Coal mine fire accurate early warning method based on dynamic comprehensive prevention and control technology

Abstract

The invention relates to the technical field of coal mine fire prediction, and discloses a coal mine fire accurate early warning method based on a dynamic comprehensive prevention and control technology, and the method comprises the steps: receiving multi-source environment sensing data of a mine monitoring end, and then carrying out the preliminary screening through combining with a coal mine global risk database, and obtaining a first risk region set; when the set meets an early warning trigger condition, a roadway risk situation map is called and sent to a monitoring center end, and the roadway risk situation map comprises a risk layout plane corresponding to the roadway network structure; then determining a risk unit area based on the division information of the monitoring center end on the risk layout plane, and determining a risk factor type according to the labeling information of the risk unit area; then calculating a comprehensive risk assessment coefficient; and finally, generating a fire risk early warning sequence according to a descending order of the coefficients, and sending the fire risk early warning sequence to a monitoring center end, thereby realizing accurate and low-false-alarm fire early warning.

Description

Technical Field

[0001] This invention relates to the field of coal mine fire prediction technology, and in particular to a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology. Background Technology

[0002] Currently, in the field of coal mine safety production, fire remains one of the major disasters threatening miners' lives and the operation of mine facilities. Existing fire early warning systems mainly rely on threshold alarm mechanisms from single or a few environmental sensors, such as temperature sensors, carbon monoxide sensors, and gas concentration sensors, to achieve fire early warning. When the monitoring data at a certain point exceeds a preset limit, the fire early warning system will immediately trigger an alarm.

[0003] However, the underlying logic of this existing fire early warning system has obvious flaws. On the one hand, when equipment failure, local heat sources, or ventilation disturbances occur, non-fire factors can easily cause false alarms in the monitoring values, leading to too many misjudgments and thus weakening the staff's trust in the early warning system. On the other hand, the true precursors of fires in coal mines often manifest as multiple risk factors co-evolving in the coal mine space according to specific patterns. For example, different risk factors such as gas accumulation, abnormal temperature, and turbulent airflow will appear according to fixed spatial patterns. However, traditional early warning methods only focus on isolated points where the values ​​exceed the standard and cannot identify complex spatiotemporal correlation characteristics.

[0004] Especially in large, modern mines, the complex network of tunnels and numerous monitoring points result in a massive amount of alarm information. This makes it difficult for monitoring personnel to quickly identify the real dangers, which severely restricts the timeliness and accuracy of the emergency response that should be achieved. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, this invention provides a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, which can solve the problems of misjudgment caused by the reliance on single or a few sensor threshold alarm mechanisms in existing traditional fire early warning systems, the inability to identify complex spatiotemporal correlation features, and the inability to quickly identify real fire hazards in large mines.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, comprising: Receive multi-source environmental sensor data from the mine monitoring terminal, and conduct a preliminary screening of the coal mine-wide risk database based on the multi-source environmental sensor data to obtain the first risk area set; When the first set of risk areas meets the dynamic comprehensive early warning triggering conditions, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane corresponding to the roadway network structure of the target mining area. Based on the spatial division information of the risk deployment plane at the monitoring center, the risk unit area is determined, and the corresponding risk factor type is determined based on the risk labeling information of the risk unit area at the monitoring center. Based on the spatial location comparison information of the risk unit area, the risk factor type, and the first area in the first risk area set, a fire risk warning sequence is generated and sent to the monitoring center.

[0008] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, the step of generating a fire risk early warning sequence based on the spatial location comparison information of the risk unit area, the risk factor type, and the first area in the first risk area set, and sending the fire risk early warning sequence to the monitoring center includes: Based on the risk unit area, the risk factor type, and the location comparison information of the first area in the first risk area set along the roadway, a preliminary risk area set is obtained. Based on the location comparison information of the risk unit area, the risk factor type and the vertical lane direction of the second area in the initial screening risk area set, the horizontal risk correlation value of the second area is obtained. The comprehensive risk assessment coefficient is obtained by weighting the vertical risk matching value and the horizontal risk correlation value of the second region in the initial screening risk region set. Based on the comprehensive risk assessment coefficient, the second region is sorted in descending order to generate a fire risk warning sequence, which is then sent to the monitoring center.

[0009] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, the step of obtaining the lateral risk correlation value of the second region based on the position comparison information of the risk unit area, the risk factor type, and the vertical roadway orientation of the second region in the initial screening risk area set includes: In the risk deployment plane, an interactive risk vector is constructed pointing to another risk unit region, starting from any risk unit region. In the second region, a data risk vector pointing to another monitoring sub-region is constructed, starting from the corresponding monitoring sub-region. Retrieve the first trend sequence corresponding to the second region in the initial screening risk region set, select the first risk factor type in the first trend sequence as the trend reference factor, and use the remaining risk factor types as trend correlation factors; By selecting any one of the directional correlation factors and combining it with the directional reference factor, multiple spatial correlation combinations can be obtained; The monitoring sub-region in the second region corresponding to the orientation reference factor is taken as the data baseline sub-region, and the monitoring sub-region in the second region corresponding to the orientation association factor is taken as the data association sub-region; Starting from the geometric center of the data baseline sub-region and ending at the geometric center of the data association sub-region, a data risk vector corresponding to the spatial association combination is constructed. The risk unit area in the risk layout plane corresponding to the direction reference factor is used as the interactive reference unit, and the risk unit area in the risk layout plane corresponding to the direction association factor is used as the interactive association unit. Starting from the geometric center of the interactive reference unit and ending at the geometric center of the interactive association unit, an interactive risk vector corresponding to the spatial association combination is constructed. Calculate the angle between the data risk vector and the interaction risk vector; When the included angle is determined to be less than or equal to a preset angle threshold, a preset high-risk association value is retrieved and configured into the corresponding spatial association combination in the second region; When the included angle is determined to be greater than the preset angle threshold, a preset low-risk association value is retrieved and configured into the corresponding spatial association combination of the second region; Based on the preset high-risk association value and the preset low-risk association value of the spatial association combination corresponding to the second region, a preset risk association value corresponding to the spatial association combination is obtained; The horizontal risk correlation value of the second region is obtained by statistically analyzing the preset risk correlation values ​​of all spatial correlation combinations corresponding to the second region.

[0010] As a preferred embodiment of the precise early warning method for coal mine fires based on dynamic integrated prevention and control technology described in this invention, the step of obtaining a preliminary risk area set based on the risk unit area, the risk factor type, and the position comparison information of the first area in the first risk area set along the roadway direction includes: The risk factor types corresponding to each risk unit area in the risk layout plane are used as the baseline risk factor types. Based on the preset roadway direction, the baseline risk factor types in the risk layout plane are counted sequentially to generate an interactive risk factor sequence. The risk factor types that are the same as the benchmark risk factor types in each of the first risk regions in the first risk region set are obtained as the matching risk factor types for each of the first regions.

