An engineering coal rock mass disaster precursor identification method, device and electronic equipment

Abstract

The application discloses an engineering coal rock mass disaster precursor identification method and device and electronic equipment. It relates to the technical field of rock water inrush precursor monitoring. The method comprises the following steps: collecting infrared thermal image data, acoustic emission data and hydraulic data of a target engineering coal rock mass sample in a hydraulic coupling loading process by using a preset hydraulic coupling loading device; using a dynamic low-temperature threshold and a benchmark low-temperature deviation amount combined judgment feature extraction method to extract low-temperature features from the infrared thermal image data to obtain infrared low-temperature cluster characteristic parameters representing the spatio-temporal evolution of infrared radiation; extracting features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission characteristic parameters and hydraulic characteristic parameters; identifying the disaster precursor of the target engineering coal rock mass sample based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters and the hydraulic characteristic parameters to obtain an identification result. The method can improve the accuracy of rock water inrush precursor identification.

Description

Technical Field

[0001] This invention relates to the field of monitoring precursors of rock water inrush, and particularly to a method, device, and electronic equipment for identifying precursors of coal and rock mass disasters. Background Technology

[0002] In underground engineering, mining, tunnel excavation, deep cavern excavation, and disturbance of water-rich rock masses, rocks are typically subjected to the coupled effects of geostress, engineering disturbance, and groundwater pressure. When internal rock fissures initiation, propagation, connection, and penetration occur under external loads and water pressure, internal seepage channels gradually form, potentially inducing seepage, water inrush, or even sudden water inrush disasters. Sudden water inrush disasters are characterized by their suddenness, large destructive range, and high difficulty in remediation. Therefore, identifying reliable precursory information before a sudden water inrush occurs is a crucial issue in geotechnical engineering safety monitoring and disaster early warning.

[0003] Infrared thermography, as a non-contact, full-field monitoring method, can continuously acquire information on the temperature field of rock surfaces. It offers advantages such as a large field of view, simple setup, and minimal disturbance to the sample. During the loading, fracturing, and seepage evolution of rocks, processes such as internal crack friction, elastic energy release, seepage heat transfer, moisture migration, and localized evaporation can all trigger changes in the surface temperature field. Therefore, infrared thermography information can be used to characterize precursory processes of rock fracturing, seepage, and water inrush.

[0004] However, existing infrared monitoring methods mostly focus on analyzing average temperature, maximum temperature, minimum temperature, local temperature difference, or temperature anomalies at a single moment. For precursors of rock water inrush under water-rich conditions, truly meaningful anomalies are often not isolated low-temperature points, but rather a process in which low-temperature anomaly areas continuously appear over time, gradually accumulate and expand spatially, and connect with potential cracks or boundaries. If only fixed thresholds or fixed percentiles are used to segment low-temperature points, they are easily affected by overall temperature drift, environmental disturbances, reflected noise, and random thermal noise, resulting in misjudgments or missed detections.

[0005] On the other hand, existing simulation tests of rock or coal fracture and water inrush have included simultaneous acquisition schemes for infrared, acoustic emission, water pressure, flow rate, deformation, or visible light images. Infrared radiation variance can also be used for fracture monitoring. However, simply combining multiple sensors in parallel or vaguely proposing "multi-parameter fusion early warning" easily remains at the level of conventional monitoring systems, making it difficult to form a clear, stable, and innovative technical solution. Summary of the Invention

[0006] In view of this, the present invention provides a method, device and electronic equipment for identifying precursors of coal and rock mass disasters. The main purpose is to solve the problems of inaccurate identification of precursors of rock water inrush caused by fixed low temperature thresholds, obvious false positives in the stable stage, large random thermal noise interference, insufficient utilization of the spatial organization and evolution of low temperature anomalies, and unclear boundaries of multi-parameter early warning rules.

[0007] To address the aforementioned problems, this application provides a method for identifying precursors of coal and rock mass disasters in engineering projects, comprising: Infrared thermographic data, acoustic emission data, and hydraulic data are collected during the hydraulic coupling loading process on the target engineering coal and rock mass sample using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. The infrared thermal image data is subjected to low-temperature feature extraction using a feature extraction method that combines dynamic low-temperature threshold and baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. Feature extraction is performed on the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters; Based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters, the catastrophic precursors of the target engineering coal and rock mass sample are identified, and the identification results are obtained.

[0008] Optionally, the feature extraction method using a combination of dynamic low-temperature threshold and baseline low-temperature deviation to extract low-temperature features from the infrared thermal image data, obtaining infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation, specifically includes: The relative temperature field of the infrared thermal image data is constructed based on a preset reference temperature; The dynamic low temperature threshold is obtained by performing calculations based on the relative temperature distribution of the relative temperature field. Obtain the average relative temperature field during the hydraulic coupling loading reference stage and determine the reference low temperature deviation; The infrared thermal image data is used to extract low-temperature features using the dynamic low-temperature threshold and the baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

[0009] Optionally, the calculation and processing based on the relative temperature distribution of the relative temperature field to obtain the dynamic low temperature threshold specifically includes: The current dynamic percentile threshold is determined based on the target percentile value of the relative temperature distribution of the effective region of the relative temperature field at the current moment; The median and median absolute deviation of the temperature field are determined based on the relative temperature distribution of the effective region of the relative temperature field at the current moment. The current statistical deviation low temperature threshold is obtained by calculating based on the median of the temperature field, the median absolute deviation, and the predetermined deviation coefficient. The current dynamic percentile threshold and the current statistical deviation low temperature threshold are filtered, and the minimum value between the current dynamic percentile threshold and the current statistical deviation low temperature threshold is determined as the dynamic low temperature threshold.

