A method and system for judging a spontaneous combustion danger area of a goaf of a coal mining face

Abstract

The present application provides a kind of judging method and system of spontaneous combustion dangerous area of goaf in coal mining face, it is related to coal mine safety monitoring technical field, the present application generates initial geographic map by using geographic information system software, and creates regular three-dimensional grid on this basis, effectively realizes the accurate positioning and data consistency management of goaf in coal mining face, by setting specific wavelength range, using infrared spectrum to capture the spectral data of each sampling point, by calculating the relative radiation intensity and radiation deviation degree of each sampling point, the abnormal sampling point can be accurately identified, by obtaining the neighbor sampling point data of abnormal sampling point, calculating the uniform distribution index, the risk level mark of abnormal sampling point is realized, the distribution density, relative radiation intensity and geographic location information of abnormal sampling point are comprehensively considered, spontaneous combustion risk index is generated, and comprehensive evaluation and timely early warning of goaf spontaneous combustion risk are realized.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety monitoring technology, specifically to a method and system for determining the risk zone of spontaneous combustion in the goaf of a coal mining face. Background Technology

[0002] The goaf in a coal mining face refers to the empty area left after all coal resources have been extracted during coal mining. This area may contain voids formed by ground movement or subsidence. Due to the lack of support, these areas are prone to accumulating methane and coal seam gases, which can easily lead to spontaneous combustion or collapse. Therefore, in modern coal mining, the goaf in a coal mining face faces significant safety hazards due to the spontaneous combustion potential of the coal seam. The accumulation of methane and coal seam gases in the goaf, especially under poor ventilation conditions, can easily trigger spontaneous combustion. Such spontaneous combustion not only wastes coal resources but also produces toxic and harmful gases, threatening the lives of miners and the normal production of the coal mine. Therefore, how to effectively monitor and prevent spontaneous combustion in goafs has become an important issue in coal mine safety management.

[0003] Current technologies primarily rely on miners' experience and simple monitoring equipment, such as temperature sensors and gas detectors. While these devices can detect early signs of spontaneous combustion to some extent, they suffer from low accuracy, slow response times, and limited coverage. Some modern coal mines have attempted to introduce infrared thermal imaging and gas analyzers for monitoring, but these methods often require complex equipment and are costly. Furthermore, in the complex mining environment, data acquisition and processing are challenging, making real-time monitoring and accurate assessment difficult.

[0004] Compared to the method proposed in this application, existing technologies have significant shortcomings. First, relying on experience and simple monitoring equipment makes it difficult to achieve comprehensive coverage of the entire goaf, easily overlooking potential spontaneous combustion points. Second, while technologies such as infrared thermal imaging have high monitoring accuracy in some cases, their data processing and analysis lack systematicity, making it difficult to comprehensively consider multiple factors for a complete assessment. Furthermore, most existing data processing methods are relatively simple, failing to effectively integrate multi-dimensional information such as spatial location, radiation intensity, and temporal variations, resulting in low accuracy and reliability of spontaneous combustion risk assessment. Therefore, there is an urgent need for a method that integrates geographic information systems, near-infrared spectral analysis, and risk assessment models to achieve accurate identification and real-time early warning of spontaneous combustion hazard areas in coal mining face goafs.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for determining the spontaneous combustion hazard zone in the goaf of a coal mining face, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for determining the spontaneous combustion hazard zone in the goaf of a coal mining face, comprising the following steps:

[0009] Step 1: Use geographic information system software to import geographic information data of the goaf area of ​​the coal mining face to be judged to generate an initial geographic map. Create a regular grid in the initial geographic map, take each grid intersection as a sampling point, obtain the geographic location information of the sampling points, and number them. Collect the dielectric constant of each sampling point to determine the signal enhancement sampling points.

[0010] Step 2: Capture the spectral data of each sampling point, set the wavelength analysis range, record the spectral intensity of each sampling point within the wavelength analysis range, map it with the geographical location information of each sampling point, and normalize the spectral intensity data of all sampling points.

[0011] Step 3: For the spectral intensity data of the normalized sampling points, calculate the relative radiance of each sampling point, calculate the mean and standard deviation of the relative radiance of all sampling points, generate the radiance shift of each sampling point, and identify the abnormal sampling points in the signal enhancement sampling points based on the radiance shift of the sampling points.

[0012] Step 4: Obtain the relative radiation intensity of the neighboring sampling points of the abnormal sampling points in each signal enhancement sampling point, calculate the uniform distribution index of each abnormal sampling point, and mark the risk level of the abnormal sampling points according to the uniform distribution index;

[0013] Step 5: Based on the distribution density of all marked abnormal sampling points in the initial geographic map, as well as the relative radiation intensity data and geographic location information of each marked abnormal sampling point, construct a risk assessment model, generate a spontaneous combustion risk index for the goaf of the coal mining face to be judged, and judge the spontaneous combustion risk of the goaf based on the spontaneous combustion risk index.

[0014] Furthermore, the logic used to determine the signal enhancement sampling points is as follows:

[0015] The dielectric constant data of each sampling point is collected using ground-penetrating radar or electromagnetic detectors. The collected dielectric constant data is then imported into geographic information system software and correlated with the geographical location information of the sampling points. A threshold for the dielectric constant is set, and points with a value less than this threshold are used as signal enhancement sampling points.

[0016] Furthermore, the geographic information data of the goaf in the coal mining face includes the shape, extent, and elevation data of the goaf, as well as the geographic coordinate data of the goaf. The specific steps for obtaining the geographic location information of the sampling points are as follows:

[0017] Import the shape, extent, elevation, and geographic coordinate data of the mined-out area into the geographic information system software, ensuring that all imported datasets use a unified coordinate system; if inconsistent, perform coordinate system conversion.

[0018] The elevation data is cropped using the shape data of the goaf to ensure that only the goaf area is retained. The size and range of the grid are set, and a regular grid is generated according to the analysis requirements to cover the goaf. The grid intersections are extracted as sampling points and their geographical location information is recorded. The grid is a three-dimensional grid.

[0019] Furthermore, the wavelength analysis range set when recording the spectral intensity of each sampling point is 900nm to 1000nm. The specific logic for normalizing the spectral intensity data of all sampling points is as follows:

[0020] Obtain the spectral intensity of each sampling point within the wavelength range of 900nm to 1000nm, and calculate the maximum and minimum spectral intensities within this range for all sampling points. Specifically:

[0021] I max =max[Iq1, Iq2, ..., Iq i ... I qn ]

[0022] I min =min[Iq1, Iq2, ..., Iq] i ... I qn ]

[0023] Among them, I max and I min Iq represents the maximum and minimum spectral intensities within the wavelength range of 900 nm to 1000 nm at the sampling points, respectively. i This represents the spectral intensity data of sampling point i before normalization in the wavelength range of 900nm to 1000nm, where i represents the sampling point number and n represents the total number of sampling points.

[0024] The spectral intensities of all sampling points in the wavelength range of 900nm to 1000nm are scaled to the range of [0, 1], specifically as follows:

[0025]

[0026] I iThis represents the normalized spectral intensity data of sampling point i within the wavelength range of 900nm to 1000nm.