[0011] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, the step of obtaining a preliminary risk area set based on the comparison information of the risk unit area, the risk factor type, and the position of the first area in the first risk area set along the roadway direction further includes: Based on the preset roadway orientation, the types of matching risk factors in the first area are counted sequentially to obtain the first orientation sequence corresponding to the first area; The matching risk factor types in the interactive risk factor sequence are extracted sequentially to obtain the benchmark trend sequence corresponding to the first region. The first region, which is consistent with the first directional sequence and the baseline directional sequence, is selected as the second region, and the second region is statistically analyzed to obtain the initial risk region set.

[0012] As a preferred embodiment of the precise early warning method for coal mine fires based on dynamic integrated prevention and control technology described in this invention, the step of weighting the longitudinal risk matching value and the horizontal risk correlation value of the second region in the initial screening risk region set to obtain a comprehensive risk assessment coefficient includes: The number of risk factor types that match the first trend sequence corresponding to the second region is taken as the vertical matching number of the second region. The number of baseline risk factor types in the interactive risk factor sequence is obtained as the longitudinal baseline number, and the longitudinal risk matching degree is obtained based on the ratio of the longitudinal matching number to the longitudinal baseline number. The vertical risk coefficient is obtained by multiplying the vertical risk matching degree and the vertical risk weight, and the horizontal risk coefficient is obtained by multiplying the horizontal risk correlation value and the horizontal risk weight. The comprehensive risk assessment coefficient is obtained by summing the longitudinal risk coefficient and the horizontal risk coefficient.

[0013] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, wherein: when the included angle corresponding to the spatial association combination is less than or equal to a preset angle threshold, the corresponding spatial association combination is taken as a unidirectional spatial combination; The magnitude of the interaction risk vector corresponding to the same-direction spatial combination is obtained as the interaction risk distance, and the magnitude of the data risk vector is obtained as the data risk distance; Based on the absolute value of the difference between the interaction risk distance and the data risk distance, a spatial scale deviation is obtained. The preset high-risk correlation value is then attenuated and adjusted according to the spatial scale deviation to obtain the actual risk correlation value.

[0014] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, the step of attenuating and adjusting the preset high-risk correlation value according to the spatial scale deviation to obtain the actual risk correlation value includes: The scale attenuation coefficient is obtained based on the ratio of the spatial scale deviation to the preset reference scale, and the risk attenuation constant corresponding to the preset reference scale is retrieved. The risk attenuation value is obtained by multiplying the scale attenuation coefficient and the risk attenuation constant. The actual risk attenuation value is obtained by the difference between the preset high-risk correlation value and the risk attenuation value.

[0015] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, wherein: when the first risk area set meets the dynamic integrated early warning triggering conditions, a roadway risk situation map is retrieved and sent to the monitoring center, the roadway risk situation map including a risk layout plane corresponding to the roadway network structure of the target mining area, including: Obtain the first quantity of the first region in the first risk region set; When the first quantity is determined to be greater than or equal to the preset area threshold, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane of the roadway network structure corresponding to the first area.

[0016] As a preferred embodiment of the coal mine fire precision early warning method based on dynamic integrated prevention and control technology described in this invention, the step of determining risk unit areas based on the spatial division information of the risk deployment plane at the monitoring center, and determining the corresponding risk factor types based on the risk labeling information of the risk unit areas at the monitoring center, includes: Based on the spatial division information of the risk deployment plane by the monitoring center, the risk unit area is determined, and the risk labeling information of the risk unit area is received from the monitoring center. Based on the risk labeling information, the types of risk factors for each risk unit area are determined. The types of risk factors include at least one of abnormal gas concentration, excessive temperature gradient, or disordered ventilation.

[0017] Compared with existing technologies, the beneficial effects of this invention are that it proposes a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology. This method receives multi-source environmental sensor data from the mine monitoring terminal, then performs preliminary screening using a coal mine-wide risk database to obtain a first set of risk areas. When this set meets the early warning triggering conditions, a roadway risk situation map is retrieved and sent to the monitoring center. This roadway risk situation map includes risk deployment planes corresponding to the roadway network structure. Then, based on the risk deployment plane division information from the monitoring center, risk unit areas are determined, and risk factor types are determined according to their labeling information. Combining the spatial location comparison information of the risk unit areas, risk factor types, and the first area, a comprehensive risk assessment coefficient is calculated through longitudinal sequence matching along the roadway direction and lateral orientation consistency judgment based on the risk vector angle. Finally, a fire risk early warning sequence is generated in descending order of this coefficient and sent to the monitoring center, achieving accurate fire early warning with low false alarm rates. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The present invention provides a flowchart of a method for precise early warning of coal mine fires based on dynamic integrated prevention and control technology, which is an embodiment of the present invention.

[0020] Figure 2 The diagram illustrates the logic flow of a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, as provided in one embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram comparing the interactive risk vector and the data risk vector of a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, provided as an embodiment of the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figures 1-3This is the first embodiment of the present invention, which provides a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, including: This invention provides a method that can effectively solve the problems mentioned above. The following sections will elaborate on how to implement this precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, using multiple embodiments. Figure 1 A flowchart illustrating a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology is shown, including: S1 receives multi-source environmental sensor data from the mine monitoring terminal, performs preliminary screening of the coal mine's overall risk database based on the multi-source environmental sensor data, and obtains the first risk area set; In some embodiments, when receiving multi-source environmental sensor data from the mine monitoring terminal, the focus can be on collecting observations of several physical quantities, including gas concentration, ambient temperature, and wind speed. Furthermore, the spatial coordinate information corresponding to each observation can be recorded simultaneously. For example, in a sensor network deployed in the main transport roadway of a mining area, data can be collected at least once per minute. If a sensor measures a gas concentration of 1.2% at coordinates (X1, Y1, Z1), it is determined to exceed the safety threshold of 1.0%. Simultaneously, if another sensor measures a local temperature of 38°C at coordinates (X2, Y2, Z2), it is considered higher than the normal ambient temperature threshold of 35°C. These multi-source data items together constitute the initial raw multi-source environmental sensor data set.

[0024] In some embodiments, after obtaining the original data set, the original data set can be compared item by item with the pre-built coal mine risk database. For example, the pre-built coal mine risk database stores a large number of characteristic parameters of historical fire event areas, as well as expert-set thresholds for various risk factors and dynamic risk boundaries output by neural network models. If the gas concentration data collected by a sensor exceeds the gas concentration anomaly threshold of the corresponding location in the database, then the spatial location of the sensor is marked as a potential risk point.

[0025] In some embodiments, once these potential risk points are marked, all potential risk points that meet the threshold conditions can be aggregated to form a first risk area set. For example, in a monitoring cycle, if the system identifies a total of 7 spatial locations, including the intersection of the return airway, the upper corner of the coal mining face, and the poorly ventilated branch roadway, where one or more risk factors exceed the standard, then these 7 locations constitute the first risk area set.

[0026] The first risk area set here refers to the set of spatial locations where all environmental parameters are abnormal and exceed the preset risk threshold during the current monitoring period, which is used to characterize the preliminary risk distribution of the mine.