[0010] Optionally, the step of extracting low-temperature features from the infrared thermal image data using the dynamic low-temperature threshold and the baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation specifically includes: Candidate low-temperature points are obtained by filtering the infrared thermal imaging data based on the dynamic low-temperature threshold and the baseline low-temperature deviation. The number of candidate low-temperature points at different times at the same location of the target engineering coal and rock mass sample is counted to obtain the continuous low-temperature points of the target engineering coal and rock mass sample. Based on the sustained low temperature points, spatial connectivity analysis is performed to obtain effective low temperature points; Based on the effective low temperature points, a neighborhood analysis is performed to obtain low temperature clusters; Feature extraction is performed on the low-temperature clusters to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

[0011] Optionally, the step of extracting features from the cryogenic cluster to obtain infrared cryogenic cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation specifically includes: The infrared radiation variance characteristic parameters are calculated based on the total number of effective pixels in the effective area, the relative temperature of each effective pixel, and the average relative temperature of the effective area. The area ratio characteristic parameter of low temperature points is obtained by calculating the number of effective low temperature points at different times and the total number of effective pixels in the effective area. Based on the area of ​​each low-temperature cluster and the area of ​​the effective region, the characteristic parameter of the maximum low-temperature cluster area ratio is obtained. Based on the maximum low-temperature cluster centroid coordinates at different times and the predetermined time window length, the characteristic parameters of the maximum low-temperature cluster centroid migration amount are obtained. The dispersion or aggregation characteristics of the low-temperature anomaly region at different times are determined based on the number of low-temperature clusters at different times. Based on the spatial overlap relationship and / or minimum distance between the maximum low-temperature cluster and the target boundary region, the connectivity index of the low-temperature cluster boundary is determined; The infrared low-temperature cluster characteristic parameters include the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, the maximum low-temperature cluster centroid migration characteristic parameter, the dispersion or aggregation state characteristic parameter, and the low-temperature cluster boundary connectivity index.

[0012] Optionally, the step of extracting features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters specifically includes: Feature extraction is performed on the acoustic emission data to obtain the acoustic emission event rate and the acoustic emission energy growth rate. Feature extraction is performed on the hydraulic data to obtain the water pressure change rate and / or flow rate change rate; The acoustic emission characteristic parameters include the acoustic emission event rate and the acoustic emission energy growth rate. The hydraulic characteristic parameters include the water pressure change rate and / or the flow rate change rate.

[0013] Optionally, the step of identifying the pre-disaster precursors of the target engineering coal and rock mass sample based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters to obtain the identification result specifically includes: The baseline mean and standard deviation are calculated for the continuous characteristic parameters, acoustic emission characteristic parameters and hydraulic characteristic parameters in the infrared low temperature cluster characteristic parameters, respectively, and anomaly judgment functions based on the baseline mean and standard deviation are used to judge the anomalies of the continuous characteristic parameters, acoustic emission characteristic parameters and hydraulic characteristic parameters. Based on the degree of deviation of the dispersion or aggregation state characteristic parameters from the average number of low-temperature clusters in the baseline stage, anomaly judgment is made on the dispersion or aggregation state characteristic parameters. Based on the current value of the low-temperature cluster boundary connectivity index, an anomaly judgment is made on the low-temperature cluster boundary connectivity index. When the low-temperature cluster boundary connectivity index is equal to 1, it is determined that it has reached an abnormal state. Based on the anomaly judgment results corresponding to different characteristic parameters, the catastrophic precursors of the target engineering coal and rock mass sample are identified, and the identification results are obtained. The continuous characteristic parameters include the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, and the maximum low-temperature cluster centroid migration characteristic parameter.

[0014] Optionally, the identification of catastrophic precursors in the target engineering coal and rock mass sample based on the anomaly judgment results corresponding to different feature parameters, to obtain the identification results, specifically includes: When the anomaly judgment result is that any two of the infrared radiation variance characteristic parameter, the low temperature point area ratio characteristic parameter, the maximum low temperature cluster area ratio characteristic parameter, and the dispersion or aggregation state characteristic parameter are abnormal and the anomaly duration is greater than or equal to the length of the first preset time window, it is determined that the current state is in a level one warning state and a level one warning is issued. When the anomaly judgment result is that any three of the following three characteristic parameters are abnormal: the low temperature point area ratio characteristic parameter, the maximum low temperature cluster area ratio characteristic parameter, the maximum low temperature cluster centroid migration characteristic parameter, and the acoustic emission event rate and acoustic emission energy growth rate characteristic parameters, and the anomaly duration is greater than or equal to the length of the second preset time window, it is determined that the current state is in a level two warning state, and a level two warning is issued. When the connectivity index of the low-temperature cluster boundary reaches an abnormal state, at least one of the characteristic parameters of the acoustic emission event rate and the acoustic emission energy growth rate is abnormal, the collected hydraulic characteristic parameters are abnormal, and at least one of the characteristic parameters of the infrared radiation variance, the low-temperature point area ratio, the maximum low-temperature cluster area ratio, the maximum low-temperature cluster centroid migration, and the dispersion or aggregation state is abnormal, and the above abnormalities continue synchronously within a third preset time window; it is determined that the current state is in a level three warning state, and a level three warning is issued.

[0015] To address the aforementioned problems, this application provides an engineering coal and rock mass disaster precursor identification device, comprising: The acquisition module is used to acquire infrared thermal image data, acoustic emission data and hydraulic data during the hydraulic coupling loading process of the target engineering coal and rock mass sample using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. The first feature extraction module is used to extract low-temperature features from the infrared thermal image data using a feature extraction method that combines dynamic low-temperature threshold and benchmark low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. The second feature extraction module is used to extract features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters. The identification module is used to identify the catastrophic precursors of the target engineering coal and rock mass sample based on the infrared low-temperature cluster feature parameters, the acoustic emission feature parameters, and the hydraulic feature parameters, and to obtain the identification results.

[0016] To address the aforementioned problems, this application provides an electronic device, comprising at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the aforementioned method for identifying precursors of coal and rock mass disasters in engineering.

[0017] The beneficial effects of this application are as follows: This invention adopts a joint determination method of dynamic low temperature threshold and benchmark low temperature deviation, avoiding the problem that a fixed proportion of candidate low temperature points will naturally be generated in the stable stage by a simple percentile threshold; This application couples the infrared low temperature cluster evolution with acoustic emission, water pressure and flow parameters within a unified time window, so that the identification of water inrush precursors no longer depends on a single temperature value or empirical observation, thereby improving the accuracy of rock water inrush precursor identification.