[0027] Furthermore, the logic used to calculate the relative radiation intensity at each sampling point is as follows:

[0028]

[0029] Among them, R i Let I be the relative radiation intensity of sampling point i. b The baseline radiation intensity represents the radiation intensity within 900 nm to 1000 nm under the same temperature and pressure without spontaneous combustion.

[0030] To generate the radiation offset at each sampling point, the mean and standard deviation of the relative radiation intensity of all sampling points are first calculated. The z-score method is then used to generate the radiation offset at each sampling point, based on the following formula:

[0031]

[0032] Where, μ R and σ R Let z represent the mean and standard deviation of the relative radiation intensity at all sampling points, respectively. i This represents the degree of radiation offset at sampling point i, when |z i |>z y Furthermore, when this point is a signal enhancement sampling point, the sampling point numbered i is identified as an abnormal sampling point, z y This indicates the offset threshold.

[0033] Furthermore, the neighboring sampling points of anomalies in the signal enhancement sampling points are the sampling points directly adjacent to the anomaly sampling point, including directly adjacent sampling points in the six directions of up, down, left, right, front, and back in the grid. The formula used to calculate the uniform distribution index of each anomaly point is as follows:

[0034]

[0035] Where, μz j and σz j Let represent the mean and variance of the relative radiance of the dataset consisting of the outlier sample point j and its neighboring sample points, respectively. j R represents the uniformity distribution index of the outlier sampling point j. j Rs j Rx j 、Rz j Ry j 、Rq j and Rh jThese represent the relative radiation intensities of the anomalous sampling point numbered j and its six neighboring sampling points (upper, lower, left, right, front, and rear). If an anomalous sampling point does not have at least one neighboring sampling point among the six neighboring sampling points, the relative radiation intensities of that sampling point are set to 0.

[0036] When marking outliers with risk levels based on the uniform distribution index, if Jf j >Jf y If the abnormal sampling point is identified, it will be marked as a high-risk sampling point and labeled; otherwise, it will be marked as a low-risk sampling point and not labeled. y denoted by the uniform distribution threshold, and j represents the number of the abnormal sampling point in the signal enhancement sampling points.

[0037] Furthermore, based on the distribution density of all marked outlier sampling points in the initial geographic map, the number of all sampling points and the number of marked outlier sampling points are counted, and the distribution density of outlier sampling points in the initial geographic map is calculated, specifically as follows:

[0038]

[0039] Where, ρ y represents the distribution density of anomalous sampling points in the initial geographic map, where n represents the total number of sampling points and M represents the number of marked anomalous sampling points.

[0040] Furthermore, the specific logic for constructing a risk assessment model and generating a spontaneous combustion risk index for the goaf of the coal mining face to be assessed is as follows:

[0041] Based on the geographical location information of each marked anomalous sampling point, its distance from the center of the coal mining face is calculated. A risk assessment model is constructed by combining the distribution density of anomalous sampling points on the initial geographic map and the relative radiation intensity data of the marked anomalous sampling points. The formula used is as follows:

[0042]

[0043] Wherein, RI represents the spontaneous combustion risk index, d m and Jf m Let represent the distance and relative radiation intensity between the marked anomalous sampling point numbered m and the center of the coal mining face, respectively. ∈ represents the denominator constant, and ∈>0. m represents the number of the marked anomalous sampling point, and e represents the base of the natural logarithm.

[0044] Furthermore, when assessing the spontaneous combustion risk of a goaf based on the spontaneous combustion risk index, if RI ≥ RI y If the risk of spontaneous combustion in the goaf is deemed high, an early warning is issued. <RI y If the risk of spontaneous combustion in the goaf is deemed low, no warning will be issued.y This indicates the threshold for spontaneous combustion risk.

[0045] The present invention also provides a system for determining the spontaneous combustion hazard zone in the goaf of a coal mining face. The system is used to execute the above-mentioned method for determining the spontaneous combustion hazard zone in the goaf of a coal mining face, comprising:

[0046] The geographic map creation module is used to import geographic information data of the goaf area of ​​the coal mining face to be judged using geographic information system software to generate an initial geographic map. A regular grid is created within the initial geographic map, and each grid intersection is used as a sampling point. The geographic location information of the sampling points is obtained and numbered. The dielectric constant of each sampling point is collected to determine the signal enhancement sampling points.

[0047] The preprocessing module is used to capture spectral data from each sampling point, set the wavelength analysis range, record the spectral intensity of each sampling point within the wavelength analysis range, map it to the geographical location information of each sampling point, and normalize the spectral intensity data of all sampling points.

[0048] The anomaly detection module is used to calculate the relative radiance of each sampling point for the spectral intensity data of the normalized sampling points, calculate the mean and standard deviation of the relative radiance of all sampling points, generate the radiance shift of each sampling point, and identify the abnormal sampling points in the signal enhancement sampling points based on the radiance shift of the sampling points.

[0049] The risk assessment module is used to obtain the relative radiation intensity of neighboring sampling points of abnormal sampling points in each signal enhancement sampling point, calculate the uniform distribution index of each abnormal sampling point, and mark the risk level of abnormal sampling points according to the uniform distribution index.

[0050] The comprehensive judgment module is used to construct a risk assessment model based on the distribution density of all marked abnormal sampling points in the initial geographic map, as well as the relative radiation intensity data and geographical location information of each marked abnormal sampling point, and to generate a spontaneous combustion risk index for the goaf of the coal mining face to be judged. The spontaneous combustion risk of the goaf is judged based on the spontaneous combustion risk index.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] This invention utilizes geographic information system software to generate an initial geographic map and then creates a regular three-dimensional grid on this basis. This effectively achieves precise positioning of the goaf in the coal mining face and data consistency management. By setting a specific wavelength range, infrared spectroscopy is used to capture the spectral data of each sampling point and normalize it to ensure the comparability and consistency of the data, thereby enhancing the accuracy of the analysis. In particular, by calculating the relative radiation intensity and radiation offset of each sampling point, abnormal sampling points in the signal enhancement sampling points can be accurately identified.

[0053] This invention acquires neighboring sampling point data from anomaly sampling points within signal enhancement sampling points, calculates a uniform distribution index, and thus labels the risk level of anomaly sampling points. This method not only considers the anomalies of individual sampling points but also integrates the data consistency of neighboring areas, significantly improving the reliability and accuracy of risk assessment. Finally, by constructing a risk assessment model that comprehensively considers the distribution density, relative radiation intensity, and geographical location information of anomaly sampling points, a spontaneous combustion risk index is generated. This enables a comprehensive assessment and timely early warning of spontaneous combustion risk in goaf areas. By achieving accurate judgment and real-time early warning of spontaneous combustion risk in goaf areas, it effectively reduces safety hazards in coal mine production. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0055] Figure 2 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0057] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0058] Example:

[0059] Please see Figure 1 The present invention provides a technical solution:

[0060] A method for determining the spontaneous combustion hazard zone in the goaf of a coal mining face, comprising the following steps:

[0061] Step 1: Use geographic information system software to import geographic information data of the goaf area of ​​the coal mining face to be judged to generate an initial geographic map. Create a regular grid in the initial geographic map, take each grid intersection as a sampling point, obtain the geographic location information of the sampling points, and number them. Collect the dielectric constant of each sampling point to determine the signal enhancement sampling points.