[0027] It should be noted that step S1 can quickly identify the spatial location of abnormal parameters in the current mine based on multi-source environmental sensing data, thereby forming a set of first risk areas, which can then provide an objective basis for whether to activate the advanced early warning mechanism.

[0028] S2, when the first risk area set meets the dynamic comprehensive early warning triggering conditions, retrieve the roadway risk situation map and send it to the monitoring center. The roadway risk situation map includes the risk layout plane corresponding to the roadway network structure of the target mining area. In some embodiments, when determining whether the first risk area set meets the dynamic comprehensive early warning triggering conditions, the first number of the first area contained in the first risk area set can be counted. After obtaining this first number, it can be compared with the preset area threshold. For example, when the system detects that a certain mining area has 8 abnormal spatial locations in a single data acquisition cycle, that is, there are 8 first areas, and the preset area threshold is set to 5, then it can be determined that the first risk area set meets the dynamic comprehensive early warning triggering conditions.

[0029] In some embodiments, the roadway risk situation map can be retrieved and sent to the monitoring center immediately after the triggering conditions are met. This roadway risk situation map is generated based on the actual roadway network structure of the target mining area and can completely present the orientation relationships of the main haulage roadway, return air roadway, and connecting roadway, the layout of intersection nodes, and the installation locations of local ventilation facilities, such as... Figure 2 As shown, but Figure 2 It's a simplified version. Figure 2 The tunnel topology is also expressed in a digital planar form, but the actual operation screenshots are more complex and involve iterative display.

[0030] In some embodiments, because the roadway risk situation map is relatively complex, the core components of the roadway risk situation map can be manipulated. These core components are referred to as the risk layout plane, and the location distribution of theoretical high-risk units will be marked on this risk layout plane in advance. These high-risk units correspond to key sections where spontaneous combustion, gas accumulation, or airflow turbulence has occurred in the past, such as roadway sections adjacent to goaf areas, areas around sealed walls, and branch roadways with abrupt changes in cross-section.

[0031] The roadway risk situation map here refers to a digital planar map that integrates the geometric structure of the target mining area roadways with the information of preset risk units, which is used to provide a structured benchmark for subsequent spatial division and risk factor comparison.

[0032] In this embodiment of the invention, when the first set of risk areas meets the dynamic comprehensive early warning triggering conditions, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane corresponding to the roadway network structure of the target mining area, including: Obtain the first quantity of the first region in the first risk region set; When the first quantity is determined to be greater than or equal to the preset area threshold, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes the risk layout plane of the roadway network structure corresponding to the first area.

[0033] It should be noted that step S2 can automatically trigger the structured analysis process when the number of risk areas reaches a preset threshold, retrieve the roadway risk situation map containing the roadway topology and theoretical risk distribution, and provide a reference plane for subsequent spatial comparison.

[0034] S3, determine the risk unit area based on the spatial division information of the risk deployment plane at the monitoring center, and determine the corresponding risk factor type based on the risk labeling information of the risk unit area at the monitoring center; In this embodiment of the invention, risk unit areas are determined based on the spatial division information of the risk deployment plane at the monitoring center, and the corresponding risk factor types are determined based on the risk labeling information of the risk unit areas at the monitoring center, including: Based on the spatial division information of the risk deployment plane at the monitoring center, the risk unit area is determined, and the risk labeling information of the risk unit area is received from the monitoring center. In some embodiments, after the roadway risk situation map is transmitted, the risk unit area can be determined based on the spatial division information of the risk layout plane at the monitoring center. For example, in the roadway network structure of a mining area, the monitoring center divides the main transport roadway into three continuous segments according to the roadway function, support status and historical accident records. The area near the intersection of the return airway is divided into an independent node area, and the branch roadway with poor local ventilation is divided into two closed units. Each divided area does not overlap and covers all high-risk potential locations. These division results constitute the risk unit area set. This step is done in advance and needs to be divided before use.

[0035] In some embodiments, after completing the above spatial division operation, the risk labeling information of each risk unit area can be obtained at the monitoring center. For example, the first section of the main transport roadway is labeled as "abnormal gas concentration", the intersection node of the return air roadway is simultaneously labeled as "temperature gradient exceeding the standard" and "ventilation status disorder", and one of the closed units of the branch roadway is labeled as a composite risk type of "abnormal gas concentration" and "temperature gradient exceeding the standard". Then all the labeling information can be manually input by professional technicians based on historical fire patterns, geological conditions and real-time monitoring trends or generated with the assistance of the system.

[0036] In some embodiments, after obtaining the risk labeling information, the types of risk factors corresponding to each risk unit area can be clearly identified based on this labeling information. For example, abnormal gas concentration means that the gas volume fraction in the area is continuously higher than the safety threshold of 1.0%; excessive temperature gradient means that the temperature difference between adjacent monitoring points exceeds the set standard of 5℃ / 10 meters; and disordered ventilation means that the wind speed fluctuation exceeds the normal value ±30% and the wind direction frequently reverses. The above three types of risk factors constitute the basic set of risk factors, which will be used for subsequent spatial comparison analysis with the first risk area.

[0037] The risk unit area here refers to a structured area with clear geometric boundaries and risk attributes delineated by the monitoring center on the roadway risk layout plane. It is used to carry the types of risk factors and serve as a benchmark unit for spatial comparison.

[0038] It should be noted that the risk factors for each risk unit area are determined based on the risk labeling information. The risk factors include at least one of the following: abnormal gas concentration, excessive temperature gradient, or disordered ventilation.

[0039] It should be noted that step S3 can be performed manually or semi-automatically by the monitoring center to divide and mark the risk deployment plane, clarify the boundaries of each risk unit area and the types of risk factors corresponding to them, establish a theoretical high-risk model, and support subsequent vertical and horizontal consistency judgments.

[0040] S4. Based on the spatial location comparison information of the risk unit area, risk factor type, and the first area in the first risk area set, generate a fire risk warning sequence and send the fire risk warning sequence to the monitoring center.

[0041] In this embodiment of the invention, a fire risk warning sequence is generated based on the spatial location comparison information of the risk unit area, the type of risk factor, and the first area in the first risk area set. The fire risk warning sequence is then sent to the monitoring center, including: S401. Based on the risk unit area, risk factor type, and the position comparison information of the first area in the first risk area set along the roadway, a preliminary risk area set is obtained. Furthermore, based on the risk unit area, risk factor type, and the positional information of the first area in the first risk area set along the roadway direction, a preliminary risk area set is obtained, including: The risk factor types corresponding to each risk unit area in the risk layout plane are used as the baseline risk factor types. Based on the preset roadway direction, the baseline risk factor types in the risk layout plane are counted sequentially to generate an interactive risk factor sequence. The risk factor types that are the same as the benchmark risk factor types in each of the first risk regions are obtained as the matching risk factor types for each first region. Based on the preset roadway orientation, the types of matching risk factors in the first region are counted sequentially to obtain the first orientation sequence corresponding to the first region; The matching risk factor types in the interactive risk factor sequence are extracted sequentially to obtain the benchmark trend sequence corresponding to the first region. The first region, whose first directional sequence is consistent with the baseline directional sequence, is selected as the second region. The second region is statistically analyzed to obtain the initial set of risk regions.