[0018] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more easily understood, specific embodiments of the present invention are described below. Attached Figure Description

[0019] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a method for identifying precursors of coal and rock mass disasters in engineering, provided in an embodiment of this application, is shown. Figure 2 A schematic diagram of the structure of a preset hydraulic coupling loading device provided in another embodiment of this application is shown; Figure 3 A flowchart illustrating a method for identifying precursors of coal and rock mass disasters in engineering, according to another embodiment of this application, is shown. Figure 4 A structural block diagram of an engineering coal and rock mass disaster precursor identification device provided in another embodiment of this application is shown.

[0020] The components include: 1. Target coal and rock mass sample; 2. Cylindrical cavity inside the sample; 3. Steel water injection bend; 4. Water pressure loading unit; 5. Mechanical loading device; 6. Infrared thermal imager; 7. Acoustic emission sensor; 8. Water pressure sensor; 9. Flow sensor; and 10. Data acquisition and processing unit. Detailed Implementation

[0021] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0022] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0023] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0024] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0025] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.

[0026] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0027] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0028] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0029] This application provides a method for identifying precursors of coal and rock mass disasters in engineering applications, such as... Figure 1 As shown, it includes: Step S101: Collect infrared thermographic data, acoustic emission data, and hydraulic data of the target engineering coal and rock mass sample during the hydraulic coupling loading process using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. In the specific implementation process of this step, the structural schematic diagram of the preset hydraulic coupling loading device is as follows: Figure 2As shown, the infrared thermal imager was positioned 1 m directly in front of the observation surface of the rock sample on a tripod, at the same height as the rock sample, with a frame rate set to 10 frames / s. The infrared thermal imager and servo press were then simultaneously started. After the rock became unstable and failed, the equipment was simultaneously shut down, and the infrared thermal images, load data, and time-series data were saved. During the experiment, the infrared thermal imager continuously acquired thermal image sequences of the observation surface of the rock sample, obtaining infrared thermal image data; mechanical and hydraulic loads were applied to the rock sample, and acoustic emission data and hydraulic data were simultaneously acquired.

[0030] Step S102: Use the feature extraction method that combines dynamic low temperature threshold and baseline low temperature deviation to extract low temperature features from the infrared thermal image data to obtain infrared low temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. In this step, a relative temperature field of the infrared thermal image data is constructed based on a preset reference temperature. The relative temperature distribution of this field is then processed to obtain a dynamic low-temperature threshold. The average relative temperature field during the hydraulic coupling loading reference stage is acquired, and the reference low-temperature deviation is determined. The dynamic low-temperature threshold and the reference low-temperature deviation are then used to extract low-temperature features from the infrared thermal image data, yielding infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

[0031] Step S103: Perform feature extraction on the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters; In this step, feature extraction is performed on the acoustic emission data to obtain the acoustic emission event rate and the acoustic emission energy growth rate; feature extraction is performed on the hydraulic data to obtain the water pressure change rate and / or the flow rate change rate; wherein, the acoustic emission feature parameters include the acoustic emission event rate and the acoustic emission energy growth rate; the hydraulic feature parameters include the water pressure change rate and / or the flow rate change rate.

[0032] Step S104: Based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters, identify the catastrophic precursors of the target engineering coal and rock mass sample, and obtain the identification results.

[0033] In the specific implementation process, this step uses corresponding anomaly criteria to judge anomalies based on the data type of different characteristic parameters, and identifies the precursors of rock water inrush based on the anomaly judgment results of each characteristic parameter.

[0034] This application adopts a joint determination method of dynamic low temperature threshold and baseline low temperature deviation to avoid the problem that a fixed proportion of candidate low temperature points will naturally be generated in the stable stage by a simple percentile threshold. This application couples the infrared low temperature cluster evolution with acoustic emission, water pressure and flow parameters within a unified time window, so that the identification of water inrush precursors no longer depends on a single temperature value or empirical observation, thereby improving the accuracy of rock water inrush precursor identification.

