Coal seam drilling gas concentration monitoring method and system based on multi-sensor fusion

Abstract

The invention discloses a coal seam drilling gas concentration monitoring method and system based on multi-sensor fusion, and the method comprises the steps: analyzing the gas concentration risks of different depth layers according to preset coal seam drilling information, and dividing the coal seam into multiple stages; for each stage of coal seam, a plurality of monitoring areas are divided by analyzing the gas concentration gradients of different areas; a plurality of sensors are arranged in each monitoring area, and a sensor network is constructed by analyzing the communication relationship among the sensors; carrying out concentration monitoring through a preset concentration safety monitoring model according to data acquired by each sensor network, and carrying out early warning when the gas concentration has an abnormal result; the coal seam is subjected to hierarchical division and regional division, the sensor network is arranged in each region to monitor the gas concentration, and the collected data is corrected in real time among the sensor networks, so that the accuracy of the gas concentration monitoring result can be improved, and the gas accident can be effectively prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine monitoring, and particularly to a method and system for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion. Background Art

[0002] Traditional methods for monitoring the gas concentration in coal seam boreholes directly monitor the gas concentration in the coal seam, lacking hierarchical and regional monitoring of the coal seam. The measured results cannot accurately reflect the actual gas concentration at different positions in the coal seam, resulting in inaccurate monitoring results; during the process of sensor data transmission, the poor signal transmission quality in the underground coal seam is not considered, resulting in insufficient accuracy of the collected data and difficulty in obtaining real gas concentration data.

[0003] The prior art has the following problems: lacking hierarchical and regional monitoring of the coal seam, the measured results cannot accurately reflect the actual gas concentration at different positions in the coal seam, resulting in inaccurate monitoring results; during the process of sensor data transmission, the poor signal transmission quality in the underground coal seam is not considered, resulting in insufficient accuracy of the collected data; to solve at least one of the above problems, the present invention proposes a method and system for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the main purpose of the present invention is to provide a method and system for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion, which can effectively solve the problems in the background art. The specific technical solutions of the present invention are as follows:

[0005] A method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion includes:

[0006] Analyze the gas concentration risks of different depth layers according to the preset coal seam borehole information, and divide them into multiple levels of coal seams;

[0007] For each level of coal seam, divide multiple monitoring areas by analyzing the gas concentration gradients in different regions;

[0008] Set multiple sensors in each monitoring area, divide the main sensor and slave sensors by analyzing the communication relationship between the sensors and combining the voting results between the sensors, and construct a sensor network;

[0009] According to the data collected by each sensor network, perform concentration monitoring through a preset concentration safety monitoring model, and give an alarm when the gas concentration shows an abnormal result to achieve gas concentration monitoring.

[0010] Specifically, the step of analyzing the gas concentration risks of different depth layers according to the preset coal seam borehole information and dividing them into multiple levels of coal seams includes:

[0011] Obtain the coal seam permeability, historical gas emission volume, and borehole trajectory data according to the preset borehole information for each layer;

[0012] According to the coal seam permeability, historical gas emission volume, and borehole trajectory data, vertically cluster the coal seam and divide it into the first-risk-level coal seam, the second-risk-level coal seam, and the third-risk-level coal seam;

[0013] When the gas concentration increase rate of the coal seam exceeds the preset concentration change threshold, upgrade the coal seam;

[0014] When the gas concentration decrease rate of the coal seam exceeds the preset concentration change threshold, downgrade the coal seam.

[0015] Specifically, for each level of coal seam, divide multiple monitoring areas by analyzing the gas concentration gradient in different regions, including:

[0016] For each level of coal seam, set the initial grid size according to the preset borehole spacing;

[0017] According to the initial grid size, divide the coal seam into multiple grid areas;

[0018] Calculate the gas concentration gradient value between adjacent grids according to the gas concentration in each grid area;

[0019] When the gas concentration gradient value is less than the preset gradient threshold, merge the corresponding adjacent grids to obtain multiple monitoring areas.

[0020] Specifically, set multiple sensors in each monitoring area, and divide the master sensor and the slave sensor and construct a sensor network by analyzing the communication relationship between the sensors and combining the voting results of the sensors, including:

[0021] Set multiple sensors in each monitoring area, and divide the master sensor and the slave sensor according to the communication relationship between different sensors and the voting results of the sensors;

[0022] Construct a communication link between the master sensor and the slave sensor according to the positional relationship and functional dependence between the master sensor and the slave sensor to obtain a sensor network.

[0023] Specifically, set multiple sensors in each monitoring area, and divide the master sensor and the slave sensor according to the communication relationship between different sensors and the voting results of the sensors, including:

[0024] In each monitoring area, set the initial master sensor according to the performance data of each sensor;

[0025] Within a preset time period, calculate the weight value of each sensor through a preset dynamic weight calculation model;

[0026] When the weight value of the initial main sensor is less than the preset first weight threshold, the sensors with weight values greater than the preset second weight threshold are used as the candidate sensor set, where the preset first weight threshold is less than the preset second weight threshold;

[0027] According to the sensor performance in the candidate sensor set, each sensor votes for each sensor in the candidate sensor set, and the sensor with the most votes is used as the updated main sensor;

[0028] The initial main sensor serves as a slave sensor and communicates with the updated main sensor.

[0029] Specifically, according to the positional relationship and functional dependence between the main sensor and the slave sensor, a communication link is constructed between the main sensor and the slave sensor to obtain a sensor network, including:

[0030] An adjacency matrix is constructed according to the position distance of each sensor;

[0031] According to the adjacency matrix, the two sensors with the shortest interval distance are connected;

[0032] When the number of interval sensors for communication between the slave sensor and the main sensor is greater than the preset number threshold, the communication link is updated, and the path with the number of interval sensors less than the preset number threshold and the shortest overall distance of the communication link is selected as the updated communication link;

[0033] According to the updated communication link, communication is established between each slave sensor and the main sensor to obtain a sensor network.

[0034] Specifically, according to the data collected by each sensor network, concentration monitoring is performed through a preset concentration safety monitoring model, and an alarm is issued when the gas concentration shows an abnormal result to achieve gas concentration monitoring, including:

[0035] According to the data collected by each sensor network, the sensor nodes with data transmission failures in the sensor network are detected and data correction is performed to obtain multi-sensor corrected data;

[0036] The multi-sensor corrected data is hierarchically fused to obtain fused data;

[0037] According to the fused data, concentration monitoring is performed through a preset concentration safety monitoring model, and an alarm is issued when the gas concentration shows an abnormal result to achieve gas concentration monitoring.