[0062] In this embodiment, the geographic information data of the goaf in the coal mining face includes the shape, extent, and elevation data of the goaf, as well as the geographic coordinate data of the goaf. The specific steps for obtaining the geographic location information of the sampling points are as follows:

[0063] Import the shape, extent, elevation, and geographic coordinate data of the mined-out area into the geographic information system software, ensuring that all imported datasets use a unified coordinate system; if inconsistent, perform coordinate system conversion.

[0064] The elevation data is cropped using the shape data of the goaf to ensure that only the goaf area is retained. The size and range of the grid are set, and a regular grid is generated according to the analysis requirements to cover the goaf. The grid intersections are extracted as sampling points and their geographical location information is recorded. The grid is a three-dimensional grid.

[0065] Elevation data provides vertical information, making the grid not just planar but three-dimensional. By cropping the elevation data and retaining only the mined-out area, and layering it vertically, a three-dimensional space is formed. A three-dimensional coordinate system is used to record the position of each sampling point, ensuring the uniqueness and accuracy of each point in space. The three-dimensional coordinates are longitude, latitude, and altitude. With length, width, and height set, the generated grid has fixed dimensions in all three directions, forming a cube with a side length of 50 meters. The three-dimensional grid is composed of multiple such cube units arranged according to a certain pattern. These cubes are arranged alternately to form a complete grid structure. The intersections of the grid are the vertices of the cube units. These intersections can be used as sampling points, and each intersection has its corresponding three-dimensional coordinates. Therefore, in this embodiment, the above steps collectively ensure that a three-dimensional grid is generated, rather than being limited to a two-dimensional grid on a plane.

[0066] Collect data on the shape, extent, elevation, and geographic coordinates of the goaf. The geographic information system software used is ArcGIS or QGIS. Create a new project in the geographic information system software to process and analyze the goaf data. Import the shape and extent data of the goaf, which exists in vector data form. Import the elevation data, which exists in raster data form.

[0067] Ensure all datasets use the same coordinate system. If they are inconsistent, perform coordinate system transformation. If the elevation data range is larger than the goaf area, the elevation data can be cropped using goaf shape data, keeping only the relevant parts. Integrate various types of data into one or more layers, ensuring that the data in each layer can be used in conjunction with each other.

[0068] Define the grid size and resolution. Select an appropriate grid size according to the analysis needs. In this embodiment, the grid size is 20 meters. Create a regular grid on the initial geographic map to cover the entire goaf area. The grid is a square cell. Take the intersection of each grid cell as a sampling point. Generate the geographic location of a sampling point for each grid intersection and record its longitude, latitude and elevation. Give each sampling point a unique number for easy identification and processing later.

[0069] In this embodiment, QGIS is used as a geographic information system software for detailed explanation. The specific operation of the above steps is as follows: Import data: Open QGIS, select "Layer" > "Add Layer" > "Add Vector Layer" to import the shape and range data of the goaf. Import elevation data by selecting "Layer" > "Add Layer" > "Add Raster Layer". Check data consistency and clipping. In "Layer Properties", check the coordinate system. If inconsistent, use the "Reproject Layer" tool to convert it. Use "Raster" > "Extraction" > "Clip Raster by MaskLayer" to clip the elevation data. Select "Processing" > "Toolbox", search for "Create Grid", set the grid size and range to generate a regular grid, obtain the geographical location information of the sampling points, use "Vector" > "Research Tools" > "Extract Nodes" to extract grid intersections and generate sampling points. Use the "Field Calculator" tool to generate a unique number for each sampling point. Overlay all layers in the map window, adjust the style and display effect, and import and process the geographic information data of the mining subsidence area in the GIS software to generate an initial geographic map containing sampling points, providing a foundation for subsequent analysis.

[0070] The purpose of identifying signal enhancement sampling points is to quickly identify areas with potential spontaneous combustion risks across a wide sampling area, thereby improving analysis efficiency. This step is achieved by analyzing changes in the dielectric constant, as an abnormally low dielectric constant may indicate changes in the composition of the coal seam, such as signs of oxidation and spontaneous combustion.

[0071] Identifying signal enhancement sampling points is a crucial step, enabling the rapid identification of potentially spontaneous combustion risk areas over a large area. This facilitates in-depth analysis in subsequent steps. The dielectric constant is closely related to the material composition and state of the underground medium. An abnormally low dielectric constant may indicate changes in the material properties of the area, such as increased air content or decreased moisture content. These changes are often associated with coal seam oxidation and spontaneous combustion processes. By setting a dielectric constant threshold, potentially anomalous areas can be quickly screened. These areas may experience enhanced electromagnetic signals due to changes in medium properties. After identifying signal enhancement sampling points, more detailed spectral analysis and radiation intensity calculations can be performed at these points, providing a reliable data foundation for subsequent risk assessment. Preliminary screening reduces the analysis scope, lowering the probability of large-scale misjudgments and improving the accuracy of subsequent analysis and judgment.

[0072] In this embodiment, identifying signal enhancement sampling points is a preliminary screening process. This step effectively narrows down the area of ​​focus. Utilizing changes in the dielectric constant to identify potential spontaneous combustion risk areas provides a more targeted basis for subsequent spectral data acquisition and relative radiation intensity analysis.

[0073] By setting a threshold for the dielectric constant, potentially risky areas can be preliminarily identified. These areas exhibit enhanced electromagnetic signal characteristics, making subsequent spectral analysis and radiation intensity calculations more targeted, thus providing a reliable data foundation for more in-depth risk assessment.

[0074] In this embodiment, the logic used to determine the signal enhancement sampling points is as follows:

[0075] The dielectric constant data of each sampling point is collected using ground-penetrating radar or electromagnetic detectors. The collected dielectric constant data is then imported into geographic information system software and correlated with the geographical location information of the sampling points. A threshold for the dielectric constant is set, and points with a value less than this threshold are used as signal enhancement sampling points.

[0076] Dielectric constant data were collected at each sampling point within the goaf of the coal mining face using ground-penetrating radar (GPR) or electromagnetic detectors. Dielectric constant is the characteristic of a material's response to an electric field, reflecting the material composition and state of the underground medium.

[0077] The collected dielectric constant data is imported into a Geographic Information System (GIS) software and associated with the geographic location information of each sampling point. This step ensures that each dielectric constant data point has its corresponding geographic location, facilitating subsequent spatial analysis.

[0078] Based on the distribution of dielectric constant within the region, a reasonable threshold is set. This threshold is typically based on analysis of historical data and experience, and may represent the lower limit of the normal range for dielectric constant. Points with dielectric constants below this threshold may indicate abnormal changes in the underground medium, such as decreased moisture or increased air content, which may be related to the oxidation and spontaneous combustion processes of coal seams.