[0035] Another embodiment of this application provides a different method for identifying precursors of coal and rock mass disasters in engineering projects, such as... Figure 3 As shown, it includes: Step S301: Collect infrared thermographic data, acoustic emission data, and hydraulic data of the target engineering coal and rock mass sample during the hydraulic coupling loading process using a preset hydraulic coupling loading device; In the specific implementation process of this step, the structural schematic diagram of the preset hydraulic coupling loading device is as follows: Figure 2As shown, the infrared thermal imager was positioned 1 m directly in front of the observation surface of the rock sample via a tripod, at the same height as the rock sample, with a frame rate of 10 frames / s. The infrared thermal imager and servo press were then simultaneously started. After the rock became unstable and failed, the equipment was simultaneously shut down, and the infrared thermal images, load data, and time-series data were saved. During the experiment, the infrared thermal imager continuously acquired thermal image sequences of the observation surface of the rock sample, obtaining infrared thermal image data; mechanical and hydraulic loads were applied to the rock sample, and acoustic emission data and hydraulic data were simultaneously acquired. The hydraulic coupling loading process for the target engineering coal and rock mass sample includes the following steps: First, preparing the engineering coal and rock mass sample. The engineering coal and rock mass is processed into a target engineering coal and rock mass sample 1 with a regular geometric shape, and a cylindrical cavity 2 of a certain size is prefabricated inside the sample to simulate the internal fracture space or water storage area of ​​the engineering coal and rock mass. Second, installing the hydraulic loading structure. A steel water injection bend 3 is installed at the inlet of the cylindrical cavity 2 inside the sample and sealed to connect the steel water injection bend 3 with the interior of the cylindrical cavity 2 inside the sample. The steel water injection bend 3 is used to supply water to the cylindrical cavity 2 inside the sample and to achieve pressure transmission between the water pressure loading unit 4 and the internal space of the sample. The third step is to establish a water pressure monitoring system. The water pressure loading unit 4 is connected to the steel water injection bend 3, and water is injected into the cylindrical cavity 2 inside the sample through the water pressure loading unit 4. A water pressure sensor 8 is used to monitor the changes in water pressure inside the sample in real time. Simultaneously, a flow sensor 9 is placed at the seepage channel location to monitor changes in seepage flow rate during the test. The fourth step is to install mechanical loading and monitoring equipment. The target engineering coal and rock mass sample 1, after hydraulic connection, is placed at the center of the mechanical loading device 5, and a predetermined mechanical load is applied to the sample through the mechanical loading device 5. Simultaneously, an infrared thermal imager 6 is positioned directly in front of the sample observation surface to collect data on the temperature field changes on the sample surface during loading; an acoustic emission sensor 7 is installed on the sample surface to acquire acoustic emission signals generated during crack propagation. The fifth step involves simultaneous loading tests and data acquisition. The mechanical loading device 5 and the hydraulic loading unit 4 are activated to apply a mechanical-hydraulic coupling effect to the target engineering coal and rock mass sample 1. During the test, the infrared thermal imager 6 collects infrared radiation information from the sample surface, the acoustic emission sensor 7 collects acoustic emission signals generated by crack propagation, the hydraulic sensor 8 records changes in internal water pressure, and the flow sensor 9 records changes in seepage flow. These monitoring signals are simultaneously transmitted to the data acquisition and processing unit 10. The sixth step involves multi-source data processing and disaster precursor identification. The data acquisition and processing unit 10 simultaneously processes the infrared thermal image data, acoustic emission data, and hydraulic parameters. Specifically, it identifies low-temperature anomaly areas based on the infrared thermal image data and analyzes their spatiotemporal evolution characteristics; combined with changes in acoustic emission events, water pressure response, and flow rate changes, it identifies precursors to the disaster evolution process of the engineering coal and rock mass. Figure 2Solid lines represent waterway connections used for water injection; dashed lines represent signal connections used for data transmission. The target engineering coal and rock mass samples used in this application are generally made from large-volume rocks found in nature. During processing, the rock should be cut perpendicular to its bedding planes, then sawn and ground to form a 100mm × 100mm × 100mm cube standard sample. A cylindrical cavity 2 with a diameter of 50mm and a depth of 50mm is drilled at the center of the sample surface using a drilling machine to leave space for water injection. Since the surface smoothness of the rock sample has a significant impact on its mechanical properties during the load-bearing process, to ensure the accuracy of the experimental monitoring data, the surface smoothness parameters are set as follows: the surface unevenness of the rock sample is less than 0.1mm, with no protrusions or depressions; the sides of the rock sample should be perpendicular to the upper and lower end faces, with a deviation of less than 0.05°. A square iron sheet measuring 80mm × 80mm with a 20mm diameter hole in the center is fixed to the opening of the rock sample using industrial adhesive. A 20mm diameter steel bend is then welded to the iron sheet, ensuring that the geometric centers of the cylindrical cavity inside the sample, the square iron sheet, and the steel water-filling bend coincide. After filling the cavity of the rock sample with water through the steel water-filling bend 3, the end of the steel bend is connected to an air compressor via a rubber conduit. Once the pressure is set to the target value, the air compressor is turned on, maintaining the water pressure in the cylindrical cavity 2 inside the sample near the target value. The pressure is adjusted in real-time by observing the pressure gauge on the air compressor. The hydraulic data includes water pressure data and / or flow rate data.

[0036] Step S302: Construct the relative temperature field of the infrared thermal image data based on a preset reference temperature; In this step, the infrared thermal imager continuously acquires thermal image sequences of the rock sample observation surface. First, the effective region is extracted from each frame of thermal image, retaining the sample observation surface area while removing the background, loading head, pressure platform, pipelines, edge obstructions, and other non-sample areas. The effective region is denoted as ROI. To reduce the impact of ambient temperature drift, infrared instrument zero drift, and background thermal reflection on the identification results, this embodiment uses relative temperature difference instead of absolute temperature for subsequent analysis, as shown in the mathematical expression below: ; in: For a moment t Next pixel ( x , y The relative temperature, usually expressed in °C; For a moment t Next pixel The infrared temperature; For a moment t The reference temperature can be taken as the average temperature of the reference area, the ambient reference temperature, or the average temperature during the initial stabilization phase of loading.x , y () represents the pixel coordinates within the effective region (ROI) of the infrared thermal image; after spatial calibration, it can also represent the actual position coordinates. t This refers to the time or the moment corresponding to the current infrared image frame. The purpose of using a relative temperature field is to minimize overall temperature drift and retain only the relative cold and hot differences within the sample observation surface. The identification of water inrush precursors focuses on low-temperature anomalies caused by localized seepage heat transfer or moisture migration, rather than slow changes in the overall laboratory temperature.

[0037] Step S303: Calculate and process the relative temperature distribution based on the relative temperature field to obtain the dynamic low temperature threshold; In this step, the current dynamic percentile threshold is determined based on the target percentile value of the relative temperature distribution of the effective region of the relative temperature field at the current moment; the mathematical expression of the current dynamic percentile threshold can be expressed by the following formula: ; in, For a moment t The dynamic percentile low temperature threshold, usually expressed in °C. The first value is calculated from all relative temperature values ​​within the ROI at the current time. p Percentile value. p This is a percentile parameter, with a given range of 20% to 35%. For a moment t A statistical sample consisting of the relative temperatures of all pixels within the ROI. Based on the median of the temperature field, the median absolute deviation, and a predetermined deviation coefficient, the current statistical deviation low-temperature threshold is calculated; the mathematical expression for the current statistical deviation low-temperature threshold is as follows: ; in, For a moment t The current statistical deviation low temperature threshold, usually in °C. This represents the median of the relative temperature distribution of the ROI at the current moment. This represents the median absolute deviation of the ROI relative to the temperature distribution at the current moment, used to characterize the robustness of the temperature field's dispersion. k The coefficient of variation ranges from 1 to 3. k The larger the value, the more stringent the low-temperature point determination. First, extract the temperature distribution set within the current valid ROI. Specifically, assume a time... t The corresponding effective region of interest (ROI) of the current infrared image contains M valid pixels. Therefore, the relative temperatures of all pixels at the current moment can be grouped into a set: ; in, This is not the temperature of a single point, but a statistical sample of the temperature field within the current valid Region of Interest (ROI). For example, if there are 40,000 valid pixels within the ROI, then... It contains 40,000 relative temperature values. This represents the relative temperature corresponding to 1-M valid pixels; subsequent percentiles, median, and MAD are calculated based on this set. The median of the temperature field is determined based on the relative temperature distribution of the valid region of the relative temperature field at the current time. and median absolute deviation Current statistical deviation low temperature threshold The mathematical expression for can also be represented by the following formula: ; in, This represents the median of all relative temperature values ​​within the current ROI. The median absolute deviation of the relative temperature distribution is calculated as follows: ; in, This represents the median absolute deviation of the relative temperature distribution. For the first i Each effective pixel at time [time] t The relative temperature. This is to take the median of the data corresponding to all pixels.