[0038] Specifically, detecting sensor nodes with data transmission failures in the sensor network based on the data collected by each sensor network and performing data correction to obtain multi-sensor corrected data, including:

[0039] Based on the data collected by each sensor network, sensor nodes with a packet loss rate greater than a preset loss threshold are regarded as failed sensor nodes;

[0040] By calculating the distance weighted values of a preset number of adjacent non-failed sensor nodes of the failed sensor nodes, a spatial correction value is obtained;

[0041] Based on the historical collected data of the failed sensor nodes, the data is predicted through a preset trend prediction model to obtain a predicted value;

[0042] The spatial correction value and the predicted value are fused through a preset weight to obtain multi-sensor corrected data.

[0043] Specifically, the multi-sensor corrected data is hierarchically fused to obtain fused data, including:

[0044] According to the distances between the master sensor node and each slave sensor node, the data of each slave sensor in the multi-sensor corrected data is calibrated to obtain slave sensor calibrated data;

[0045] The master sensor data and the slave sensor calibrated data are fused to obtain regional fused data;

[0046] Data with different trends in the gas concentration gradient between the gas concentration in the regional fused data and the adjacent regions is corrected to obtain fused data.

[0047] A coal seam borehole gas concentration monitoring system based on multi-sensor fusion, used to implement the coal seam borehole gas concentration monitoring method based on multi-sensor fusion, includes:

[0048] A coal seam division module, analyzing the gas concentration risks of different depth layers according to the preset coal seam borehole information and dividing them into multiple levels of coal seams;

[0049] A regional division module, for each level of coal seam, dividing multiple monitoring regions by analyzing the gas concentration gradients of different regions;

[0050] A sensor network construction module, setting multiple sensors in each monitoring region, dividing the master sensor and the slave sensors by analyzing the communication relationships between the sensors and combining the voting results between the sensors, and constructing a sensor network;

[0051] The gas concentration monitoring module monitors the gas concentration according to the data collected by each sensor network through a preset concentration safety monitoring model, and gives an early warning when abnormal results of the gas concentration occur, so as to realize the monitoring of the gas concentration.

[0052] Compared with the prior art, the present application has the following beneficial effects:

[0053] By hierarchically dividing and regionally dividing the coal seam, setting up a sensor network in each region to monitor the gas concentration, and performing real-time correction on the collected data between the sensor networks, the present application can obtain data that accurately reflects the gas concentration changes in different regions. Identifying the gas concentration based on accurate monitoring data can improve the accuracy of the gas concentration monitoring results, effectively prevent the occurrence of gas accidents, and ensure the safety of coal miners and the safe and stable operation of coal mine production. Description of the Drawings

[0054] Figure 1 It is the working flow chart of the coal seam borehole gas concentration monitoring method based on multi-sensor fusion in Embodiment 1 of the present invention;

[0055] Figure 2 It is the schematic diagram of the coal seam monitoring area division in Embodiment 1 of the present invention;

[0056] Figure 3 It is the schematic diagram of the sensor network construction in Embodiment 1 of the present invention;

[0057] Figure 4 It is the schematic diagram of the structure of the coal seam borehole gas concentration monitoring system based on multi-sensor fusion in Embodiment 2 of the present invention. Detailed Embodiments

[0058] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be given with reference to the accompanying drawings of the specification.

[0059] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0060] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that exclude each other with other embodiments.

[0061] Embodiment 1:

[0062] This embodiment provides a method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion, as Figure 1 shown. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion includes:

[0063] S101. Analyze the gas concentration risks of different depth layers according to the preset coal seam borehole information, and divide them into multiple levels of coal seams;

[0064] S102. For each level of coal seam, divide multiple monitoring areas by analyzing the gas concentration gradients in different regions;

[0065] S103. Set multiple sensors in each monitoring area. By analyzing the communication relationships between the sensors and combining the voting results among the sensors, divide the master sensor and slave sensors, and construct a sensor network;

[0066] S104. According to the data collected by each sensor network, perform concentration monitoring through a preset concentration safety monitoring model, and give an early warning when abnormal gas concentration results occur to achieve gas concentration monitoring.

[0067] Through the multi-level coal seam division, monitoring area division based on gas gradient, sensor network construction, and multi-level fusion and monitoring early warning of multi-sensor data, this embodiment realizes the comprehensive and accurate monitoring of the gas concentration in coal seam boreholes compared with the traditional method of monitoring the gas concentration from the whole coal seam. It covers coal seams at different depths and in different regions, can accurately obtain gas concentration information, quickly respond when the gas concentration is abnormal, and provides a strong guarantee for coal mine safety production.

[0068] In this embodiment, first, evaluate the gas concentration risks of coal seams at different depths by analyzing the coal seam permeability, historical gas emission volume, and borehole trajectory data. Based on the coal seam permeability, historical gas emission volume, and borehole trajectory data, use the vertical clustering algorithm to divide coal seams with similar risk characteristics into the same level; at the same time, dynamically adjust the coal seam level according to the real-time change of the gas concentration. By setting a concentration change threshold, when the gas concentration increase rate or decrease rate exceeds the concentration change threshold, dynamically adjust the coal seam level. Through the precise layering and dynamic adjustment of the coal seam, the change trend of the gas concentration can be reflected in a timely manner, and the accuracy and effectiveness of risk assessment can be improved.

[0069] Specifically, after dividing the coal seams, regional division is carried out for each grade of coal seam. The coal seam is meshed by setting the initial mesh size to obtain multiple mesh regions. Then, the gas concentration gradient value between adjacent meshes is calculated to analyze the degree of change of gas concentration in space. When the gas concentration gradient value is less than the preset gradient threshold, it indicates that the gas concentration difference between adjacent meshes is small, and the adjacent meshes are merged, so that a more reasonable monitoring area can be obtained, which can more accurately reflect the distribution of gas concentration in different regions, avoid monitoring blind spots or duplicate monitoring caused by simple division, and improve the monitoring efficiency.

