conditions, and the optimized field monitoring data and the historical data of the mine location and geological information are input into FLAC 3D Simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters; According to the numerical simulation data of mine disasters, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in high-risk areas of mines to form a risk sensor group; Numerical optimization is performed on the data returned by the risk sensor group obtained through real-time monitoring, and the optimized data returned by the risk sensor group is input into the short-term prediction formula to obtain risk prediction data, and the risk prediction data is compared with the numerical simulation data of mine disasters for difference, and an early warning signal is issued according to the difference comparison result; According to the numerical simulation data of mine disasters, vibration sensors, temperature sensors, pressure sensors and humidity sensors are densely laid in high-risk areas of mines to form a risk sensor group, including: According to the numerical simulation data of mine disasters within a complete data cycle of the normal sensor group, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid at the layers with the risk of mine disasters, the vibration sensors, temperature sensors, pressure sensors and humidity sensors are organized into a risk sensor group and connected to the edge data processing center through a wired network and a wireless network for edge data processing, and the edge data processing center is connected to the comprehensive data processing center through a wired network and a wireless network; The risk sensor group is set to transmit data back in a period of one minute, and data monitoring and data transmission are performed at the beginning of each period; Numerical optimization of the data returned by the risk sensor group obtained through real-time monitoring, including: The data returned by the risk sensor group transmitted through two independent wired and wireless network paths are compared. If the data of the corresponding periods are the same, they are directly retained. If the data of the corresponding periods are different, the forward period time average value is used for data correction. The forward period time average value is expressed as: ； Among them, Y t-1 Indicates the result of data comparison and retention or data correction in the previous cycle, Y t-2 Indicates the result of data comparison and retention or data correction in the previous two cycles, Y t-3 Indicates the results of data comparison and retention or data correction in the previous three cycles; The absolute data difference is calculated according to the forward period time average value, and the absolute data difference expression is: ； in, Indicates the absolute data difference between the wireless network return data of the current period and the average value of the forward period time, Indicates the absolute data difference between the wired network return data of the current cycle and the average value of the forward cycle time, U t Indicates the wireless network backhaul data in the current cycle, V t Indicates the wired network return data in the current cycle; like Less than , then U t The value of Y t ,like Greater than , then V t The value of Y t ; Input the optimized risk sensor group feedback data into the short-term prediction formula to obtain risk prediction data, including: The optimized risk sensor group feedback data is processed using a short-term prediction formula, and the short-term prediction formula is expressed as: ； in, Represents the short-term forecast value of the next period, t i Represents a cycle sequence, and N represents the number of the current cycle.

2. A mine disaster early warning method based on combining on-site monitoring and numerical simulation according to claim 1, characterized in that: The on-site monitoring data of the mine geological conditions is obtained through a normal sensor group consisting of vibration sensors, temperature sensors, pressure sensors and humidity sensors distributed on the mine surface, inside the mine and underground, including: According to the actual terrain of the mine, vibration sensors, temperature sensors, pressure sensors and humidity sensors are laid in the center of the area with slow terrain changes, at the junction of the area with drastic terrain changes, and in the center and boundary of the mining operation area; The vibration sensor, temperature sensor, pressure sensor and humidity sensor are organized into a normal sensor group and connected to a comprehensive data processing center via a wired network; The normal sensor group data return cycle is set to 12 hours, and data monitoring and data return are performed at the beginning of each cycle.

3. A mine disaster early warning method based on combining on-site monitoring and numerical simulation according to claim 2, characterized in that: Numerical optimization of the field monitoring data of the mine geology, including: Perform fast Fourier transform and Butterworth filtering on the data sent back by the vibration sensor and the pressure sensor to filter out the frequency components generated by human interference; The data missing rate is calculated for the data returned by the temperature sensor and the humidity sensor. If the data missing rate is less than 5%, no data interpolation operation is performed and the original data is retained. If the data missing rate is greater than 5% and less than 30%, the returned data at two time points on the left and right of the missing data are selected, and a polynomial is constructed using the Lagrange interpolation method. The value of the polynomial at the time point corresponding to the missing data is inserted into the returned data. If the data missing rate is greater than 30%, the data is retransmitted.

4. A mine disaster early warning method based on combining on-site monitoring and numerical simulation according to claim 3, characterized in that: Input the optimized field monitoring data and the historical data of the mine location and geological information into FLAC 3D The simulation software is used to simulate in time series to obtain numerical simulation data of mine disasters, including: Set the length of the simulation model according to the mine location and historical data of geological information ,width ,high , according to the optimized on-site monitoring data, set the number of layers and layer heights of the simulation model, and mark the vibration data, pressure data, temperature data and humidity data of each layer; The current normal sensor group feedback data cycle is divided into twelve equal parts into one-hour simulation cycles, and simulation runs are performed within the simulation cycles. The simulation results generated by each simulation cycle are used as the initial conditions for the next simulation cycle. In this way, a complete simulation is performed within the normal sensor group feedback data cycle, and the time nodes and simulation results of each simulation cycle are recorded to form numerical simulation data of mine disasters.

5. The mine disaster early warning method based on the combination of on-site monitoring and numerical simulation according to claim 1 is characterized in that: The risk prediction data is compared with the numerical simulation data of mine disasters, and an early warning signal is issued according to the difference comparison result, including: The risk prediction data is numerically compared with the numerical simulation data of mine disasters. If the comparison result is that the numerical difference between the two is less than 20%, the warning state is a red warning. If the comparison result is that the numerical difference between the two is greater than 20% and less than 40%, the warning state is an orange warning. If the comparison result is that the numerical difference between the two is greater than 40%, the warning state is a green warning.

Images