[0038] For example, the following is an example of a set of relative temperatures: ; The mathematical expression for calculating the median is as follows: ℃; When the coefficient of variation is 2, the mathematical formula for calculating the current statistical deviation low temperature threshold is as follows: ℃; Therefore, according to the current statistical deviation low-temperature threshold, when the relative temperature corresponding to an effective pixel is less than or equal to -0.265℃, it is considered to have a significant low-temperature anomaly. Compared to ℃, More stringent.

[0039] The current dynamic percentile threshold and the current statistical deviation low temperature threshold are filtered, and the minimum value between the current dynamic percentile threshold and the current statistical deviation low temperature threshold is determined as the dynamic low temperature threshold. The dynamic low temperature threshold... The mathematical expression is as follows: ; The criteria for determining low temperature anomalies are usually as follows: ,so The smaller the value, the stricter the judgment. The meaning is that the candidate low temperature point must not only fall into the low temperature percentile region of the current temperature field, but also have a sufficiently significant low temperature deviation relative to the current median temperature.

[0040] Logically speaking, the dual criteria for candidate low-temperature points can be represented by the following formula: ; in, For a moment t Next pixel ( x , y The relative temperature of ). Current dynamic low temperature threshold. For a moment t Next pixel ( x , y The low-temperature deviation relative to the corresponding pixel temperature or the average temperature of the corresponding area in the reference stage. The minimum low temperature deviation threshold is a positive value; the formula uses... This means the current pixel must be at least [temperature lower than the reference temperature]. The minimum low temperature deviation threshold satisfies the following minimum low temperature deviation threshold constraint condition: ; in, NETD is the minimum low temperature deviation threshold. NETD is the infrared thermal imager noise equivalent temperature difference, used to characterize the minimum temperature difference level that the instrument can resolve. This represents the standard deviation of pixel temperature fluctuation during the initial reference stage of loading. 2 is a safety factor used to ensure that the effective low temperature deviation exceeds the instrument noise or random fluctuations of the reference temperature. The minimum low temperature deviation threshold is the larger of twice the equivalent temperature difference of the instrument noise and twice the standard deviation of the temperature fluctuation during the reference stage.

[0041] Step S304: Obtain the average relative temperature field during the hydraulic coupling loading reference stage and determine the reference low temperature deviation; In this step, a relatively stable initial stage of hydraulic coupling loading, characterized by a stable temperature field, acoustic emission activity, and water pressure and / or flow rate, is selected as the baseline stage. Infrared thermal images within this baseline stage are numbered chronologically, and the baseline stage is defined as containing [number of stages]. Frame infrared thermal image.

[0042] For any pixel point within the effective region ROI ( x , yThe reference relative temperature of the pixel is calculated based on the relative temperature of each frame within the reference phase, and the low temperature deviation of the pixel relative to the reference phase at the current moment is further determined. The mathematical expression for the reference low temperature deviation is as follows: ; in, For a moment t Next pixel ( x , y The low-temperature deviation relative to the corresponding pixel temperature or the average temperature of the corresponding region in the reference stage; Δ T ( x , y , t (time) t Next pixel ( x , y The relative temperature of Δ T ( x , y , j (This refers to the baseline phase time) j The corresponding pixels in the infrared thermal image ( x , y The relative temperature; The number of infrared thermal images included in the baseline phase; j The sampling time corresponds to the infrared thermal image within the baseline phase.

[0043] when "Time" indicates the current moment. t Next pixel ( x , y The low-temperature deviation relative to the corresponding pixel temperature or the average temperature of the corresponding region in the baseline stage is less than zero. The smaller the value, the more significant the temperature drop of that pixel relative to the baseline stage.

[0044] when Time indicates the current moment. t Next pixel ( x , y The low-temperature deviation relative to the baseline pixel temperature or the average temperature of the corresponding region is not less than the minimum low-temperature deviation threshold. The opposite number; used in the formula This means the current pixel must be at least [temperature lower than the reference temperature]. .

[0045] If it is not possible to perform a one-to-one correspondence calculation for each pixel in the reference stage, the average relative temperature of the region corresponding to the pixel in the reference stage can be used to replace the average relative temperature of the corresponding pixel in order to calculate the reference low temperature deviation.

[0046] By setting the baseline low temperature deviation, the problem of generating candidate low temperature points according to the preset percentile ratio during the temperature field stabilization stage can be avoided when only the dynamic percentile threshold is used, and stable low temperature regions caused by the inherent temperature non-uniformity of the sample surface can be eliminated.

[0047] Step S305: Use the dynamic low temperature threshold and the baseline low temperature deviation to extract low temperature features from the infrared thermal image data to obtain infrared low temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation; In the specific implementation process of this step, the infrared thermal imaging data is screened based on the dynamic low temperature threshold and the baseline low temperature deviation to obtain candidate low temperature points; For any pixel point within the effective region ROI ( x , y When the current relative temperature of a pixel is not higher than the dynamic low-temperature threshold, and the low-temperature deviation of the pixel relative to the baseline stage is not higher than the negative value of the minimum low-temperature deviation threshold, the pixel is determined as a candidate low-temperature point. The binary determination function for the candidate low-temperature point is expressed as: ; in, B ( x , y , t (time) t Next pixel ( x , y Is it a binary variable representing a candidate low temperature point? B ( x , y , t When )=1, it indicates that the pixel simultaneously meets the dynamic low temperature threshold condition and the baseline low temperature deviation condition, and is determined to be a candidate low temperature point; when B ( x , y , t When )=0, it means that the pixel does not meet both of the above conditions at the same time and is not judged as a candidate low temperature point; θ ( t (time) t The dynamic low temperature threshold; ( x , y , t (time) t Next pixel ( x , y The low-temperature deviation relative to the corresponding pixel temperature or the average temperature of the corresponding region in the baseline stage; The minimum low-temperature deviation threshold is a positive value. The dynamic low-temperature threshold condition is used to determine whether the current pixel is located in the low-temperature portion of the relative temperature distribution of the effective region of interest (ROI) at the current time. The baseline low-temperature deviation condition is used to determine whether the temperature of the current pixel has decreased sufficiently significantly relative to its baseline stage. Only pixels that simultaneously meet both of the above conditions are identified as candidate low-temperature points.