[0042] In some embodiments, after obtaining the risk unit area and its corresponding risk factor type, the risk factor types of each risk unit area in the risk layout plane can be arranged sequentially according to the preset roadway direction to generate an interactive risk factor sequence. For example, in the direction from the main transport roadway to the return air roadway in a certain mining area, there are three risk unit areas in sequence, and their risk factor types are abnormal gas concentration, excessive temperature gradient, and disordered ventilation. Then the interactive risk factor sequence is [abnormal gas concentration, excessive temperature gradient, disordered ventilation].

[0043] In some embodiments, the types of risk factors corresponding to the monitoring data within each first area in the first risk area set can be obtained, and items that are consistent with the types of risk factors that have appeared in the interactive risk factor sequence can be selected. Then, they are sorted according to the same roadway direction, and finally the first direction sequence corresponding to the first area can be obtained. For example, a certain first area contains three monitoring sub-areas, which respectively detect abnormal gas concentration, excessive temperature gradient and disordered ventilation. The spatial position of these three monitoring sub-areas along the roadway direction is consistent with the order of the interactive risk factor sequence. Then the first direction sequence of this first area is [abnormal gas concentration, excessive temperature gradient, disordered ventilation].

[0044] In some embodiments, a subsequence that perfectly matches the first orientation sequence can be extracted from the interactive risk factor sequence as the baseline orientation sequence corresponding to the first region. For example, if the interactive risk factor sequence is [abnormal gas concentration, excessive temperature gradient, disordered ventilation], and the first orientation sequence of a certain first region is [abnormal gas concentration, excessive temperature gradient], then the baseline orientation sequence [abnormal gas concentration, excessive temperature gradient] can be extracted from the first two items of the interactive risk factor sequence. This ensures that during the comparison process, not only is the existence of factor types considered, but also the consistency of their arrangement logic in the roadway orientation is strictly verified.

[0045] In some embodiments, a first region whose first orientation sequence is completely consistent with the baseline orientation sequence can be marked as a second region and included in the initial screening risk region set. For example, if the first orientation sequence of a certain first region is [abnormal gas concentration, excessive temperature gradient, disordered ventilation], and is completely consistent with the interactive risk factor sequence, then it is determined that the region not only has multiple factor anomalies, but its spatial distribution pattern is also highly consistent with the theoretical high-risk structure, and has the potential for fire evolution, so it is selected into the initial screening risk region set.

[0046] The initial risk area set refers to the set of second-order regions selected after longitudinal sequence consistency testing along the roadway direction. This set is used to exclude isolated or random outliers and retain candidate regions with structured risk characteristics.

[0047] S402. Based on the location comparison information of the risk unit area, risk factor type and the vertical roadway direction of the second area in the initial screening risk area set, the horizontal risk correlation value of the second area is obtained. Based on the location comparison information of the risk unit area, risk factor type, and the vertical roadway orientation of the second area in the initial screening risk area set, the horizontal risk correlation value of the second area is obtained, including: In the risk deployment plane, an interactive risk vector is constructed pointing to another risk unit area, starting from any risk unit area. In the second region, a data risk vector pointing to another monitoring sub-region is constructed, starting from the corresponding monitoring sub-region. Retrieve the first trend sequence corresponding to the second region in the initial screening risk region set, select the first risk factor type in the first trend sequence as the trend reference factor, and use the remaining risk factor types as trend correlation factors; By selecting any directional correlation factor and combining it with the directional reference factor, multiple spatial correlation combinations can be obtained; The monitoring sub-regions in the second region corresponding to the directional reference factor are used as the data baseline sub-regions, and the monitoring sub-regions in the second region corresponding to the directional correlation factor are used as the data correlation sub-regions. Starting from the geometric center of the data baseline sub-region and ending at the geometric center of the data association sub-region, a data risk vector corresponding to the spatial association combination is constructed. The risk unit area in the risk layout plane corresponding to the direction reference factor is used as the interactive benchmark unit, and the risk unit area in the risk layout plane corresponding to the direction correlation factor is used as the interactive correlation unit. Starting from the geometric center of the interactive baseline unit and ending at the geometric center of the interactive associated unit, an interactive risk vector corresponding to the spatial association combination is constructed. Calculate the angle between the data risk vector and the interaction risk vector; When the included angle is determined to be less than or equal to a preset angle threshold, a preset high-risk association value is retrieved and configured into the corresponding spatial association combination in the second region. When the included angle is determined to be greater than a preset angle threshold, a preset low-risk association value is retrieved and configured into the corresponding spatial association combination in the second region. Based on the preset high-risk correlation value and preset low-risk correlation value of the spatial correlation combination corresponding to the second region, the preset risk correlation value corresponding to the spatial correlation combination is obtained; The horizontal risk correlation value of the second region is obtained by statistically analyzing the preset risk correlation values ​​of all spatial correlation combinations corresponding to the second region.

[0048] In some embodiments, after obtaining the set of initial risk areas, lateral spatial relationship modeling can be performed for each of the second areas. For example, taking a second area that includes three monitoring sub-areas: abnormal gas concentration, excessive temperature gradient, and disordered ventilation, the first risk factor type that appears in the first directional sequence is selected as the directional reference factor, i.e., abnormal gas concentration, and the other two factors are used as directional correlation factors.

[0049] In some embodiments, the orientation reference factor can be combined with each orientation-related factor separately to form multiple spatial association combinations. For example, abnormal gas concentration and excessive temperature gradient constitute the first spatial association combination, and abnormal gas concentration and disordered ventilation constitute the second spatial association combination. Each combination is used to independently assess the spatial orientation consistency of the two risk factors in the direction perpendicular to the roadway orientation.

[0050] In some embodiments, risk unit areas corresponding to the direction reference factors can be located in the risk layout plane as interactive reference units, and risk unit areas corresponding to the direction association factors can be located as interactive association units. Then, an interactive risk vector is constructed with the geometric center of the interactive reference unit as the starting point and the geometric center of the interactive association unit as the ending point. For example, if the interactive reference unit is located at coordinates (X1, Y1) in the middle section of the main transport roadway and the interactive association unit is located at coordinates (X2, Y2) in the branch roadway to its north, then the interactive risk vector is a two-dimensional vector pointing from (X1, Y1) to (X2, Y2).

[0051] In some embodiments, a monitoring sub-region corresponding to the directional reference factor can be located within the second region as a data baseline sub-region, and a monitoring sub-region corresponding to the directional association factor can be located as a data association sub-region. Then, a data risk vector is constructed with the geometric center of the data baseline sub-region as the starting point and the geometric center of the data association sub-region as the ending point. For example, if the measured position of the data baseline sub-region is (x1, y1) and the measured position of the data association sub-region is (x2, y2), then the data risk vector is a vector pointing from (x1, y1) to (x2, y2).