[0070] Specifically, multiple sensors are set in each monitoring area. By analyzing the communication relationship between the sensors, the master sensor and the slave sensors are divided. The master sensor is responsible for collecting and processing the data of the slave sensors and communicating with the external system. According to the positional relationship and functional dependence of the master and slave sensors, a communication link is constructed to build a sensor network. At the same time, the master sensor is dynamically adjusted according to the change of the master sensor performance, which can ensure the reliability and stability of the sensor network and ensure that the network can operate continuously and efficiently when the sensor performance changes. By dynamically dividing the master and slave sensors and optimizing the communication link, the adaptability and reliability of the sensor network are improved, which can ensure the efficient transmission and accurate collection of data, provide accurate data for gas concentration monitoring, and ensure the normal operation of the monitoring system.

[0071] Specifically, during the process of the sensor network transmitting data, the data transmission result is corrected. First, the sensor nodes with data transmission failure in the sensor network are detected, and the data of the failed sensor nodes are corrected by the adjacent sensor nodes. The corrected data is hierarchically fused, including correcting data according to the distance between the master and slave sensors, fusing the data of the master and slave sensors, and correcting abnormal data, etc. The fused data is input into a preset concentration safety monitoring model for concentration monitoring. When an abnormal result of gas concentration is detected, a warning is issued in time to achieve effective monitoring of gas concentration. By detecting and correcting the sensor nodes with data transmission failure and hierarchically fusing the data of multiple sensors, the accuracy and reliability of the data are improved, ensuring that the monitoring result can truly reflect the gas concentration situation. The timely and effective warning mechanism can quickly notify relevant personnel when the gas concentration is abnormal, facilitating timely measures to prevent the occurrence of gas accidents and ensuring the safety of coal mine production.

[0072] By hierarchically and regionally dividing the coal seam, a sensor network is set up in each region to monitor the gas concentration, and the collected data is corrected in real time between the sensor networks, enabling the acquisition of data that accurately reflects the gas concentration changes in different regions. Based on the accurate monitoring data, the gas concentration is identified, which can improve the accuracy of the gas concentration monitoring results, effectively prevent the occurrence of gas accidents, and ensure the safety of coal miners and the safe and stable operation of coal mine production.

[0073] Further, based on the preset coal seam borehole information, the gas concentration risks at different depths are analyzed, and the coal seam is divided into multiple levels, including:

[0074] S201. Obtain the coal seam permeability, historical gas emission volume, and borehole trajectory data according to the preset borehole information for each layer;

[0075] S202. According to the coal seam permeability, historical gas emission volume, and borehole trajectory data, perform vertical clustering on the coal seam to divide it into the first-risk-level coal seam, the second-risk-level coal seam, and the third-risk-level coal seam;

[0076] S203. When the gas concentration increase rate of the coal seam exceeds the preset concentration change threshold, upgrade the coal seam;

[0077] S204. When the gas concentration decrease rate of the coal seam exceeds the preset concentration change threshold, downgrade the coal seam.

[0078] In this embodiment, through geological exploration means, specifically laboratory core analysis, the core samples taken from the coal seam are tested, and the permeability value of the coal seam is calculated using relevant principles such as Darcy's law; historical data is retrieved from the coal mine production record system, including gas emission volume monitoring data for different time periods and different mining areas; during the borehole construction process, a borehole trajectory measuring instrument, such as a measurement-while-drilling system or a wired single-shot inclinometer, is used to measure parameters such as the depth, inclination angle, and azimuth angle of the borehole in real time, and detailed borehole trajectory data is generated through software analysis, which can visually display the borehole's trend and spatial distribution in the coal seam. By analyzing multi-dimensional data, the one-sidedness that may be brought by a single data source is avoided, making the evaluation process more scientific and comprehensive.

[0079] Specifically, the obtained coal seam permeability, historical gas emission volume, and borehole trajectory data are cleaned to remove outliers and missing values. For missing values, methods such as mean filling and regression prediction can be used for supplementation; for outliers, they can be corrected or removed according to the data distribution. Then, the data is standardized to convert data with different dimensions to the same scale. The processed data is clustered in the vertical direction using a clustering algorithm, and coal seams with similar characteristics are grouped into the same category, thereby dividing coal seams into different risk levels. For example, a coal seam with high permeability, high historical gas emission volume, and a complex borehole trajectory has a relatively high gas concentration risk and will be classified into a higher risk level. The first risk-level coal seam, the second risk-level coal seam, and the third risk-level coal seam are obtained through clustering. By classifying the risk levels of coal seams, coal mine managers can quickly and intuitively understand the risk status of coal seams in different areas.

[0080] At the same time, the gas concentration in the coal seam is in a dynamic change process. The classification results and performance of the coal seam are dynamically adjusted according to the change of the gas concentration. When the increase rate exceeds the preset concentration change threshold, it indicates that the gas concentration risk of this coal seam has increased significantly in the short term, and the original risk level can no longer accurately reflect the current actual risk status. The risk level of the coal seam is upgraded to attract the high attention of relevant personnel and take more stringent preventive measures in a timely manner. The preset concentration change threshold can be set according to the actual situation and safety standards of the coal mine. For example, it is set that the gas concentration increases by 5% per hour. When the decrease rate of the gas concentration in the coal seam exceeds the preset concentration change threshold, it indicates that the gas concentration risk of this coal seam is decreasing, and the original risk level is relatively too high. The risk level of the coal seam is downgraded, which can more accurately reflect the current actual risk status of the coal seam, rationally allocate monitoring and preventive resources, and avoid waste of resources. By dynamically adjusting the risk level of the coal seam, the resource allocation is made more reasonable and efficient.

[0081] Furthermore, for each level of coal seam, by analyzing the gas concentration gradient in different areas, multiple monitoring areas are divided, including:

[0082] S301: For each level of coal seam, set the initial grid size according to the preset borehole spacing;

[0083] S302: According to the initial grid size, divide the coal seam into grid regions to obtain multiple grid regions;

[0084] S303: Calculate the gas concentration gradient value between adjacent grids according to the gas concentration in each grid region;

[0085] S304: When the gas concentration gradient value is less than the preset gradient threshold, merge the corresponding adjacent grids to obtain multiple monitoring areas.

[0086] Such asFigure 2 , in this embodiment, the regional division is carried out for each level of coal seam, and the regional division result is dynamically optimized according to the gas concentration in the grid area. The optimized monitoring area division reduces unnecessary monitoring duplication, avoids resource waste, and at the same time highlights the key monitoring of the areas with obvious gas concentration changes, improving the efficiency and response speed of the entire gas concentration monitoring system.