[0079] Sampling points below a set threshold are marked as signal-enhanced sampling points. These points exhibit enhanced electromagnetic signals due to abnormal dielectric constants, thus requiring further attention. The dielectric constant reflects changes in the material properties of the underground medium. Setting a threshold filters out points with abnormally low dielectric constants, indicating potential risk areas. These abnormal points are marked as signal-enhanced sampling points, serving as key areas for further in-depth analysis. This logic effectively allows for the initial screening of potentially spontaneous combustion risk areas within a broad sampling region, improving the efficiency and accuracy of subsequent analysis.

[0080] Ground-penetrating radar (GPR) generates images of underground structures by emitting high-frequency electromagnetic waves and receiving their reflected signals. By analyzing these images, the dielectric constant variations at different levels can be inferred. GPR performs a systematic radar scan of the area to be measured, generating two-dimensional or three-dimensional images of the underground structure. GPR data processing software is used to process the acquired data, extracting the time and intensity of the reflected signals. Based on the time delay and intensity of the reflected signals, combined with known underground structure information, the dielectric constant at different locations is calculated. Electromagnetic detectors (EMDs) generate images of the electromagnetic properties of underground media by emitting low-frequency electromagnetic waves and measuring their conduction and reflection characteristics underground. EMDs perform a systematic electromagnetic scan of the area to be measured, acquiring continuous images of the underground electromagnetic properties. Specialized software is used to process the acquired data, generating two-dimensional or three-dimensional images of the underground structure. Based on the propagation speed and attenuation characteristics of electromagnetic waves, combined with known underground structure information, the dielectric constant at different locations can be inferred.

[0081] Although GPR and electromagnetic detectors typically provide continuous images, by selecting a local region of interest from the acquired image data, subdividing the data within that region, extracting the reflected signals and propagation characteristics near specific points, and then using the subdivided data to calculate the dielectric constant of the local points, it is possible to extract the dielectric constant information of local points from these images through appropriate processing and analysis, even though GPR and electromagnetic detectors primarily provide continuous electromagnetic property images.

[0082] Sampling points with enhanced signals typically have a higher risk of spontaneous combustion because the abnormally low dielectric constant at these points may indicate changes in the material within the underground coal seam, such as oxidation or thermal accumulation—changes that are precursors to spontaneous combustion. These changes may lead to easier oxygen penetration, promoting oxidation reactions and thus increasing the risk of spontaneous combustion. Furthermore, these material changes may be accompanied by increased temperature, further exacerbating the likelihood of spontaneous combustion. Therefore, identifying these signal-enhanced sampling points allows for earlier detection of potential spontaneous combustion areas, enabling effective monitoring and prevention.

[0083] Step 2: Capture the spectral data of each sampling point, set the wavelength analysis range, record the spectral intensity of each sampling point within the wavelength analysis range, map it with the geographical location information of each sampling point, and normalize the spectral intensity data of all sampling points.

[0084] Using a near-infrared spectrometer capable of covering the wavelength range of 900nm to 1000nm, ensure the equipment is calibrated before measurement to reduce measurement errors. Select specific sampling points and correctly position the spectrometer probe to ensure measurement accuracy and consistency. The spectrometer will scan within the set wavelength range and record the spectral intensity corresponding to each wavelength. Typically, this data is output in wavelength-intensity pairs, forming a complete spectral curve. The acquired data can be processed, including noise reduction and smoothing, to improve data quality. Then, spectral intensity data within the 900nm to 1000nm range is extracted. In this embodiment, the maximum spectral intensity value within this range is selected as representative and calibrated as the spectral intensity within the 900nm to 1000nm wavelength range.

[0085] In this embodiment, the wavelength analysis range set when recording the spectral intensity of each sampling point is 900nm to 1000nm. The specific logic for normalizing the spectral intensity data of all sampling points is as follows:

[0086] Obtain the spectral intensity of each sampling point within the wavelength range of 900nm to 1000nm, and calculate the maximum and minimum spectral intensities within this range for all sampling points. Specifically:

[0087] I max =Max[Iq1, Iq2, ..., Iq i ... I qn ]

[0088] I min =min[Iq1, Iq2, ..., Iq] i ... I qn ]

[0089] Among them, I max and Imin Iq represents the maximum and minimum spectral intensities within the wavelength range of 900 nm to 1000 nm at the sampling points, respectively. i This represents the spectral intensity data of sampling point i before normalization in the wavelength range of 900nm to 1000nm, where i represents the sampling point number and n represents the total number of sampling points.

[0090] The spectral intensities of all sampling points in the wavelength range of 900nm to 1000nm are scaled to the range of [0, 1], specifically as follows:

[0091]

[0092] I i This represents the normalized spectral intensity data of sampling point i within the wavelength range of 900nm to 1000nm.

[0093] Within the wavelength range of 900nm to 1000nm, the first step is to calculate the maximum value I of the spectral intensity at all sampling points. max and minimum value I min These values ​​represent the range of spectral intensities in the dataset. Determining the maximum and minimum values ​​lays the foundation for subsequent normalization. For each sampling point's spectral intensity value, a normalization formula is applied to scale each intensity value to between 0 and 1. This process ensures consistent scale across the data, facilitating subsequent analysis. The original spectral data may have different dimensions due to acquisition conditions, equipment, or other factors. Normalization eliminates the influence of dimensions on data analysis by standardizing the data range, making the intensity values ​​of different sampling points comparable on the same scale. In data analysis or model training, an excessively large range of original data can lead to unstable algorithm performance. Normalization adjusts the data to a uniform standard, enhancing the stability and convergence speed of the algorithm, making it easier for the model to capture the essential characteristics of the data. Through normalization, the spectral intensities of different sampling points are within the same range, greatly enhancing data comparability. This is not only beneficial for numerical analysis but also makes data visualization more intuitive, facilitating the identification of trends and patterns, thereby supporting deeper scientific analysis and decision-making.

[0094] Different wavelengths of light exhibit specific sensitivities to different substances and phenomena. The 900nm to 1000nm wavelength range, in the near-infrared region, is commonly used to detect temperature changes, flames, and radiation characteristics. Within this range, changes in thermal radiation can effectively reflect high-temperature phenomena such as spontaneous combustion and fires. Therefore, by monitoring the spectral intensity within this wavelength range, naturally hazardous areas can be effectively identified. In conclusion, selecting the 900nm to 1000nm wavelength range for analysis and normalizing the spectral intensity data can significantly improve the effectiveness and reliability of naturally hazardous area identification. Through the sensitivity of a specific wavelength range and the standardization of normalization processing, abnormal spectral changes within the area can be captured and analyzed more accurately, ensuring the efficient operation of the monitoring system.

[0095] Step 3: For the spectral intensity data of the normalized sampling points, calculate the relative radiance of each sampling point, calculate the mean and standard deviation of the relative radiance of all sampling points, generate the radiative shift of each sampling point, and identify the abnormal sampling points in the signal enhancement sampling points based on the radiative shift of the sampling points.

[0096] In this embodiment, the logic used to calculate the relative radiation intensity of each sampling point is as follows:

[0097]

[0098] Among them, R i Let I be the relative radiation intensity of sampling point i. b The baseline radiation intensity represents the radiation intensity within 900 nm to 1000 nm under the same temperature and pressure without spontaneous combustion.