[0052] In some embodiments, the directionality of spatial association combinations can be determined by calculating the angle between the data risk vector and the interaction risk vector. If this angle is less than or equal to a preset angle threshold (e.g., 30 degrees), the spatial association combination is considered to have good directional consistency, and a preset high-risk association value (e.g., 1.0) is configured. If the angle is greater than the preset angle threshold, the direction is considered to be significantly deviated, and a preset low-risk association value (e.g., 0.2) is configured. For example, when the interaction risk vector points due north and the data risk vector points northeast with an angle of 25 degrees, a high-risk association value is assigned; if the data risk vector points southeast with an angle of 60 degrees, a low-risk association value is assigned.

[0053] In some embodiments, it is necessary to accumulate the preset risk correlation values ​​corresponding to all spatial correlation combinations in the second region, and then obtain the horizontal risk correlation value of the second region. For example, if a second region contains two spatial correlation combinations with correlation values ​​of 1.0 and 0.2 respectively, the horizontal risk correlation value is 1.2. This value reflects the matching strength between the spatial layout of multiple risk factors in the vertical lane direction within the region and the theoretical high-risk pattern.

[0054] The horizontal risk correlation value here refers to a value quantified based on the consistency between the relative spatial orientation and theoretical structure of risk factors, and is used to measure the risk synergy characteristics of the second region in the horizontal dimension.

[0055] S403, the vertical risk matching value and horizontal risk correlation value of the second region in the initial screening risk region set are weighted and calculated to obtain the comprehensive risk assessment coefficient; The comprehensive risk assessment coefficient is obtained by weighting the vertical risk matching value and the horizontal risk correlation value of the second region in the initial risk region set, including: The number of risk factor types that match the first trend sequence corresponding to the second region is taken as the vertical matching number of the second region. The number of baseline risk factor types in the interactive risk factor sequence is obtained as the longitudinal baseline number. The longitudinal risk matching degree is obtained based on the ratio of the longitudinal matching number to the longitudinal baseline number. The vertical risk coefficient is obtained by multiplying the vertical risk matching degree and the vertical risk weight, and the horizontal risk coefficient is obtained by multiplying the horizontal risk correlation value and the horizontal risk weight. The comprehensive risk assessment coefficient is obtained by summing the vertical risk coefficient and the horizontal risk coefficient.

[0056] The comprehensive risk assessment coefficient is obtained by summing the vertical and horizontal risk coefficients, including: The comprehensive risk assessment coefficient is obtained using the following formula. ; in, The comprehensive risk assessment coefficient, which represents the first... The overall risk level of the second region For the number of vertical matches, For the vertical baseline quantity, For vertical risk weights, This represents the upper limit of the number of spatial association combinations in all second regions. For horizontal risk weights, Indicates the second region The final horizontal risk correlation value.

[0057] In an embodiment of the present invention, when the included angle corresponding to a spatial association combination is determined to be less than or equal to a preset angle threshold, the corresponding spatial association combination is taken as a spatial combination in the same direction. Obtain the magnitude of the interaction risk vector corresponding to the same-direction spatial combination as the interaction risk distance, and the magnitude of the data risk vector as the data risk distance; Based on the absolute value of the difference between the interaction risk distance and the data risk distance, the spatial scale deviation is obtained. The preset high-risk correlation value is attenuated and adjusted according to the spatial scale deviation to obtain the actual risk correlation value.

[0058] The preset high-risk correlation value is attenuated and adjusted based on the spatial scale deviation to obtain the actual risk correlation value, including: The scale attenuation coefficient is obtained based on the ratio of the spatial scale deviation to the preset benchmark scale, and the risk attenuation constant corresponding to the preset benchmark scale is retrieved. The risk attenuation value is obtained by multiplying the scale attenuation coefficient and the risk attenuation constant. The actual risk attenuation value is obtained by the difference between the preset high-risk correlation value and the risk attenuation value. The actual risk correlation value can be obtained using the following formula. ; in, This is the actual risk correlation value, which represents the correlation value after distance deviation correction. To preset high-risk correlation values, To mitigate the risk of interaction, To mitigate data risk distance, To pre-set a benchmark scale, The attenuation of the correlated value caused by the unit deviation.

[0059] S404: Based on the comprehensive risk assessment coefficient, the second area is sorted in descending order to generate a fire risk warning sequence, which is then sent to the monitoring center.

[0060] In some embodiments, after calculating the comprehensive risk assessment coefficient for all second regions, the second regions can be sorted in descending order based on this comprehensive risk assessment coefficient. For example, if there are three second regions with comprehensive risk assessment coefficients of 0.87, 0.65 and 0.92 respectively, the sorting result is the third second region, the first second region, and the second second region. This sorting ensures that the region with the highest risk level is placed at the beginning of the sequence.

[0061] In some embodiments, the sorted second region can be sequentially encapsulated into a data structure containing region identifiers and corresponding risk scores, and then used as a fire risk warning sequence. For example, the first element in the sequence is "return airway intersection node (comprehensive risk assessment coefficient 0.92)", the second element is "main transport roadway middle section (comprehensive risk assessment coefficient 0.87)", and the third element is "local branch roadway closure section (comprehensive risk assessment coefficient 0.65)". Each element can clearly identify the spatial location and quantify the risk value.

[0062] In some embodiments, the complete fire risk warning sequence can also be transmitted to the monitoring center in real time through the mine communication network. For example, after the system identifies five high-risk secondary areas and completes the assessment during a certain monitoring cycle, it is immediately pushed to the dispatch screen and duty terminal of the monitoring center in the form of structured data packets, so that emergency command personnel can quickly grasp the risk priority.

[0063] In some embodiments, after receiving the fire risk warning sequence, the monitoring center can also directly use it to guide on-site inspection route planning and resource scheduling. For example, the scheduling instruction can prioritize assigning the fire prevention and extinguishing team to the intersection of the return airway with the highest comprehensive risk assessment coefficient to check the gas accumulation and temperature anomalies. At the same time, ventilation engineers can be arranged to check the airflow status in the area simultaneously, so as to achieve graded response according to risk level.

[0064] The fire risk warning sequence refers to the second zone list arranged from high to low according to the comprehensive risk assessment coefficient. Each entry contains a clear spatial location identifier and a quantitative risk score, which is used to support the monitoring center to carry out accurate and efficient fire prevention and control decisions and emergency response.

[0065] It should be noted that step S4 can integrate the spatial matching relationship between the theoretical risk structure and the actual monitored anomalies, and generate an orderly fire risk early warning sequence through comprehensive risk assessment, so that the monitoring center can respond accurately according to risk priority, significantly reduce the false alarm rate and improve the efficiency of emergency response.