[0087] In this embodiment, at each level of coal seam, the initial grid size is set according to the borehole spacing. The borehole spacing is closely related to the accuracy of coal seam gas concentration monitoring. The smaller the borehole spacing, the denser the gas concentration data obtained. For example, in the mining plan of a certain coal mine, it is clearly stipulated that the borehole spacing is 5 meters, and the initial grid size is set as an integer multiple or an appropriate ratio of the borehole spacing. The initial grid size is set to 2 times the borehole spacing, that is, 10 meters. Setting the initial grid size according to the borehole spacing can make full use of the existing borehole data, reduce the data acquisition cost, and at the same time ensure that each grid area has sufficient data to support the gas concentration analysis.

[0088] Specifically, grid division is carried out according to the initial grid size. The coal seam is divided into regular grid areas according to the initial grid size, transforming the complex coal seam gas concentration distribution problem into the analysis of the gas concentration in multiple discrete grid areas. Each grid area is regarded as a relatively independent unit, and multiple grid areas are obtained. Through the regular grid area division, it is convenient to calculate the gas concentration gradient value between adjacent grids, providing data support for accurately dividing the monitoring area; calculate the gas concentration data in each grid area, and obtain the average gas concentration value of each grid area through interpolation calculation based on the borehole data. For each pair of adjacent grid areas, calculate the gas concentration gradient value. By calculating the gas concentration gradient value, it is possible to timely discover the areas with drastic gas concentration changes, which helps to take preventive measures in advance and ensure the safe production of coal mines.

[0089] Optimize the grid area according to the calculated gas concentration gradient value. When the gas concentration gradient value between adjacent grids is less than the preset gradient threshold, it means that the gas concentration difference between adjacent grid areas is small, and they are merged into one monitoring area. By merging the monitoring areas, the number of monitoring areas can be reduced, and at the same time, it will not affect the accuracy of the gas concentration distribution area division; for adjacent grids with a gas concentration gradient value greater than the preset gradient threshold, they are retained as independent monitoring areas. The preset gradient threshold can be set according to the actual situation of the coal mine, historical monitoring data, and safety standards. For example, the preset gradient threshold is set to 0.5 (unit: concentration value / meter); by merging adjacent grids with similar gas concentrations, the number of monitoring areas is reduced, and the complexity of the monitoring system is lowered.

[0090] Further, a plurality of sensors are arranged in each monitoring area, and the master sensor and the slave sensor are divided by analyzing the communication relationship between the sensors and combining the voting results between the sensors, and a sensor network is constructed, including:

[0091] S401. Arrange a plurality of sensors in each monitoring area, and divide the master sensor and the slave sensor according to the communication relationship between different sensors and the voting results between the sensors;

[0092] S402. Construct a communication link between the master sensor and the slave sensor according to the positional relationship and functional dependence between the master sensor and the slave sensor to obtain a sensor network.

[0093] Such as Figure 3 , in this embodiment, a plurality of sensors are arranged in each monitoring area. According to the differences in the performance and communication capabilities of different sensors, the sensors are reasonably divided to achieve efficient data collection and transmission, and are divided into master sensors and slave sensors. Among them, the master sensor is responsible for the core tasks of data aggregation, processing, and communication with the external system, and the slave sensor is responsible for collecting local data and transmitting it to the master sensor. By analyzing the communication relationship between the sensors, including signal strength, transmission rate, stability, etc., combined with the performance of the sensors themselves (measurement accuracy, power consumption, etc.), the master sensor and the slave sensor are divided, and the allocation of the sensors is dynamically adjusted according to the real-time performance of the master sensor; by dynamically adjusting the master sensor, the sensor network can adapt to environmental changes and sensor performance degradation. When the performance of the master sensor decreases, the system can replace the master sensor in time to ensure the continuous and stable operation of the network, and improve the accuracy of the gas concentration monitoring data.

[0094] Specifically, according to the positional relationship and functional dependence between the master sensor and the slave sensor, a communication link is constructed between the master sensor and the slave sensor. The positional relationship reflects the distance and path of data transmission. The closer the distance, the smaller the signal attenuation and interference during the transmission process; the functional dependence requires that the slave sensor can accurately and timely transmit the data to the master sensor for processing. By comprehensively analyzing the positional relationship and functional dependence, an optimal communication link is constructed to ensure the efficiency and stability of data transmission in the sensor network, so that the master sensor can quickly collect the data of all slave sensors and realize the real-time monitoring of the gas concentration in the monitoring area. By constructing a communication link, signal attenuation and interference can be reduced, the probability of errors occurring during data transmission can be reduced, the stability of the sensor network can be enhanced, and reliable operation can be ensured even in a complex coal mine environment.

[0095] Further, arranging a plurality of sensors in each monitoring area, and dividing the master sensor and the slave sensor according to the communication relationship between different sensors and the voting results between the sensors, includes:

[0096] S501. In each monitoring area, set an initial master sensor based on the performance data of each sensor;

[0097] S502: Calculate the weight value of each sensor using a preset dynamic weight calculation model within a preset time period;

[0098] S503: When the weight value of the initial primary sensor is less than a preset first weight threshold, sensors having weight values greater than a preset second weight threshold are selected as a candidate sensor set, wherein the preset first weight threshold is less than the preset second weight threshold;

[0099] S504: Based on the sensor performance in the candidate sensor set, each sensor votes for each sensor in the candidate sensor set, and the sensor with the most votes is used as the updated main sensor;

[0100] S505 : The initial master sensor serves as a slave sensor and communicates with the updated master sensor.

[0101] In this embodiment, in each monitoring area, an initial main sensor is set according to the different performance data of the sensors, and the performance data of each sensor is obtained through the parameter configuration file provided by the sensor. A weight is set for each performance indicator according to the actual needs of gas concentration monitoring; for example, the measurement accuracy weight accounts for 40%, the data transmission rate weight accounts for 30%, the battery life weight accounts for 20%, and the anti-interference ability weight accounts for 10%; the performance indicator values of each sensor are multiplied by the corresponding weights, and then summed to obtain the comprehensive performance score of each sensor; the comprehensive performance scores of all sensors are compared, and the sensor with the highest score is selected as the initial main sensor. When there are multiple sensors with the same comprehensive performance scores, the sensor with a higher data transmission rate is preferentially selected as the initial main sensor; selecting the sensor with the best performance as the initial main sensor can efficiently collect and process data from the sensors in the early stage of sensor network operation and transmit it to the external monitoring system in a timely manner.