[0099] To generate the radiation offset at each sampling point, the mean and standard deviation of the relative radiation intensity of all sampling points are first calculated. The z-score method is then used to generate the radiation offset at each sampling point, based on the following formula:

[0100]

[0101] Where, μ R and σ R Let z represent the mean and standard deviation of the relative radiation intensity at all sampling points, respectively. i This indicates the degree of radiation shift at sampling point i, when |z i |>z y Furthermore, when this point is a signal enhancement sampling point, the sampling point numbered i is identified as an abnormal sampling point, z yThe z-score represents the offset threshold. When using the z-score method to identify outlier sampling points, the reason for taking the absolute value of the z-score is that we are concerned with the degree of deviation of the sampling points, regardless of whether the deviation is positive or negative. By taking the absolute value, we can unify the positive and negative deviations, focusing only on the magnitude of the deviation and not considering the direction of the deviation. Absolute value processing makes the anomaly detection standard consistent and will not miss any significantly deviated data points due to different directions. This processing method enables the detection to capture both extremely high and extremely low value anomalies.

[0102] Relative radiation intensity specifically reflects the change in radiation intensity at a sampling point relative to the baseline radiation intensity. It represents the relative change between the radiation intensity at each sampling point and the baseline radiation intensity under non-spontaneous combustion conditions. This indicator allows observation of whether the radiation intensity has increased or decreased, and what changes it has compared to the baseline. A large relative radiation intensity indicates a significant increase in radiation intensity at that sampling point compared to the baseline, which may signify some change under those conditions, such as spontaneous combustion. A higher relative radiation intensity may indicate an increased risk of spontaneous combustion, as spontaneous combustion is usually accompanied by energy release, reflected in the increase in radiation intensity. A higher relative radiation intensity indicates a greater magnitude of energy release, potentially indicating the occurrence of a spontaneous combustion reaction. Spontaneous combustion involves chemical reactions, and the heat and light radiation released during spontaneous combustion increase. A significant increase in radiation intensity can be a sign of these reactions intensifying. Radiation intensity is closely related to temperature. Increased temperature is a crucial condition for spontaneous combustion; a high relative radiation intensity may indicate that the temperature has exceeded a certain threshold, leading to spontaneous combustion. Therefore, significant changes in radiation intensity above the baseline can serve as a basis for anomaly detection, suggesting a risk of spontaneous combustion. By monitoring changes in relative radiation intensity, potential spontaneous combustion risks can be identified in advance.

[0103] The baseline radiation intensity is a standard intensity value measured under conditions without spontaneous combustion. This is achieved by maintaining stable temperature and pressure in a well-controlled experimental environment (with standard values ​​set for temperature and pressure) and performing multiple spectral measurements in the absence of spontaneous combustion. For example, by maintaining constant temperature and pressure conditions in a laboratory and using a spectrometer to measure the radiation intensity in the wavelength range of 900 nm to 1000 nm, recording multiple measurements, and statistically processing the measured radiation intensity data to calculate the average value. This ensures a stable baseline value. If the measured radiation intensities are 100, 102, and 101 respectively, then the average of these measurements yields a baseline radiation intensity of 101. The offset threshold is determined through statistical analysis of historical data. During the analysis, the z-score distribution under normal conditions is observed, and a threshold that can effectively distinguish between normal and abnormal conditions is set. If the z-score of most normal sampling points is between -2 and 2, a threshold of 3 might be chosen as the offset threshold.

[0104] Step 4: Obtain the relative radiation intensity of neighboring sampling points of the abnormal sampling points in each signal enhancement sampling point, calculate the uniform distribution index of each abnormal sampling point, and mark the risk level of the abnormal sampling points according to the uniform distribution index.

[0105] In this embodiment, the neighboring sampling points of anomalies in the signal enhancement sampling points are the sampling points directly adjacent to the anomaly sampling point, including directly adjacent sampling points in the six directions of up, down, left, right, front, and back in the grid. The formula used to calculate the uniform distribution index of each anomaly point is as follows:

[0106]

[0107] Where, μz j and σz j Let Jf represent the mean and variance of the relative radiance of the dataset consisting of the outlier sample point j and its neighboring sample points, respectively. j R represents the uniformity distribution index of the outlier sampling point j. j Rs j Rx j 、Rz j Ry j 、Rq j and Rh j These represent the relative radiation intensities of the anomalous sampling point numbered j and its six neighboring sampling points (upper, lower, left, right, front, and rear). If an anomalous sampling point does not have at least one neighboring sampling point among the six neighboring sampling points, the relative radiation intensities of that sampling point are set to 0.

[0108] A higher radiation intensity at an anomaly point indicates an increased risk of spontaneous combustion at that point. A smaller variance means that the radiation intensity differences between neighboring points are smaller, indicating a more uniform distribution of radiation intensity among neighboring points. The larger the uniformity index, the more consistent the radiation intensity of not only the anomaly point itself but also its neighboring points. A higher radiation intensity at an anomaly point suggests a potential risk of localized high temperatures or fire. A smaller difference in radiation intensity between an anomaly point and its neighboring points indicates that the radiation intensity of these points is relatively uniform, suggesting that not only is the anomaly point itself highly radiated, but the points around it also have similarly high radiation intensity levels.

[0109] When the radiation intensity of a point is abnormally high and the radiation intensity of surrounding points is relatively uniform, this consistency may reflect a local anomaly area rather than just an isolated anomaly point. Small variance indicates that the data distribution of neighboring points is uniform. Combined with the anomaly point with high radiation intensity, this situation is more likely to be a real high-risk area rather than a random fluctuation or measurement error. The uniformity index, as a comprehensive indicator, combines the intensity and distribution characteristics of a local area to more accurately assess the spontaneous combustion risk of the area.

[0110] The uniformity distribution index can effectively identify anomalous points that not only have high radiation intensity themselves but also exhibit consistently high radiation intensity in their surrounding areas. This method can more accurately assess and mark high-risk areas, helping to provide early warnings and responses to potential hazards such as spontaneous combustion. The smaller the difference in radiation intensity between neighboring points, the higher the risk of spontaneous combustion. When an anomalous point has high radiation intensity and its neighboring points have small differences in radiation intensity, it indicates that the radiation intensity is relatively uniform and consistent within a local area. This usually means that it is not just an isolated anomalous point but may be a large area of ​​high radiation. Heat has the characteristic of diffusion in matter. If the radiation intensity of surrounding neighboring points is also relatively uniform and high, it may mean that there is no obvious concentration point in the area, but rather that the heat has diffused throughout the entire region. In this case, the temperature of the entire area is high, increasing the risk of spontaneous combustion. If only one point has high radiation intensity, it may be an accidental phenomenon. However, when multiple adjacent points show high radiation intensity with small differences, it indicates that the total heat at these points is high, and local heat accumulation is severe. Uniform and high radiation intensity means that the area is in a state of sustained high temperature. Under such conditions, materials or the environment are more likely to reach their spontaneous combustion point.