[0066] Example 2: Based on the above examples, a specific implementation of a precise early warning method for coal mine fires based on dynamic integrated prevention and control technology can be designed as follows: At the beginning, multi-source environmental sensor data can be received from the mine monitoring terminal, and then the original dataset can be constructed based on this data. : ; In the formula, Indicates the first Each sensor collects data items. This represents the total number of sensors, each of which... All contain three-dimensional spatial coordinates And the corresponding physical quantity observation values, the types of physical quantities mentioned here generally include coal mine fire-related indicators such as gas concentration, ambient temperature, and wind speed.

[0067] In practical applications, the sensor data stream of a real-time multi-source environmental sensor with a sampling frequency of not less than once per minute can be selected as the input source to ensure the diversity of the data set.

[0068] Furthermore, after obtaining the set, you can then... With the pre-built coal mine full-domain risk database To conduct a comparison item by item, here... This refers to a database that stores historical fire events, characteristics of high-risk areas, and threshold standards for various risk factors in advance. It can be obtained based on expert experience and some neural network models, and there are no restrictions here.

[0069] Furthermore, when making comparisons, if a certain data item... When a physical quantity exceeds the preset threshold of its corresponding risk factor in the coal mine's overall risk database, its spatial location needs to be marked as a potential risk point. Then, all such potential risk points in spatial locations are combined to form the first risk region set, denoted as: ; In the formula, Indicates the first The spatial location of the initially identified risk areas. This represents the total number of risk points that meet the threshold condition.

[0070] It should be noted that this set of first risk areas obtained after the assessment can reflect the geographical distribution of abnormal environmental parameters in the current mine, which is also the basic input for further refined assessment.

[0071] Furthermore, once the first set of risk areas is obtained, it is necessary to determine whether to activate the advanced early warning mechanism. Here, M can be set as the first quantity of the first region in the first set of risk areas, and then a region threshold can be preset. If the value is a positive integer, this preset area threshold is used to control the response of the early warning system. ; If this is the case, it can be determined that the current risk situation has reached a level that cannot be analyzed manually, and therefore it is necessary to trigger the dynamic comprehensive early warning process.

[0072] It should be noted that this design prevents human operators from making misjudgments due to frequent system responses when encountering frequent alarms.

[0073] Furthermore, once this Then it is necessary to retrieve the tunnel risk situation map. Then, the risk situation map of this tunnel... The data is transmitted to the monitoring center and then used for subsequent spatial partitioning and risk labeling. It is a simplified view of the digital plan generated based on the actual roadway network structure of the target mining area. Figure 2 Similar to the dashed lines in the diagram, this is a planar structural map containing topological information such as the direction of the tunnels, intersections, and the layout of ventilation facilities. This is a tunnel risk situation map. The core component of the diagram is the risk deployment plane. This risk deployment plane The preset positions of theoretically high-risk units can be marked on the screen.

[0074] It should be noted that this step is to transform the traditional point-like anomalies into a structural diagram, and also to lay a preliminary foundation for the subsequent introduction of lane spatial semantics.

[0075] Furthermore, when the monitoring center receives the roadway risk situation map... After that, the monitoring center will view the risk situation map of this alley. Risk layout plan inside The space is divided into several non-overlapping risk unit regions.

[0076] It should be noted that the area divided in this partitioning operation can be manually or semi-automatically partitioned by professionals based on factors such as roadway function, support status, and historical accident records. The partitioning results can be recorded as a set. Specifically, it is expressed as follows: ; In the formula, Indicates the first The geometric boundary range of each risk unit region Indicates the total number of units, each In physical space, each corresponds to a continuous alleyway or a key node area.

[0077] Furthermore, after the segmentation process, the monitoring center will automatically analyze each risk unit area. The risks are labeled with information, and the types of risk factors are determined based on this labeling information. Here, it can be assumed that the types of risk factors are taken from a finite set. gas, temp, vent The specific contents in this set represent abnormal gas concentration, excessive temperature gradient, and disordered ventilation. Relevant technical personnel can add or remove the types of risk factors in these sets according to actual needs.

[0078] In practical applications, a single While associating multiple factor categories, to simplify sequence alignment, it's also possible to process each category according to its dominant factor in subsequent steps. The corresponding dominant factor is denoted as .

[0079] Furthermore, once the information labeling and risk factor classification are completed, the vertical consistency comparison stage can begin. This can be done by first defining a predefined directional vector for the tunnel's orientation. This is a unit vector used to unify the spatial sorting benchmark. This vector is determined by the direction of the main haulage roadway or the main intake airway in the mine, so as to ensure that the processing direction is consistent before and after the entire system is processed.

[0080] Furthermore, one can follow Directional traversal risk deployment plane Extract the types of risk factors from all risk unit regions to form an interactive risk factor sequence: ; In the formula, Indicates the sequence length. Indicates by Sorted risk unit index.

[0081] Furthermore, once the sorted risk unit index is obtained, it is necessary to analyze the first risk region set. Each region in Performing local factor extraction can be performed in the first risk region. In the internal monitoring sub-area, filter out those related to Observations of the same type that have appeared in the data, and following the same trend... Sort the sequences to obtain the first-order sequence: ; In the formula, Indicates the first Length of observation sequence for each region Indicates the first in this region The types of risk factors that were matched.

[0082] Furthermore, after obtaining the first directional sequence, it is possible to perform a sequence for each... Construct the corresponding baseline trajectory sequence This baseline trend sequence can be established based on expert experience and can be understood as a theoretical high-risk structure. During its establishment, it is also necessary to ensure that it originates from an interactive risk factor sequence. Extracting from and in sequence Terms with completely identical factor types, and strictly maintaining their position in The original order is used for construction.

[0083] It should be noted that this construction process can ensure that, during the comparison process, not only the existence of the types of factors of interest can be monitored, but also whether the arrangement logic in the direction of the lanes is the same.

[0084] Furthermore, this invention performs a sequence consistency check if the following conditions are met: ; Then the area can be identified. Not only do risk factors exist, but its spatial distribution pattern is also highly consistent with theoretical high-risk structures, making it more likely to evolve into a fire.

[0085] Furthermore, it can be Marked as the second region and included in the initial screening risk region set: ; In the formula, This indicates the number of regions that passed the longitudinal consistency test. Indicates the first The second region is a set of input objects designed for subsequent cross-correlation analysis, which can effectively filter out isolated or random outliers or abnormal regions.

[0086] Furthermore, in obtaining the initial screening risk area set After that, we can proceed to the stage of horizontal risk correlation analysis, where... Indicates the first The second region that passed the longitudinal consistency test. This indicates the total number of such regions.

[0087] It should be noted that the vertical risk correlation analysis is to determine whether the arrangement order of risk factors in the first region is consistent with the preset directional sequence of the baseline risk factors in the roadway risk situation map, while the horizontal risk correlation analysis is to determine whether the relative spatial orientation between different risk factors in the second region is consistent with the relative spatial orientation between the corresponding risk unit areas in the roadway risk situation map.

[0088] Going further, it is necessary to analyze each second region. Perform the internal factor role classification operation. The first risk factor type appearing in the first trend sequence within this region can be selected as the trend reference factor, denoted as... The remaining factor types constitute the set of trend-related factors. .