[0102] Specifically, during the data transmission process, the master sensor and slave sensors are dynamically adjusted according to the changes in sensor performance. Considering the complex environment in coal mines, the performance of sensors changes over time, including unstable operation caused by battery power consumption, measurement accuracy affected by equipment aging, and data transmission quality changed by environmental interference. Adjusting the master sensor can ensure that the master sensor always has the optimal data transmission performance, guaranteeing the accuracy and efficiency of data transmission. According to the actual requirements of coal mine gas concentration monitoring and the law of sensor performance changes, a suitable time period is set. In this embodiment, the time period is set to one hour. The preset dynamic weight calculation model synthesizes multiple factors, including the current measurement error of the sensor, data transmission success rate, remaining battery power, signal strength, etc. The weights of each performance index can be set according to the actual calculation accuracy requirements. For example, the measurement accuracy weight is 0.4, the data transmission weight is 0.3, the battery weight is 0.2, and the signal weight is 0.1. During each time period, data such as the measurement error, data transmission success rate, remaining battery power, and signal strength of each sensor are collected in real time. The collected data is input into the preset dynamic weight calculation model, and the weight value of each sensor in this time period is calculated through weighted calculation. By dynamically calculating the weight value, the sensor network can adapt to the complex and changeable environment underground in coal mines, timely reflect the dynamic changes in sensor performance, and ensure that the network always maintains a good operating state.

[0103] Specifically, according to the actual situation of coal mine gas concentration monitoring and sensor performance standards, a preset first weight threshold and a preset second weight threshold are set. For example, the preset first weight threshold is set to 0.6, and the preset second weight threshold is set to 0.7. After calculating the weight values of the sensors in each preset time period, the weight value of the initial master sensor is compared with the preset first weight threshold, and at the same time, the weight values of other sensors are compared with the preset second weight threshold. When the weight value of the initial master sensor is less than the preset first weight threshold and there are other sensors whose weight values are greater than the preset second weight threshold, the sensors with weight values greater than the preset second weight threshold are selected to form a candidate sensor set. When the weight value of the initial master sensor is less than the preset first weight threshold, it indicates that its performance can no longer meet the working requirements of the master sensor; while for the sensors with weight values greater than the preset second weight threshold, it shows that their performance is relatively good and they have the potential to become the master sensor. By comparing with the thresholds, the situation of the performance decline of the initial master sensor can be detected in time, avoiding the interruption or inaccuracy of monitoring data caused by the failure of the master sensor and ensuring the continuity of the monitoring work.

[0104] Meanwhile, after screening out the candidate sensor set, each sensor votes for the sensors in the candidate sensor set. Through the voting mechanism, the opinions of multiple sensors are integrated to select the sensor with better performance and more recognition in actual work as the updated master sensor; the voting basis includes the actual working performance such as the measurement accuracy stability, data transmission reliability, and anti-interference ability of the sensor. Each sensor votes for each sensor in the candidate sensor set within a specified time, compares the number of votes obtained by the candidate sensors, and determines the sensor with the most votes as the updated master sensor. When there is a tie in the number of votes, the comprehensive performance scores of the sensors are further compared, and the sensor with a higher score is selected as the updated master sensor; the voting mechanism fully considers the actual working experiences and evaluations of multiple sensors, can select the master sensor that better meets the requirements in actual operation, and improves the scientificity and rationality of the master sensor selection.

[0105] Specifically, after determining the updated master sensor, a role conversion instruction is sent to the initial master sensor to notify it to convert from the master sensor role to the slave sensor role. After receiving the role conversion instruction, the initial master sensor establishes a communication connection with the updated master sensor according to the preset communication protocol; after the initial master sensor is converted into a slave sensor, it continues to collect data in the monitoring area and transmits the data to the updated master sensor according to the new communication link. The updated master sensor receives and processes these data and then transmits them to the external monitoring system; by replacing the master sensor, the initial master sensor with degraded performance is converted into a slave sensor and continues to be used for data collection, quickly completing the role conversion and communication connection establishment of the master and slave sensors, reducing the impact on data collection and transmission work during the replacement of the master sensor in the sensor network, and ensuring the continuity of the monitoring work and the stability of the network.

[0106] Furthermore, constructing a communication link between the master sensor and the slave sensor according to the positional relationship and functional dependence between the master sensor and the slave sensor to obtain a sensor network includes:

[0107] S601. Construct an adjacency matrix according to the positional distances of each sensor;

[0108] S602. Connect the two sensors with the shortest interval distance according to the adjacency matrix;

[0109] S603. When the number of interval sensors connecting the slave sensor and the master sensor is greater than the preset number threshold, update the communication link, and select the path with the number of interval sensors less than the preset number threshold and the shortest overall distance of the communication link as the updated communication link;

[0110] S604. Establish communication between each slave sensor and the master sensor according to the updated communication link to obtain a sensor network.

[0111] In this embodiment, the position distances between sensors are calculated based on the position coordinates of each sensor, and the calculated position distances are used to construct an adjacency matrix. In the sensor network, each sensor is regarded as a node, and the position distance between sensors is the connection weight between nodes. By constructing the adjacency matrix, the distance relationship between sensors is intuitively presented. When constructing a communication link, the distance information between any two sensors can be quickly obtained, so as to efficiently plan the connection path. Specifically, according to the adjacency matrix, two sensors with the shortest interval distance are selected for connection. The shorter the distance, the smaller the attenuation of the signal during transmission, and the relatively less interference it receives, and the higher the stability and accuracy of data transmission. Connect the sensors step by step in this way to construct the communication network topology structure, and at the same time, the energy consumption of the sensors can be reduced to a certain extent because short-distance transmission consumes less energy. By preferentially connecting sensors with the shortest distance, signal attenuation and interference can be effectively reduced, the success rate and accuracy of data transmission can be improved, and it is ensured that the monitoring data can be reliably transmitted to the main sensor.

[0112] Specifically, the communication link is updated according to the number of interval sensors in the connection path between the slave sensor and the master sensor, avoiding the risk of increased data transmission delay and data loss caused by too many interval sensors in the communication link from the slave sensor to the master sensor. According to the actual communication requirements and sensor performance in the coal mine, a preset number threshold is set, which is set to 3 in this embodiment. Traverse all communication links from the slave sensor to the master sensor, and count the number of interval sensors in each link. For example, the link from slave sensor A to master sensor B passes through sensors C, D, and E, and the number of interval sensors is 3. When the number of interval sensors in a certain link is greater than the preset number threshold, a new communication path is searched from the adjacency matrix, and a set of paths with the number of interval sensors less than the preset number threshold is screened out. For each path in the screened path set, according to the distance between sensors recorded in the adjacency matrix, calculate the total distance of the path, compare the total distances of each path in the path set, and select the path with the shortest total distance as the updated communication link. When there are multiple paths with the same and shortest total distance, select a path with the best sensor transmission performance as the updated communication link.