[0111] A single high radiation point indicates the possible existence of a single hot spot. However, if multiple adjacent points show high radiation with little difference, it indicates that the entire local area has a high level of thermal radiation. This suggests that the risk in the area is more concentrated and severe. In the early warning of spontaneous combustion risk in goaf areas, the spread of fire is often due to the gradual increase in temperature in a local area, forming a high-temperature zone. If the radiation intensity difference between neighboring points is small and high, it means that the fire may be spreading or has already spread to multiple points in the area.

[0112] When marking outliers with risk levels based on the uniform distribution index, if Jf j >Jf y If the abnormal sampling point is identified, it will be marked as a high-risk sampling point and labeled; otherwise, it will be marked as a low-risk sampling point and not labeled. y Let J represent the uniform distribution threshold, and j represent the number of the outlier sampling point in the signal enhancement sampling points. j >Jf yWhen the radiation intensity of an anomaly is significantly higher than that of its neighboring points, and the intensity differences between neighboring points are small, it indicates that there is a high probability of an anomaly in the area where the point is located, such as localized high temperatures or potential fires. Therefore, it is marked as a high-risk sampling point. Conversely, when the radiation intensity of an anomaly is not significantly different from that of its neighboring points, the risk is relatively low, and it is marked as a low-risk sampling point.

[0113] Through this logic, the system can more accurately identify potential high-risk areas, provide early warnings, concentrate resources and attention on marked high-risk areas, improve monitoring and management efficiency, reduce false alarm rates, analyze historical abnormal data, identify uniform distribution index values ​​corresponding to different risk levels, set reasonable uniform distribution thresholds, and test different uniform distribution thresholds through experiments in real-world scenarios to select values ​​that can effectively distinguish between high and low risks.

[0114] Step 5: Based on the distribution density of all marked abnormal sampling points in the initial geographic map, as well as the relative radiation intensity data and geographic location information of each marked abnormal sampling point, construct a risk assessment model, generate a spontaneous combustion risk index for the goaf of the coal mining face to be judged, and judge the spontaneous combustion risk of the goaf based on the spontaneous combustion risk index.

[0115] In this embodiment, based on the distribution density of all marked abnormal sampling points in the initial geographic map, the number of all sampling points and the number of marked abnormal sampling points are counted, and the distribution density of abnormal sampling points in the initial geographic map is calculated, specifically as follows:

[0116]

[0117] Where, ρ y represents the distribution density of anomalous sampling points in the initial geographic map, where n represents the total number of sampling points and M represents the number of marked anomalous sampling points.

[0118] Distribution density represents the proportion of anomalous sampling points out of all sampling points in the entire initial geographic map. It reflects the spatial distribution of anomalous sampling points within the study area. A higher distribution density indicates a higher proportion of anomalous sampling points in the entire geographic map. This may mean that there are more high-risk points in the area, such as high-temperature areas or fire hazards. A high distribution density means that anomalous points are not just localized phenomena, but may be widely distributed throughout the entire goaf area of ​​the working face. Distribution density provides a macro-level risk assessment indicator, helping to comprehensively understand the risk level within the goaf area of ​​the working face. By calculating the distribution density of anomalous sampling points, a comprehensive understanding of the spatial distribution of anomalous points in the initial geographic map can be obtained. The higher the distribution density, the more anomalous points there are in the area, indicating a higher potential risk.

[0119] Furthermore, the specific logic for constructing a risk assessment model and generating a spontaneous combustion risk index for the goaf of the coal mining face to be assessed is as follows:

[0120] Based on the geographical location information of each marked anomalous sampling point, its distance from the center of the coal mining face is calculated. A risk assessment model is constructed by combining the distribution density of anomalous sampling points on the initial geographic map and the relative radiation intensity data of the marked anomalous sampling points. The formula used is as follows:

[0121]

[0122] Wherein, RI represents the spontaneous combustion risk index, d m and Jf m Let represent the distance and relative radiation intensity between the marked anomalous sampling point numbered m and the center of the coal mining face, respectively. ∈ represents the denominator constant, and ∈>0. m represents the number of the marked anomalous sampling point, and e represents the base of the natural logarithm.

[0123] When determining the center of a coal mining face, for regularly shaped faces, the coordinates of its center point can be directly calculated. For example, for a rectangular coal mining face, its center point can be determined by the intersection of its diagonals. For irregularly shaped faces, it is difficult to determine the center point directly using geometric methods. Therefore, the centroid calculation method can be used. This method assumes that the coal mining face is divided into multiple small units, each with equal volume or area. The centroid of each unit (the centroid refers to the geometric center of each small unit) is calculated. The centroid of the entire coal mining face can be determined by averaging the coordinates of the centroids of all small units. The centroid is then marked as the center. When calculating the distance between the abnormal sampling point and the center of the coal mining face, since the distance is small, the spherical properties of the Earth do not need to be considered, and the Euclidean distance formula in three-dimensional space can be used.

[0124] Anomalies with high radiation intensity have a higher risk of spontaneous combustion because they indicate greater localized heat accumulation. (Jf) m As a molecule, it directly reflects the risk level of the anomaly point. Generally, the farther away from the center of the coal face, the lower the risk of spontaneous combustion. m The existence of this reflects the risk attenuation effect of distance. The formula uses d... m e This makes the impact of distance on the risk index more significant, especially when the distance is large. The constant ∈ avoids the case where the denominator is zero, while ensuring the stability of the model and the feasibility of the calculation. The distribution density reflects the distribution of outliers in the whole region. A high density of outliers means that the overall risk in the region is higher. Therefore, the distribution density is used as a multiplier to further adjust the risk index so that it can better reflect the actual situation.

[0125] As the distance between the abnormal sampling point and the center of the coal face increases, the risk of spontaneous combustion generally decreases. (Using d) m e It can be emphasized that distance has a diminishing effect on risk. The exponent e is the base of the natural logarithm, which makes the influence of distance non-linear and more accurately simulates the actual physical situation. In this way, the influence of points that are farther away on the risk index is significantly weakened.

[0126] The ∈ ensures that the denominator is not zero, thus avoiding computational instability. It is usually set to a positive number less than but close to zero. The value of ∈ needs to ensure the numerical stability of the model. It can usually be adjusted according to specific data to find a suitable value. The common value range may be 0.01 to 1. With these settings, the model can more stably and accurately assess the spontaneous combustion risk index of the goaf in the coal mining face.

[0127] By summing the risk contributions of all anomalies, the overall spontaneous combustion risk index can be obtained by comprehensively considering the impact of multiple anomalies on the entire coal mining face area. Relative radiation intensity directly affects the risk index, reflecting the risk level of anomalies. Distance attenuation reduces the impact of anomalies farther from the center on the overall risk, which is consistent with the actual physical meaning. Distribution density, as an amplification factor, reflects the impact of the spatial distribution of anomalies on the overall risk. The spontaneous combustion risk index is a comprehensive indicator that combines relative radiation intensity, geographical location, and distribution density, and can comprehensively and accurately assess the spontaneous combustion risk of the coal mining face.