[0089] In practical applications, it is necessary to ensure Risk deployment plane There is a unique corresponding interactive reference unit.

[0090] Furthermore, it is also necessary to provide for each pair Construct a spatial association combination Here, q represents the index of the qth correlation factor. This spatial correlation combination can be used to measure the directional consistency of two risk factors in spatial distribution, and can also be regarded as the basic unit of horizontal correlation calculation.

[0091] Furthermore, in the risk deployment plane Mid-positioning and The corresponding risk unit area is recorded as the interactive baseline unit. Simultaneously positioning with The corresponding risk unit area is denoted as the interactive association unit. .

[0092] Furthermore, preset The function represents the coordinates of the geometric center. This function can be used to obtain the location of the cell center point and then construct the interaction risk vector, as shown below: ; It should be noted that the interactive risk vector can reflect the spatial orientation relationship between two factors in the theoretical risk structure, and this spatial orientation relationship can provide a basis for subsequent determination of relative positions.

[0093] Furthermore, it is also necessary to go further in the second region. The monitoring sub-area corresponding to the internal positioning can be compared with... The corresponding monitoring sub-region is denoted as the data baseline sub-region. , will with The corresponding monitoring sub-region is denoted as the data association sub-region. Then use the same method. Obtain its geometric center and simultaneously construct a data risk vector: ; This vector represents the spatial relative position of two risk factors in actual observation.

[0094] Going further, it is necessary to calculate the angle between the two vectors. Alright, as Figure 3 As shown: ; In the formula, This represents the vector dot product operation. This angle represents the horizontal length of the vector and is used to quantify the degree of consistency between the theoretical structure and actual observations in direction.

[0095] Furthermore, a preset angle threshold can be set. Then, the analysis results of lateral risk are finally obtained through this preset angle threshold.

[0096] It should be noted that it is possible to set It is an acute angle constant, with a typical value range of... to In this regard, relevant technical personnel can design according to actual needs, and this invention does not impose any limitations.

[0097] It should be noted that if the angle between the two vectors is calculated... satisfy: ; Then the spatial association combination can be determined. If the direction is consistent, a preset high-risk correlation value can be assigned. ; Otherwise, the spatial association combination can be determined. If the direction consistency is insufficient, a preset low-risk correlation value can be assigned. .

[0098] In practical applications, one can select , Such an extreme choice can effectively reflect the differences.

[0099] Furthermore, after obtaining all the association values, it is necessary to sum the association values ​​of all spatial association combinations to obtain the second region. The final horizontal risk correlation value : ; In the formula, express The total number of spatial association combinations in the middle Indicates the first The correlation value corresponding to each combination reflects the strength of the spatial distribution of multiple factors within the region and the theoretical high-risk pattern.

[0100] Furthermore, longitudinal risk matching values ​​can be calculated to extract the second region. The number of risk factor types matched in the first-order sequence is denoted as . Simultaneously obtain the interactive risk factor sequence The total number of benchmark risk factors, denoted as Then the longitudinal risk matching degree can be calculated: ; In the formula, This represents the proportion of the area covered by the theoretical risk model along the direction of the alley, with a value range of [0, 1].

[0101] Furthermore, vertical risk weights can be set. , are positive real numbers and satisfy Then the longitudinal risk coefficient can be calculated: ; It should be noted that this coefficient is a reference value reflecting the contribution of vertical structural consistency to overall risk.

[0102] Furthermore, it is also necessary to introduce horizontal risk weights. , It is a positive real number and satisfies This is to ensure that the normalization performance of memory comprehensive evaluation is achieved.

[0103] Furthermore, an upper limit can be set for the number of spatially related combinations in all second regions. This value is used to normalize the lateral components and is denoted as... Then, the final expression for calculating the comprehensive risk assessment coefficient can be as follows: ; In the formula, Indicates the first The comprehensive risk level of the second region is a combination of vertical pattern coverage and horizontal spatial correlation strength. This coefficient is the most important basis for the final warning ranking.

[0104] It should be noted that the preceding comprehensive risk assessment coefficient operation is to determine the directional relationship between specific coal mine roadway risk units. However, this operation can only determine the directional relationship, so the location needs to be determined later in order to finally achieve the final risk determination.

[0105] Furthermore, a scale decay mechanism can be enabled, which can also be understood as a distance-based judgment mechanism. This requires correcting the correlation value. Under these conditions, it is also necessary to calculate the interaction risk distance. Distance from data risks The spatial scale deviation can be defined as: ; In practical applications, it is also necessary to set a preset benchmark scale. and risk decay constant , This indicates the maximum acceptable distance deviation reference value. This indicates the magnitude of the decay of the correlated value caused by a unit deviation.

[0106] Furthermore, it is necessary to calculate the specific scale attenuation coefficient: ; Then, the actual risk correlation value can be obtained based on this attenuation coefficient: ; In the formula, This represents the correlation value after distance deviation correction. The final horizontal risk correlation value is then updated as follows: ; Then this Substitution The calculation formula ensures that even if the directions are consistent, if the spatial scale deviates significantly from the theoretical structure, the final risk correlation will be reasonably adjusted.

[0107] Furthermore, once the comprehensive risk assessment of all second-region areas is completed, the early warning sequence generation phase can begin, which can be based on the comprehensive risk assessment coefficient. Based on the sorting criteria, for all Sort in descending order.

[0108] Furthermore, a final fire risk warning sequence can be generated: ; In the formula, Each element in the sequence contains a region identifier and its corresponding risk score.

[0109] It should be noted that this sequence clearly expresses the priority order of each high-risk area.

[0110] Furthermore, this could ultimately lead to a fire risk early warning sequence. The data is sent to the monitoring center, whereby relevant dispatchers can then verify and handle it according to the risk level.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention. All such modifications and equivalent substitutions should be covered within the scope of the claims of the present invention.

[0112] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0113] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology, characterized in that, include: Receive multi-source environmental sensor data from the mine monitoring terminal, and conduct a preliminary screening of the coal mine-wide risk database based on the multi-source environmental sensor data to obtain the first risk area set; When the first set of risk areas meets the dynamic comprehensive early warning triggering conditions, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane corresponding to the roadway network structure of the target mining area. Based on the spatial division information of the risk deployment plane at the monitoring center, the risk unit area is determined, and the corresponding risk factor type is determined based on the risk labeling information of the risk unit area at the monitoring center. Based on the spatial location comparison information of the risk unit area, the risk factor type, and the first area in the first risk area set, a fire risk warning sequence is generated and sent to the monitoring center.