[0113] According to the updated communication link, reconfigure the communication connections between sensors, disconnect the redundant connections in the original link, establish new connection relationships, establish stable and efficient communication connections between each slave sensor and the master sensor, connect all sensors to form a complete sensor network, and ensure that data can be transmitted according to the new link. By reducing the number of intermediate sensors in the communication link from the slave sensor to the master sensor, the data transmission delay can be significantly reduced, enabling the master sensor to obtain the data of the slave sensor faster and improving the response speed of the entire sensor network. The shorter communication link and fewer intermediate links reduce the risk of data transmission failure caused by sensor failures or signal interference, enhance the stability of the communication link, and ensure the continuity and reliability of the monitoring data transmission.

[0114] Further, based on the data collected by each sensor network, perform concentration monitoring through a preset concentration safety monitoring model, and give an early warning when abnormal results of the gas concentration occur to achieve gas concentration monitoring, including:

[0115] S701. According to the data collected by each sensor network, detect the sensor nodes with data transmission failures in the sensor network and perform data correction to obtain multi-sensor corrected data;

[0116] S702. Perform hierarchical fusion on the multi-sensor corrected data to obtain fused data;

[0117] S703. According to the fused data, perform concentration monitoring through a preset concentration safety monitoring model, and give an early warning when abnormal results of the gas concentration occur to achieve gas concentration monitoring.

[0118] In this embodiment, according to the data collected by each sensor network, the data is corrected, the nodes with data transmission failures in the sensor network are detected, and the data of the failed sensor nodes is corrected by combining the spatial correlation and time trend of the data collected by the sensor nodes to obtain more accurate and reliable multi-sensor corrected data, avoiding the influence of data transmission failures of sensors due to electromagnetic interference, equipment failures, signal blocking, etc. in the complex environment of the coal mine underground, effectively avoiding the interference of abnormal data on the monitoring results, enabling the multi-sensor corrected data to truly reflect the actual gas concentration situation, and improving the accuracy of the data.

[0119] Specifically, based on multiple sensor data collected by the sensor network, data fusion is performed. By performing hierarchical processing on the data, first, the data of the slave sensors is corrected according to the positional relationship between the master and slave sensors to eliminate the measurement deviation caused by distance; then, the data of the master and slave sensors is fused to obtain the comprehensive data within the region; finally, the data between regions is analyzed to correct abnormal data, enabling the fused data to more accurately and comprehensively reflect the gas concentration distribution in the monitored area and providing more accurate data for the concentration safety monitoring model; through hierarchical fusion, the multi-sensor corrected data is comprehensively processed, reducing data redundancy and inconsistency, improving the quality and usability of the data, providing more accurate data for the concentration safety monitoring model, and enhancing the reliability of the monitoring results.

[0120] Further, based on the fused data, a preset concentration safety monitoring model is used to monitor the gas concentration. The concentration safety monitoring model in this embodiment is specifically a random forest model, which is trained using a large amount of historical data to obtain a pre-trained concentration safety monitoring model; the fused data is input into the pre-trained concentration safety monitoring model, and the model outputs a gas concentration safety judgment result according to the safety threshold of the gas concentration and the abnormal judgment rule. When the gas concentration exceeds 80% of the safety threshold or the gas concentration rises by more than 50% within 10 minutes, it is determined that the gas concentration is abnormal; when the model detects an abnormal result, the warning mechanism is immediately triggered, and warning information is sent to relevant personnel such as coal mine safety management personnel and underground operators through methods such as text messages, audible and visual alarms, and pushing through the coal mine safety production management system, reminding them to take corresponding preventive and treatment measures; through the preset concentration safety monitoring model, abnormal gas concentration situations can be quickly and accurately identified, early warnings can be issued in a timely manner, precious time can be gained for coal mine safety production, and the occurrence of gas accidents can be effectively prevented.

[0121] Further, based on the data collected by each sensor network, the sensor nodes with data transmission failures in the sensor network are detected and data correction is performed to obtain multi-sensor corrected data, including:

[0122] S801. Based on the data collected by each sensor network, the sensor nodes with a packet loss rate greater than the preset loss threshold are regarded as failed sensor nodes;

[0123] S802. By calculating the distance weighted values of a preset number of adjacent non-failed sensor nodes of the failed sensor nodes, a spatial correction value is obtained;

[0124] S803. Based on the historical collected data of the failed sensor nodes, the data is predicted through a preset trend prediction model to obtain a predicted value;

[0125] S804. The spatial correction value and the predicted value are fused through a preset weight to obtain multi-sensor correction data.

[0126] In this embodiment, first, sensor nodes with a high data loss rate during the data transmission process in the sensor network are identified, and the data of the failed sensor nodes is corrected. According to the actual calculation accuracy requirements, a data loss threshold is set. In this embodiment, the preset loss threshold is set to 15%. The sensor network management system continuously monitors the data transmission status of each sensor node. Within a unit time (such as every minute), the total number of data packets sent by each sensor node and the number of data packets successfully transmitted to the target node (such as the main sensor) are recorded; the data packet loss rate = (total number of sent data packets - number of successfully transmitted data packets) ÷ total number of sent data packets × 100%, and the data packet loss rate of each sensor node is calculated; the calculated data packet loss rate of each sensor node is compared with the preset loss threshold. When the data packet loss rate of the sensor node is greater than 15%, the failed sensor node is obtained; the failed sensor node is discovered in time to prevent its unreliable data from being mixed into the concentration safety monitoring data, ensuring the overall quality of the data used for gas concentration monitoring and improving the credibility of the monitoring results.

[0127] Specifically, after identifying the failed sensor node, with the failed sensor node as the center, according to a preset range (such as a circular area with a radius of 5 meters centered on the failed sensor node), the non-failed sensor nodes within this range are screened out. At the same time, the number of adjacent non-failed sensor nodes participating in the calculation is set. For example, 3 adjacent non-failed sensor nodes closest to the failed sensor node are selected, and the adjacent non-failed sensor nodes are weighted and corrected according to the distance between the nodes. The correction formula is:

[0128] ;

[0129] In the formula, S is the spatial correction value, n is the number of adjacent non-failed sensor nodes, di is the distance between the i-th adjacent non-failed sensor node and the failed sensor node, dj is the distance between the j-th adjacent non-failed sensor node and the failed sensor node, and Ci is the gas concentration data value collected by the i-th adjacent non-failed sensor node. According to the data and spatial position relationship of the adjacent non-failed sensor nodes, the spatial information of the data is fully exploited, making the corrected data more in line with the actual gas concentration distribution and improving the accuracy of data correction.