[0048] To eliminate random thermal noise and instantaneous low-temperature fluctuations, candidate low-temperature states of the same pixel across multiple consecutive frames of infrared thermal images are statistically analyzed. (Based on time...) t The corresponding infrared thermal image serves as the endpoint, and a length of [length missing] is established. N A sliding frame window, wherein the continuous frame filtering satisfies: ; in, For pixels ( x , y )exist j Whether the time is a candidate low temperature point is a binary variable. N To continuously examine the number of frames, The minimum number of occurrences. t - N +1 to t For time t The length of the endpoint is N A sliding frame window. Preferably, N 3 to 15 frames Not less than N 60% of the candidate low-temperature points were identified as persistent low-temperature points. Spatial connectivity analysis was performed on these persistent low-temperature points to obtain valid low-temperature points; subsequently, connectivity analysis was performed on the persistent low-temperature points to eliminate those with areas smaller than [a certain value]. In the designated area, only the effective low temperature points are retained. The ROI area can be set to 0.05%–1%, or 5–50 pixels depending on the spatial resolution of the infrared thermal image. Through the above processing, random thermal noise, transient cold spots, and edge artifacts are less likely to simultaneously satisfy the four conditions of dynamic threshold, baseline deviation, continuous occurrence, and spatial connectivity, thus significantly reducing false positives. Neighborhood analysis is performed based on the effective low-temperature points to obtain low-temperature clusters; adjacent effective low-temperature points are grouped into low-temperature clusters. Let time... t There are a total of K ( t ) low-temperature cluster, the first k The area of ​​each low-temperature cluster is The effective area is The total number of effective pixels is M The total number of effective low temperature points is m ( tFeature extraction is performed on the low-temperature clusters to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. Infrared radiation variance feature parameters are calculated based on the total number of effective pixels within the effective region of interest (ROI), the relative temperature corresponding to each effective pixel, and the average relative temperature of the ROI. IRV The mathematical expression is as follows: ; in, For a moment t The average relative temperature of the lower effective region ROI. For a moment t The infrared radiation variance characteristic parameter characterizes the infrared radiation variance and is used to characterize the spatial dispersion of the temperature field of the ROI. The mathematical expression is as follows: ; in, M This represents the total number of valid pixels within the ROI. For the first i Each effective pixel at time [time] t The relative temperature. i These are pixel labels.

[0049] The low-temperature point area ratio feature parameter is obtained by calculating the number of effective low-temperature points at different times and the total number of effective pixels within the effective region ROI. LTPSF The mathematical expression is as follows: ; in, For a moment t The effective low-temperature point area ratio characteristic parameter is dimensionless. For a moment t The number of valid low-temperature points after filtering by dynamic threshold, baseline deviation, continuous frames, and connected components. M This represents the total number of valid pixels within the ROI.

[0050] The maximum low-temperature cluster area ratio characteristic parameter is obtained by calculating based on the area of ​​each low-temperature cluster and the area of ​​the effective region. LTAF The mathematical expression is as follows: ; in, For a moment t The maximum low-temperature cluster area ratio characteristic parameter is dimensionless. For a moment t Next k The area of ​​each low-temperature cluster. k This refers to the low-temperature cluster number; and the coefficient of variation in the statistical deviation formula. kThey belong to different contexts. This represents the maximum area among all low-temperature clusters at the current moment. The total area of ​​the effective region ROI can be represented by pixel area or actual area after calibration.

[0051] The characteristic parameters of the maximum cryogenic cluster centroid migration amount are calculated based on the maximum cryogenic cluster centroid coordinates at different times and the predetermined time window length. LTCM ( t The mathematical expression is as follows: ; in, , For a moment t The x and y coordinates of the centroid of the largest low-temperature cluster. This represents the number of valid pixels contained in the largest low-temperature cluster. , The largest low-temperature cluster i The horizontal and vertical coordinates of each pixel. i This is the pixel number inside the largest low-temperature cluster. The coordinates of the centroid of the maximum low-temperature cluster at the corresponding moment of the previous time window. The time window length is specified. The infrared cryogenic cluster characteristic parameters include the infrared radiation variance characteristic parameter, the cryogenic point area ratio characteristic parameter, the maximum cryogenic cluster area ratio characteristic parameter, the maximum cryogenic cluster centroid migration characteristic parameter, the dispersion or aggregation state characteristic parameter, and the cryogenic cluster boundary connectivity index.

[0052] Based on the number of low-temperature clusters at different times, the dispersion or aggregation characteristics of the low-temperature anomaly region at different times are determined. LTCN The mathematical expression is as follows: ; in, For a moment t The total number of valid low-temperature clusters retained after connectivity component filtering. Define the low-temperature cluster boundary connectivity index. LTCI When the minimum distance between the maximum low-temperature cluster and the sample boundary, orifice, projected area of ​​the cylindrical cavity inside the sample, pre-fabricated crack, or seepage outlet is less than a threshold, or when the connected region has already touched the above-mentioned objects, it is determined that... LTCI An abnormal state has been reached. Low-temperature cluster boundary connectivity index. LTCI The connectivity criterion for the boundary of a low-temperature cluster can be expressed by the following formula: ; in, LTCI For a moment tThe low-temperature cluster boundary connectivity index; a value of 1 indicates that the connectivity anomaly has been reached, and a value of 0 indicates that it has not been reached. For a moment t The spatial region occupied by the largest low-temperature cluster. The target boundary area includes the sample boundary, orifice, cavity projection area, pre-fabricated cracks or seepage outlets, etc. It represents the minimum distance from the boundary of the maximum low-temperature cluster to the target boundary region. For the preset distance threshold, the corresponding distance threshold To avoid confusion with the actual minimum distance, this is denoted as […]. .