2. The method for precise early warning of coal mine fires based on dynamic integrated prevention and control technology as described in claim 1, characterized in that, The step of generating a fire risk warning sequence based on the spatial location comparison information of the risk unit area, the risk factor type, and the first area in the first risk area set, and sending the fire risk warning sequence to the monitoring center includes: Based on the risk unit area, the risk factor type, and the position comparison information of the first area in the first risk area set along the roadway, a preliminary risk area set is obtained. Based on the location comparison information of the risk unit area, the risk factor type and the vertical lane direction of the second area in the initial screening risk area set, the horizontal risk correlation value of the second area is obtained. The comprehensive risk assessment coefficient is obtained by weighting the vertical risk matching value and the horizontal risk correlation value of the second region in the initial screening risk region set. Based on the comprehensive risk assessment coefficient, the second region is sorted in descending order to generate a fire risk warning sequence, which is then sent to the monitoring center.

3. The method for precise early warning of coal mine fires based on dynamic integrated prevention and control technology as described in claim 2, characterized in that, The method of obtaining the lateral risk correlation value of the second region based on the location comparison information of the risk unit area, the risk factor type, and the vertical lane orientation of the second region in the initial screening risk area set includes: In the risk deployment plane, an interactive risk vector is constructed pointing to another risk unit region, starting from any risk unit region. In the second region, a data risk vector pointing to another monitoring sub-region is constructed, starting from the corresponding monitoring sub-region. Retrieve the first trend sequence corresponding to the second region in the initial screening risk region set, select the first risk factor type in the first trend sequence as the trend reference factor, and use the remaining risk factor types as trend correlation factors; By selecting any one of the directional correlation factors and combining it with the directional reference factor, multiple spatial correlation combinations can be obtained; The monitoring sub-region in the second region corresponding to the orientation reference factor is taken as the data baseline sub-region, and the monitoring sub-region in the second region corresponding to the orientation association factor is taken as the data association sub-region; Starting from the geometric center of the data baseline sub-region and ending at the geometric center of the data association sub-region, a data risk vector corresponding to the spatial association combination is constructed. The risk unit area in the risk layout plane corresponding to the direction reference factor is used as the interaction reference unit, and the risk unit area in the risk layout plane corresponding to the direction association factor is used as the interaction association unit. Starting from the geometric center of the interactive reference unit and ending at the geometric center of the interactive association unit, an interactive risk vector corresponding to the spatial association combination is constructed. Calculate the angle between the data risk vector and the interaction risk vector; When the included angle is determined to be less than or equal to a preset angle threshold, a preset high-risk association value is retrieved and configured into the corresponding spatial association combination in the second region; When the included angle is determined to be greater than the preset angle threshold, a preset low-risk association value is retrieved and configured into the corresponding spatial association combination of the second region; Based on the preset high-risk association value and the preset low-risk association value of the spatial association combination corresponding to the second region, a preset risk association value corresponding to the spatial association combination is obtained; The horizontal risk correlation value of the second region is obtained by statistically analyzing the preset risk correlation values ​​of all spatial correlation combinations corresponding to the second region.

4. The method for precise early warning of coal mine fires based on dynamic integrated prevention and control technology as described in claim 3, characterized in that, The process of obtaining a preliminary risk area set based on the risk unit area, the risk factor type, and the positional comparison information of the first area in the first risk area set along the roadway direction includes: The risk factor types corresponding to each risk unit area in the risk layout plane are used as the baseline risk factor types. Based on the preset roadway direction, the baseline risk factor types in the risk layout plane are counted sequentially to generate an interactive risk factor sequence. The risk factor types that are the same as the benchmark risk factor types in each of the first risk regions in the first risk region set are obtained as the matching risk factor types for each of the first regions.

5. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology as described in claim 4, characterized in that, The step of obtaining the initial risk area set based on the risk unit area, the risk factor type, and the position comparison information of the first area in the first risk area set along the roadway direction further includes: Based on the preset roadway orientation, the types of matching risk factors in the first area are counted sequentially to obtain the first orientation sequence corresponding to the first area; The matching risk factor types in the interactive risk factor sequence are extracted sequentially to obtain the benchmark trend sequence corresponding to the first region. The first region, which is consistent with the first directional sequence and the baseline directional sequence, is selected as the second region, and the second region is statistically analyzed to obtain the initial risk region set.

6. The method for precise early warning of coal mine fires based on dynamic integrated prevention and control technology as described in claim 5, characterized in that, The comprehensive risk assessment coefficient is obtained by weighting the vertical risk matching value and the horizontal risk correlation value of the second region in the initial screening risk region set, including: The number of risk factor types that match the first trend sequence corresponding to the second region is taken as the vertical matching number of the second region. The number of baseline risk factor types in the interactive risk factor sequence is obtained as the longitudinal baseline number, and the longitudinal risk matching degree is obtained based on the ratio of the longitudinal matching number to the longitudinal baseline number. The vertical risk coefficient is obtained by multiplying the vertical risk matching degree and the vertical risk weight, and the horizontal risk coefficient is obtained by multiplying the horizontal risk correlation value and the horizontal risk weight. The comprehensive risk assessment coefficient is obtained by summing the longitudinal risk coefficient and the horizontal risk coefficient.

7. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology as described in claim 6, characterized in that, When the included angle corresponding to the spatial association combination is determined to be less than or equal to a preset angle threshold, the corresponding spatial association combination is regarded as a spatial combination in the same direction; Obtain the magnitude of the interaction risk vector corresponding to the same-direction spatial combination as the interaction risk distance, and the magnitude of the data risk vector as the data risk distance; Based on the absolute value of the difference between the interaction risk distance and the data risk distance, a spatial scale deviation is obtained. The preset high-risk correlation value is then attenuated and adjusted according to the spatial scale deviation to obtain the actual risk correlation value.

8. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology as described in claim 7, characterized in that, The step of attenuating and adjusting the preset high-risk correlation value based on the spatial scale deviation to obtain the actual risk correlation value includes: The scale attenuation coefficient is obtained based on the ratio of the spatial scale deviation to the preset reference scale, and the risk attenuation constant corresponding to the preset reference scale is retrieved. The risk attenuation value is obtained by multiplying the scale attenuation coefficient and the risk attenuation constant. The actual risk attenuation value is obtained by the difference between the preset high-risk correlation value and the risk attenuation value.

9. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology as described in claim 8, characterized in that, When the first set of risk areas meets the dynamic comprehensive early warning triggering conditions, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane corresponding to the roadway network structure of the target mining area, including: Obtain the first quantity of the first region in the first risk region set; When the first quantity is determined to be greater than or equal to the preset area threshold, the roadway risk situation map is retrieved and sent to the monitoring center. The roadway risk situation map includes a risk layout plane of the roadway network structure corresponding to the first area.

10. A precise early warning method for coal mine fires based on dynamic integrated prevention and control technology as described in claim 9, characterized in that, The process of determining risk unit areas based on the spatial division information of the risk deployment plane at the monitoring center, and determining corresponding risk factor types based on the risk labeling information of the risk unit areas at the monitoring center, includes: Based on the spatial division information of the risk deployment plane by the monitoring center, the risk unit area is determined, and the risk labeling information of the risk unit area is received from the monitoring center. Based on the risk labeling information, the types of risk factors for each risk unit area are determined. The types of risk factors include at least one of abnormal gas concentration, excessive temperature gradient, or disordered ventilation.

Images