[0130] Meanwhile, based on the historical acquisition data of the failed sensor nodes, analyze the trend information of the gas concentration at their locations over time, and make predictions through a preset trend prediction model. The trend prediction model is specifically an LSTM model. Use a large amount of historical data to train the LSTM model to obtain a pre-trained trend prediction model. Input the time information (such as timestamp) of the failure time period into the trained trend prediction model. The model predicts the gas concentration value within this time period according to the learned historical data patterns to obtain a predicted value. Use the historical data and the trend prediction model to predict and supplement the data of the failed sensor nodes from the time dimension, fully exploiting the time series characteristics of the data, improving the integrity of the data, and enabling the monitoring data to more comprehensively reflect the change of gas concentration over time.

[0131] Specifically, according to the actual calculation requirements and the analysis of the monitoring data, set the fusion weights of the spatial correction value and the predicted value. For example, if it is considered that the spatial correlation has a greater impact on data correction, set the weight of the spatial correction value to 0.7 and the weight of the predicted value to 0.3. Fuse the spatial correction value and the predicted value according to the set weight values to calculate the multi-sensor corrected data. By fusing the spatial correction value and the predicted value, the information in both the spatial and time dimensions of the data is fully utilized, making up for the deficiencies of a single correction method, making the corrected data closer to the actual gas concentration, and improving the quality and accuracy of the data.

[0132] Further, the step of performing hierarchical fusion on the multi-sensor corrected data to obtain the fusion data includes:

[0133] S901. According to the distances between the master sensor node and each slave sensor node, correct the data of each slave sensor in the multi-sensor corrected data to obtain the corrected data of the slave sensors;

[0134] S902. Fuse the master sensor data and the corrected data of the slave sensors to obtain the regional fusion data;

[0135] S903. Correct the data with different gas concentration and adjacent area gas concentration gradient trends in the regional fusion data to obtain the fusion data.

[0136] In this embodiment, first, according to the distances between the master sensor and the slave sensors, correct the data of each slave sensor. According to the calculated distances and combined with the calibration empirical formula, the calibration empirical formula is set according to the sensor calibration experience, calculate the calibration coefficient, and perform weighted calculation with the correction data to calculate the corrected data of each slave sensor after correction. By correcting the data of the slave sensors, the data deviation caused by the difference in sensor positions is reduced, making the data of the slave sensors closer to the true gas concentration value.

[0137] Specifically, according to factors such as the reliability and accuracy of the main sensor and the slave sensor, corresponding weights are assigned to each sensor. The data of each sensor is weighted and summed with the corresponding weight to obtain the regional fusion data. By fusing the data of multiple sensors, the data of the main sensor and the slave sensor are fully utilized, so that the fused data contains more information and can better reflect the overall situation of the gas concentration in the region. For each region, the gas concentration gradient between it and the adjacent region is calculated, and a gradient difference threshold is set. When the gradient difference of the gas concentration in the region exceeds the gradient difference threshold compared with the gradient of the adjacent region, it is considered that the fusion data of this region is abnormal. The fusion data of this region is corrected by the interpolation correction method, and the specific interpolation correction method is the spline interpolation algorithm. The corrected fusion data can more accurately reflect the actual gas concentration distribution, which helps to detect the abnormal changes of the gas concentration in time and ensure safe production.

[0138] Embodiment 2:

[0139] In this embodiment, as Figure 4 , a coal seam borehole gas concentration monitoring system based on multi-sensor fusion is provided for implementing the coal seam borehole gas concentration monitoring method based on multi-sensor fusion, including:

[0140] A coal seam division module analyzes the gas concentration risks of different depth layers according to the preset coal seam borehole information and divides them into multiple levels of coal seams;

[0141] A regional division module divides multiple monitoring regions for each level of coal seam by analyzing the gas concentration gradients of different regions;

[0142] A sensor network construction module sets multiple sensors in each monitoring region, divides the main sensor and the slave sensor by analyzing the communication relationship between the sensors and combining the voting results between the sensors, and constructs a sensor network;

[0143] A gas concentration monitoring module monitors the concentration according to the data collected by each sensor network through a preset concentration safety monitoring model and gives an alarm when an abnormal gas concentration result appears to achieve gas concentration monitoring.

[0144] In this embodiment, the coal seam division module analyzes the gas concentration risk of different depth layers according to the preset coal seam drilling information, comprehensively considers various characteristics of the coal seam and factors such as historical data, etc., and divides the coal seam into multiple levels according to different gas concentration risks, so as to take targeted monitoring measures for different levels of coal seams subsequently, providing a basis for accurately monitoring the gas concentration; for each level of coal seam, the area division module further divides multiple monitoring areas by analyzing the gas concentration gradient of different areas. There will be differences in the gas concentration changes in different areas of the same level of coal seam. Through area division, the distribution of gas concentration in the coal seam can be understood more carefully, making the monitoring work more targeted and accurate, and helping to detect abnormal gas concentration areas in time.

[0145] Specifically, the sensor network construction module sets multiple sensors in each monitoring area, constructs a sensor network by analyzing the communication relationship between sensors. First, determine the initial master sensor according to the performance data of the sensors, then calculate the weight value of each sensor through a dynamic weight calculation model, screen candidate sensors according to the weight value and determine the updated master sensor by voting, and finally establish a communication link between the master and slave sensors. The constructed sensor network can optimize the cooperation and communication between sensors, and improve the efficiency and reliability of data collection and transmission.

[0146] Specifically, the gas concentration monitoring module monitors the gas concentration according to the data collected by each sensor network, uses a preset concentration safety monitoring model to monitor the gas concentration. First, detect and correct the data of sensor nodes with data transmission failures, then perform hierarchical fusion on the corrected data, and finally perform concentration monitoring according to the fused data. When the gas concentration is abnormal, an alarm is issued in time, realizing real-time and accurate monitoring of the gas concentration in the coal seam drilling, being able to detect potential safety hazards in time, and providing strong support for ensuring the safe production of coal mines.