[0053] Step S306: Perform feature extraction on the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters; In this step, feature extraction is performed on the acoustic emission data to obtain the acoustic emission event rate and the acoustic emission energy growth rate; the mathematical expression for the acoustic emission event rate is as follows: ; in, For a moment t The acoustic emission event rate within the corresponding time window, usually expressed as times per second. N AE ( t Δ t , t ) represents the time interval [ t -Δ t , t The total number of acoustic emission events recorded within the [database]. This represents the length of the time window.

[0054] The mathematical expression for the acoustic emission energy growth rate is as follows: ; in, Let be the acoustic emission energy growth rate at time t. Deadline t The cumulative acoustic emission energy. This represents the cumulative acoustic emission energy at the corresponding moment of the previous time window. This represents the length of the time window.

[0055] Feature extraction was performed on the hydraulic data to obtain the rate of change of water pressure and the rate of change of flow rate; the mathematical expression for the rate of change of water pressure is as follows: ; in, For a moment tThe rate of change of water pressure is measured in units that depend on the units of water pressure and time, such as MPa / s. For a moment t Water pressure. This represents the water pressure at the corresponding moment of the previous time window. This represents the length of the time window.

[0056] The mathematical expression for the rate of change of flow is as follows: ; in, For a moment t The magnitude of the rate of change of flow. For a moment t The water flow rate or volumetric flow rate. This represents the water flow rate at the corresponding moment of the previous time window. The time window length is defined as follows. The acoustic emission characteristic parameters include the acoustic emission event rate and the acoustic emission energy growth rate; the hydraulic characteristic parameters include the water pressure change rate and / or the flow rate change rate. The acoustic emission event rate and acoustic emission energy growth rate reflect the frequency of internal crack activity and the intensity of energy release, respectively; the water pressure change rate and / or flow rate change rate reflect changes in seepage state, respectively. If the expansion of the low-temperature cluster and the enhancement of acoustic emission activity, as well as water pressure or flow anomalies, occur synchronously within a unified time window, the reliability of the precursor to water inrush is significantly improved.

[0057] Step S307: Select a stage in which the initial temperature field, acoustic emission activity, and water pressure and / or flow rate are relatively stable as the baseline stage; In this step, during implementation, for continuous characteristic parameters such as infrared radiation variance characteristic parameters, low-temperature point area ratio characteristic parameters, maximum low-temperature cluster area ratio characteristic parameters, maximum low-temperature cluster centroid migration characteristic parameters, acoustic emission event rate, acoustic emission energy growth rate, water pressure change rate, and flow rate change rate, the baseline mean and standard deviation are calculated within the baseline period. For characteristic parameters of dispersed or aggregated states... LTCN ( t ), calculate its baseline mean during the baseline period. For the connectivity index of low-temperature cluster boundaries LTCI ( t (The baseline mean and standard deviation are not calculated.)

[0058] Step S308: Based on the data type of the feature parameters, anomaly judgments are made using the continuous feature parameter anomaly criterion, the low-temperature cluster quantity anomaly criterion, and the low-temperature cluster boundary connectivity anomaly criterion, respectively. In the specific implementation process of this step, for continuous characteristic parameters such as infrared radiation variance characteristic parameters, low-temperature point area ratio characteristic parameters, maximum low-temperature cluster area ratio characteristic parameters, maximum low-temperature cluster centroid migration characteristic parameters, acoustic emission event rate, acoustic emission energy growth rate, water pressure change rate, and flow rate change rate, an anomaly judgment function based on the benchmark mean and standard deviation is used for judgment. The mathematical expression is as follows: ; When the anomaly detection function is satisfied, the feature parameter is determined to be abnormal. For any continuous monitoring characteristic parameter at time t, we can take... IRV , LTPSF , LTAF , LTCM , R AE , G AE G P or G Q . For characteristic parameters X The mean value during the initial stable baseline phase of loading. Let X be the standard deviation of the characteristic parameter X during the initial stable baseline stage of loading. This is the abnormal threshold coefficient, and its value can range from 2 to 3.

[0059] For characteristic parameters of dispersed or aggregated states LTCN ( t Anomaly detection is performed based on the degree of deviation of the current number of low-temperature clusters from the average number of low-temperature clusters in the baseline period. The mathematical expression is as follows: ; When the above conditions are met, the characteristic parameters for determining the dispersion or aggregation state are determined. LTCN ( t An anomaly occurred. Among them, This represents the baseline mean of the characteristic parameters of the dispersed or clustered state during the baseline phase. This is the minimum deviation threshold for the number of low-temperature clusters, and it is a positive value, preferably not less than 1.

[0060] when When this occurs, it indicates an increase in the number of effective low-temperature clusters, and the low-temperature anomaly area may exhibit a multi-point dispersed state; when LTCN ( t )< When this occurs, it indicates a decrease in the number of effective low-temperature clusters. Whether this corresponds to cluster aggregation or merging should be determined in conjunction with the characteristic parameter of the maximum low-temperature cluster area ratio. LTAF ( t Judge based on the changes in ).

[0061] For the low-temperature cluster boundary connectivity index LTCI ( t Since it is a binary state characteristic parameter, the benchmark mean and standard deviation are not used for anomaly detection. When the low-temperature cluster boundary connectivity index reaches an abnormal state, it is determined that the low-temperature cluster boundary connectivity index has reached an abnormal state; when At that time, it was determined that the connectivity index of the low-temperature cluster boundary did not reach an abnormal state.

[0062] Step S309: Based on the anomaly judgment results corresponding to different feature parameters, identify the catastrophic precursors of the target engineering coal and rock mass sample, and obtain the identification results.

[0063] In the specific implementation process of this step, the anomaly judgment results of each characteristic parameter obtained in step S308 are acquired. Among them, the anomalies of the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, the maximum low-temperature cluster centroid migration characteristic parameter, the acoustic emission event rate, the acoustic emission energy growth rate, the water pressure change rate, and the flow rate change rate are determined according to the continuous characteristic parameter anomaly judgment function; the dispersion or aggregation state characteristic parameters... LTCN The abnormal state is determined based on its deviation from the average number of cryogenic clusters in the baseline stage; when the cryogenic cluster boundary connectivity index LTCI ( t When )=1, it is determined that the boundary connectivity index of the low-temperature cluster has reached an abnormal state.