[0128] A higher spontaneous combustion risk index indicates a greater risk of spontaneous combustion in the goaf of the coal mining face. The warning signal indicates a potential spontaneous combustion hazard and suggests the need for preventative measures to reduce the likelihood of spontaneous combustion accidents. By constructing the aforementioned risk assessment model, factors such as the relative radiation intensity of abnormal sampling points, their distance from the center of the coal mining face, and their distribution density can be comprehensively considered to generate a comprehensive spontaneous combustion risk index. This comprehensive index can help to more accurately assess the spontaneous combustion risk of the goaf in the coal mining face, supporting scientific decision-making and effective risk management.

[0129] In this embodiment, when judging the spontaneous combustion risk of the goaf based on the spontaneous combustion risk index, if RI≥RI y If the risk of spontaneous combustion in the goaf is deemed high, an early warning is issued. <RI y If the risk of spontaneous combustion in the goaf is deemed low, no warning will be issued. y This indicates the threshold for spontaneous combustion risk.

[0130] When RI≥RI yA high-risk warning signal is issued, indicating a high risk of spontaneous combustion in the goaf of the coal mining face. Immediate measures are required, such as increased monitoring, improved ventilation, and the application of extinguishing agents, to prevent spontaneous combustion accidents. <RI y The risk of spontaneous combustion in the goaf of a coal mining face is low and no immediate emergency measures are required. However, regular monitoring and maintenance are still necessary to prevent a sudden increase in risk. The spontaneous combustion risk threshold is determined by collecting data on historical spontaneous combustion events and using statistical methods (such as distribution analysis and regression analysis) to determine a significant threshold.

[0131] Geographic information data of the goaf area of ​​the coal mining face was imported using Geographic Information System (GIS) software (such as ArcGIS or QGIS) to generate an initial geographic map and create a regular 3D grid. Each grid intersection was used as a sampling point, and its geographic location information was recorded. This step ensured data consistency and spatial positioning accuracy, providing an accurate foundation for subsequent analysis. Near-infrared spectrometers were used to capture spectral data for each sampling point in the wavelength range of 900nm to 1000nm, and the spectral intensity was recorded. Subsequently, the spectral intensity data of all sampling points were normalized, adjusting the data to the range of 0 to 1 to ensure that the spectral intensities of different sampling points were comparable on the same scale.

[0132] By calculating the relative radiance at each sampling point and using the z-score method to generate the degree of radiance offset, anomalous sampling points are identified. Relative radiance reflects the difference between the sampling point and the baseline radiance, while the z-score helps quantify this difference, thereby accurately identifying areas with a high risk of spontaneous combustion. For each anomalous sampling point, the relative radiance of its neighboring sampling points is obtained, and a uniform distribution index is calculated. Based on the uniform distribution index, anomalous sampling points are labeled with risk levels, distinguishing between high-risk and low-risk sampling points. The uniform distribution index not only considers the radiance intensity of the anomalous point but also incorporates the data consistency of neighboring areas, providing a more reliable risk assessment standard.

[0133] By combining the distribution density, relative radiation intensity, and geographical location information of all marked abnormal sampling points, a risk assessment model is constructed to generate a spontaneous combustion risk index. The spontaneous combustion risk in the goaf is determined based on this index, leading to early warnings or continued monitoring. This comprehensive assessment method allows for a thorough and accurate evaluation of the spontaneous combustion risk in coal mining faces, providing a solid basis for scientific decision-making and risk management.

[0134] Please see Figure 2 The present invention also provides a system for determining the spontaneous combustion hazard zone in the goaf of a coal mining face. The system is used to execute the above-mentioned method for determining the spontaneous combustion hazard zone in the goaf of a coal mining face, including:

[0135] The geographic map creation module is used to import geographic information data of the goaf area of ​​the coal mining face to be judged using geographic information system software to generate an initial geographic map. A regular grid is created within the initial geographic map, and each grid intersection is used as a sampling point. The geographic location information of the sampling points is obtained and numbered. The dielectric constant of each sampling point is collected to determine the signal enhancement sampling points.

[0136] The preprocessing module is used to capture spectral data from each sampling point, set the wavelength analysis range, record the spectral intensity of each sampling point within the wavelength analysis range, map it to the geographical location information of each sampling point, and normalize the spectral intensity data of all sampling points.

[0137] The anomaly detection module is used to calculate the relative radiance of each sampling point for the spectral intensity data of the normalized sampling points, calculate the mean and standard deviation of the relative radiance of all sampling points, generate the radiance shift of each sampling point, and identify the abnormal sampling points in the signal enhancement sampling points based on the radiance shift of the sampling points.

[0138] The risk assessment module is used to obtain the relative radiation intensity of neighboring sampling points of abnormal sampling points in each signal enhancement sampling point, calculate the uniform distribution index of each abnormal sampling point, and mark the risk level of abnormal sampling points according to the uniform distribution index.

[0139] The comprehensive judgment module is used to construct a risk assessment model based on the distribution density of all marked abnormal sampling points in the initial geographic map, as well as the relative radiation intensity data and geographical location information of each marked abnormal sampling point, and to generate a spontaneous combustion risk index for the goaf of the coal mining face to be judged. The spontaneous combustion risk of the goaf is judged based on the spontaneous combustion risk index.

[0140] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0141] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0142] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0143] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for judging a spontaneous combustion danger zone of a goaf of a coal mining face, characterized in that, The specific steps include: Step 1: Import the geographic information data of the goaf of the coal mining face to be judged into the geographic information system software to generate an initial geographic map, create a regular grid in the initial geographic map, take each grid intersection as a sampling point, obtain the geographic position information of the sampling point, and number it, collect the dielectric constant of each sampling point, and determine the signal enhancement sampling point; Step 2: Capture the spectral data of each sampling point, set the wavelength analysis range, record the spectral intensity of each sampling point within the wavelength analysis range, and map it with the geographic position information of each sampling point, and normalize the spectral intensity data of all sampling points; Step 3: For the normalized spectral intensity data of the sampling points, calculate the relative radiation intensity of each sampling point, calculate the mean and standard deviation of the relative radiation intensity of all sampling points, generate the radiation deviation degree of each sampling point, and identify the abnormal sampling points in the signal enhancement sampling points according to the radiation deviation degree of the sampling points; Step 4: Obtain the relative radiation intensity of the neighbor sampling points of each abnormal sampling point in the signal enhancement sampling point, calculate the uniform distribution index of each abnormal sampling point, and mark the risk level of the abnormal sampling points according to the uniform distribution index; Step 5: According to the distribution density of all marked abnormal sampling points in the initial geographic map, as well as the relative radiation intensity data and geographic position information of each marked abnormal sampling point, a risk assessment model is constructed to generate a spontaneous combustion risk index of the goaf of the coal mining face to be judged, and the spontaneous combustion risk of the goaf is judged according to the spontaneous combustion risk index.

2. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 1, characterized in that: The geographic information data of the goaf of the coal mining face includes the shape, range and elevation data of the goaf, and the geographic coordinate data of the goaf, and the geographic position information of the sampling point is obtained as follows: Import the shape, range, elevation and geographic coordinate data of the goaf into the geographic information system software, ensure that all imported data sets use a unified coordinate system, and if they are inconsistent, convert the coordinate systems; Use the goaf shape data to crop the elevation data to ensure that only the goaf area is retained, set the size and range of the grid, generate a regular grid according to the analysis requirements, cover the goaf, extract the grid intersections as sampling points, and record their geographic position information. The grid is a three-dimensional grid.

3. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 2, characterized in that: The logic for determining the signal enhancement sampling point is as follows: Use the geological radar or electromagnetic detector to collect the dielectric constant data of each sampling point, import the collected dielectric constant data into the geographic information system software, associate it with the geographic position information of the sampling point, set a threshold value for the dielectric constant, and points less than this value are signal enhancement sampling points.

4. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 1, characterized in that: The wavelength analysis range set when recording the spectral intensity of each sampling point is 900nm to 1000nm, and the specific logic for normalizing the spectral intensity data of all sampling points is as follows: Obtain the spectral intensity of each sampling point within the wavelength range of 900nm to 1000nm, calculate the maximum and minimum values of the spectral intensity within the wavelength range of 900nm to 1000nm of all sampling points, and the specific process is as follows: I max = max [ Iq1, Iq2,..., Iq i ,..., Iq n ] I min = min [Iq1, Iq2,..., Iq i ,..., Iq n ] where I max and I min respectively represent the maximum and minimum values of the spectral intensity in the wavelength range of 900 nm to 1000 nm in the sampling points, Iq i represents the spectral intensity data before normalization of the sampling point numbered i in the wavelength range of 900 nm to 1000 nm, i represents the number of the sampling point, and n represents the total number of the sampling points; The spectral intensity of all the sampling points in the wavelength range of 900nm to 1000nm is scaled to the range of [0, 1], specifically: I i represents the normalized spectral intensity data of the sampling point numbered i in the wavelength range from 900 nm to 1000 nm.

5. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 4, characterized in that: The logic for calculating the relative radiation intensity of each sampling point is: wherein R i is the relative intensity of the radiation of the sampling point numbered i, I b represents the baseline intensity, representing the intensity of the radiation in the range from 900 nm to 1000 nm at the same temperature and pressure intensity, in the absence of spontaneous combustion phenomena; When generating the radiation deviation degree of each sampling point, first, the mean and standard deviation of the relative radiation intensity of all the sampling points are calculated, and the radiation deviation degree of each sampling point is generated using the z-score method, and the formula is: wherein μ R and σ R respectively represent the mean and standard deviation of the relative radiation intensity of all sampling points, z i represents the radiation deviation degree of the sampling point numbered i, when |z i |>z y , and the point is a signal-enhanced sampling point, the sampling point numbered i is identified as an abnormal sampling point, and z y represents a deviation threshold.

6. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 1, characterized in that: The neighbor sampling points of the abnormal sampling points in the signal enhanced sampling points are the sampling points directly adjacent to the abnormal sampling points, including the sampling points directly adjacent in the six directions of up, down, left, right, front and back in the grid, and the formula for calculating the uniform distribution index of each abnormal point is: wherein μz j and σz j respectively represent the mean and variance of the relative radiation intensity of the data set consisting of the abnormal sampling point numbered j and its adjacent sampling points, Jf j represents the uniform distribution index of the abnormal sampling point numbered j, R j , Rs j , Rx j , Rz j , Ry j , Rq j , and Rh j respectively represent the relative radiation intensity of the abnormal sampling point numbered j and its upper, lower, left, right, front, and rear six sampling points, and if the abnormal sampling point does not have at least one neighbor sampling point in the upper, lower, left, right, front, and rear six sampling points, the relative radiation intensity of the sampling point is set to 0. According to the uniform distribution index, the risk level of the abnormal point is marked, if Jf j >Jf y , the abnormal sampling point is marked as a high-risk sampling point, and the marking is performed, otherwise the sampling point is marked as a low-risk sampling point, and the marking is not performed, Jf y represents the uniform distribution threshold, and j represents the number of abnormal sampling points in the signal enhancement sampling point.

7. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 6, characterized in that: When determining the distribution density of all the labeled abnormal sampling points in the initial geographic map, the number of all the sampling points and the number of the labeled abnormal sampling points are counted, the distribution density of the abnormal sampling points in the initial geographic map is calculated, and the formula is: wherein p y is the distribution density of the abnormal sampling points in the initial geographic map, n represents the total number of sampling points, and M represents the number of marked abnormal sampling points.

8. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 7, characterized in that: The specific logic for constructing a risk assessment model to generate the spontaneous combustion risk index of the goaf of the coal mining face to be judged is: According to the geographic location information of each labeled abnormal sampling point, the distance from the center of the coal mining face is calculated, and a risk assessment model is constructed in combination with the distribution density of the abnormal sampling points in the initial geographic map and the relative radiation intensity data of the labeled abnormal sampling points, and the formula is: wherein, RI represents the risk index of spontaneous combustion, d m and Jf m respectively represent the distance and relative radiation intensity of the labeled abnormal sampling point numbered m from the center of the coal mining face, represents the denominator constant, and, m represents the number of the labeled abnormal sampling point, and e represents the base number of the natural logarithm.

9. The method for judging the spontaneous combustion danger zone of the goaf of the coal mining face according to claim 8, characterized in that: When judging the spontaneous combustion risk of the goaf according to the spontaneous combustion risk index, if RI≥RI y , it is judged that the spontaneous combustion risk of the goaf is high, a warning is issued, if RI<RI y , it is judged that the spontaneous combustion risk of the goaf is low, no warning is issued, and RI y represents a spontaneous combustion risk threshold value.

10. A system for determining a spontaneous combustion danger zone of a goaf of a coal mining face, characterized in that: The judgment system is used to execute the judgment method of the spontaneous combustion danger area of the goaf of the coal mining face according to any one of claims 1-9, comprising: A geographic map creation module is configured to import geographic information data of a goaf of a coal mining face to be judged into an initial geographic map using geographic information system software, create a regular grid in the initial geographic map, take each grid intersection as a sampling point, obtain geographic location information of the sampling points, and number the sampling points, collect dielectric constants of the sampling points, and determine signal enhanced sampling points; A preprocessing module is configured to capture spectral data of the sampling points, set a wavelength analysis range, record spectral intensity of each sampling point in the wavelength analysis range, map the spectral intensity with the geographic location information of each sampling point, and normalize spectral intensity data of all the sampling points; An abnormality identification module is configured to calculate relative radiation intensity of each sampling point based on the normalized spectral intensity data of the sampling points, calculate a mean and a standard deviation of the relative radiation intensity of all the sampling points, generate a radiation deviation degree of each sampling point, and identify abnormal sampling points in the signal enhanced sampling points based on the radiation deviation degree of the sampling points; A risk assessment module is configured to obtain relative radiation intensity of neighbor sampling points of each abnormal sampling point in the signal enhanced sampling points, calculate a uniform distribution index of each abnormal sampling point, and label the abnormal sampling points according to the risk level based on the uniform distribution index; A comprehensive judgment module is configured to construct a risk assessment model based on distribution density of all the labeled abnormal sampling points in the initial geographic map, and geographic location information and relative radiation intensity data of each labeled abnormal sampling point, generate a spontaneous combustion risk index of a goaf of a coal mining face to be judged, and judge the spontaneous combustion risk of the goaf based on the spontaneous combustion risk index.

Images