[0147] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion, characterized in that, Including: Analyze the gas concentration risks of different depth layers according to the preset coal seam borehole information, and divide them into multiple levels of coal seams; For each level of coal seam, divide multiple monitoring areas by analyzing the gas concentration gradients in different areas; Set multiple sensors in each monitoring area. By analyzing the communication relationships between the sensors and combining the voting results among the sensors, divide the master sensors and slave sensors, and construct a sensor network; According to the data collected by each sensor network, conduct concentration monitoring through a preset concentration safety monitoring model, and give an alarm when abnormal results of the gas concentration occur to achieve gas concentration monitoring.

2. The coal seam borehole gas concentration monitoring method based on multi-sensor fusion according to claim 1, wherein The step of analyzing the gas concentration risks of different depth layers according to the preset coal seam borehole information and dividing them into multiple levels of coal seams includes: Obtain the coal seam permeability, historical gas emission volume, and borehole trajectory data according to the preset borehole information for each layer; Perform vertical clustering on the coal seam according to the coal seam permeability, historical gas emission volume, and borehole trajectory data, and divide it into the first-risk-level coal seam, the second-risk-level coal seam, and the third-risk-level coal seam; When the gas concentration increase rate of the coal seam exceeds the preset concentration change threshold, upgrade the coal seam; When the gas concentration decrease rate of the coal seam exceeds the preset concentration change threshold, downgrade the coal seam.

3. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 1, wherein The step of dividing multiple monitoring areas by analyzing the gas concentration gradients in different areas for each level of coal seam includes: For each level of coal seam, set the initial grid size according to the preset borehole spacing; Divide the coal seam into multiple grid areas according to the initial grid size; Calculate the gas concentration gradient value between adjacent grids according to the gas concentration in each grid area; When the gas concentration gradient value is less than the preset gradient threshold, merge the corresponding adjacent grids to obtain multiple monitoring areas.

4. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 1, characterized in that, The step of setting multiple sensors in each monitoring area, dividing the master sensors and slave sensors by analyzing the communication relationships between the sensors and combining the voting results among the sensors, and constructing a sensor network includes: Set multiple sensors in each monitoring area, and divide the master sensors and slave sensors according to the communication relationships between different sensors and the voting results among the sensors; Construct a communication link between the master sensor and the slave sensor according to the positional relationship and functional dependence between the master sensor and the slave sensor to obtain a sensor network.

5. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 4, characterized in that, The step of setting multiple sensors in each monitoring area, and dividing the master sensors and slave sensors according to the communication relationships between different sensors and the voting results among the sensors includes: In each monitoring area, set the initial master sensor according to the performance data of each sensor; Within a preset time period, calculate the weight value of each sensor through a preset dynamic weight calculation model; When the weight value of the initial master sensor is less than the preset first weight threshold, use the sensors with weight values greater than the preset second weight threshold as the candidate sensor set, where the preset first weight threshold is less than the preset second weight threshold; According to the performance of the sensors in the candidate sensor set, each sensor votes for each sensor in the candidate sensor set, and use the sensor with the most votes as the updated master sensor. The initial master sensor acts as a slave sensor and communicates with the updated master sensor.

6. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 4, characterized in that, Constructing a communication link between the master sensor and the slave sensor according to the positional relationship and functional dependence between the master sensor and the slave sensor to obtain a sensor network, including: Constructing an adjacency matrix according to the position distances of each sensor; Connecting the two sensors with the shortest interval distance according to the adjacency matrix; When the number of interval sensors for communication connection between the slave sensor and the master sensor is greater than a preset number threshold, updating the communication link, and selecting a path with the number of interval sensors less than the preset number threshold and the shortest overall distance of the communication link as the updated communication link; Establishing communication between each slave sensor and the master sensor according to the updated communication link to obtain a sensor network.

7. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 1, characterized in that, Performing concentration monitoring on the data collected by each sensor network through a preset concentration safety monitoring model, and giving an alarm when the gas concentration shows an abnormal result to achieve gas concentration monitoring, including: Detecting sensor nodes with data transmission failures in the sensor network according to the data collected by each sensor network, and performing data correction to obtain multi-sensor corrected data; Performing hierarchical fusion on the multi-sensor corrected data to obtain fusion data; Performing concentration monitoring on the fusion data through a preset concentration safety monitoring model, and giving an alarm when the gas concentration shows an abnormal result to achieve gas concentration monitoring.

8. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 7, wherein, Detecting sensor nodes with data transmission failures in the sensor network according to the data collected by each sensor network, and performing data correction to obtain multi-sensor corrected data, including: Regarding sensor nodes with a packet loss rate greater than a preset loss threshold as failed sensor nodes according to the data collected by each sensor network; Obtaining a spatial correction value by calculating the distance weighted values of a preset number of adjacent non-failed sensor nodes of the failed sensor nodes; Predicting the data through a preset trend prediction model according to the historical collected data of the failed sensor nodes to obtain a predicted value; Fusing the spatial correction value and the predicted value through a preset weight to obtain multi-sensor corrected data.

9. The method for monitoring the gas concentration in coal seam boreholes based on multi-sensor fusion according to claim 7, characterized in that, Performing hierarchical fusion on the multi-sensor corrected data to obtain fusion data, including: Correcting the data of each slave sensor in the multi-sensor corrected data according to the distance between the master sensor node and each slave sensor node to obtain slave sensor corrected data; Fusing the master sensor data and the slave sensor corrected data to obtain regional fusion data; Correcting the data with different gas concentration and adjacent area gas concentration gradient trends in the regional fusion data to obtain fusion data.

10. A coal seam borehole gas concentration monitoring system based on multi-sensor fusion, characterized in that, Used to implement the multi-sensor fusion-based coal seam borehole gas concentration monitoring method according to any one of claims 1 to 9, including: A coal seam division module, analyzing the gas concentration risks of different depth layers according to preset coal seam borehole information, and dividing them into multiple levels of coal seams; A regional division module, for each level of coal seam, dividing multiple monitoring regions by analyzing the gas concentration gradients of different regions; Sensor network construction module, which sets multiple sensors in each monitoring area, divides the master sensor and slave sensors by analyzing the communication relationship between sensors and combining the voting results between sensors, and constructs a sensor network; Gas concentration monitoring module, which monitors the concentration through a preset concentration safety monitoring model based on the data collected by each sensor network, and issues a warning when an abnormal result of the gas concentration appears to achieve gas concentration monitoring.

Images