[0064] In the specific implementation of this step, when the anomaly judgment result shows that any two of the following characteristic parameters are abnormal: the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, and the dispersion or aggregation state characteristic parameter, and the anomaly duration is greater than or equal to the length of the first preset time window, it is determined that the current state is in a level one warning state, and a level one warning is issued; that is... IRV , LTPSF , LTAF , LTCN If any two of the following conditions are abnormal and persist for the length of a first preset time window: Local seepage anomaly is initiated, and a stable low-temperature anomaly appears on the surface. A Level 1 warning is issued. If the anomaly assessment result indicates that any three of the following three characteristic parameters are abnormal (among the infrared low-temperature cluster characteristic parameters: the low-temperature point area ratio, the maximum low-temperature cluster area ratio, and the maximum low-temperature cluster centroid migration) and the following two acoustic emission characteristic parameters (among the acoustic emission event rate and acoustic emission energy growth rate), and the anomaly duration is greater than or equal to the length of a second preset time window, a Level 2 warning is determined, and a Level 2 warning is issued; that is... LTPSF , LTAF , LTCM , R AE ,G AE Any three anomalies that persist for a second preset time window duration indicate that the low-temperature anomaly is expanding from a point-like pattern to a cluster or band-like pattern, with enhanced internal crack activity, triggering a level-two warning. When the low-temperature cluster boundary connectivity index... LTCI ( t A value of 1 indicates that the maximum cryogenic cluster has spatially overlapped with the target boundary region, or that the minimum distance between the boundary of the maximum cryogenic cluster and the target boundary region is not greater than a preset distance threshold; simultaneously, the acoustic emission event rate... R AE ( t (and acoustic emission energy growth rate) G AE ( t At least one of the following is abnormal: water pressure change rate G P ( t ) and flow rate G Q ( t At least one of the following is abnormal, and the infrared radiation variance characteristic parameter IRV ( t Low temperature point area ratio characteristic parameters LTPSF ( t ), Maximum low-temperature cluster area ratio characteristic parameter LTAF ( t ), characteristic parameters of maximum low-temperature cluster centroid migration LTCM ( t and characteristic parameters of dispersion or aggregation state. LTCN ( t If at least one of the following conditions is abnormal, and the above-mentioned abnormality continues synchronously within the third preset time window, it is determined that the current state is in a level three warning state, and a level three warning is issued.

[0065] The three-level warning indicates that the maximum low-temperature cluster has approached or connected to the sample boundary, pore, cylindrical cavity projection area, pre-fabricated crack, or seepage outlet, accompanied by enhanced internal crack activity and abnormal water pressure or flow conditions, suggesting that potential crack-seepage channels may be nearing completion or have already formed, and the rock sample has entered a high-risk stage for water inrush. To facilitate continuous characterization of the water inrush risk level of rock samples, a comprehensive warning index can be constructed in addition to the graded warning rules. FWI ( t This index does not replace the aforementioned tiered early warning criteria. Instead, it performs dimensionless weighted fusion of extracted infrared cryogenic cluster features, acoustic emission features, and hydraulic response features to output a continuous risk curve. The mathematical expression for the comprehensive early warning index is shown below: ; in: For a moment tThe comprehensive flood warning index is dimensionless. These are the standardized outliers of the infrared radiation variance. This represents the standardized outlier of the area ratio at low temperatures. This represents the standardized outlier of the maximum low-temperature cluster area ratio. The standardized outliers of the acoustic emission event rate. These are the standardized outliers of the rate of change of water pressure. These are standardized outliers representing the rate of change in flow. For characteristic parameters X At any moment t The standardized outliers are dimensionless. The mathematical expression is as follows: ; in, and These are the mean and standard deviation of the baseline period, respectively. - Let be the weighting coefficient, and satisfy: ; in, Indicates the first i The weight coefficients corresponding to each feature parameter. These weights can be determined based on experience, experimental samples, entropy weighting, analytic hierarchy process (AHP), or data training methods. When FWI ( t When the risk level continues to rise and exceeds the preset risk threshold, it can be used as an auxiliary basis for judging the results of Level 1, Level 2 or Level 3 early warnings.

[0066] This application employs a joint determination method combining dynamic low-temperature threshold and baseline low-temperature deviation to avoid the problem of a fixed proportion of candidate low-temperature points naturally generated during the stable phase using a simple percentile threshold. This invention uses continuous frame persistence filtering and connected component area filtering to eliminate transient cold spots, isolated noise points, and edge artifacts, improving the stability of low-temperature anomaly identification. This invention organizes low-temperature points into low-temperature clusters and extracts... LTAF , LTCM , LTCI These indicators can characterize the evolution of low-temperature anomalies from point-like appearances to clustered aggregations, directional expansion, and boundary connectivity. This invention couples the evolution of infrared low-temperature clusters with acoustic emission, water pressure, and flow parameters within a unified time window, enabling the identification of flood precursors to no longer rely on a single temperature value or empirical observation. This invention retains existing... IRV and LTPSF The monitoring foundation is enhanced by dynamic thresholds, continuous screening, and low-temperature cluster parameters to improve its noise resistance and generalization ability.

[0067] Another embodiment of this application provides an engineering coal and rock mass disaster precursor identification device 400, such as... Figure 4 As shown, it includes: The acquisition module 401 is used to acquire infrared thermal image data, acoustic emission data and hydraulic data during the hydraulic coupling loading process of the target engineering coal and rock mass sample using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. The first feature extraction module 402 is used to extract low-temperature features from the infrared thermal image data using a feature extraction method that combines dynamic low-temperature threshold and benchmark low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. The second feature extraction module 403 is used to extract features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters. The identification module 404 is used to identify the catastrophic precursors of the target engineering coal and rock mass sample based on the infrared low-temperature cluster feature parameters, the acoustic emission feature parameters, and the hydraulic feature parameters, and to obtain the identification results.

[0068] In specific implementation, the first feature extraction module 402 is specifically used to construct the relative temperature field of the infrared thermal image data based on a preset reference temperature; perform calculations based on the relative temperature distribution of the relative temperature field to obtain a dynamic low-temperature threshold; obtain the average relative temperature field of the hydraulic coupling loading reference stage, and determine the reference low-temperature deviation. Using the dynamic low-temperature threshold and the reference low-temperature deviation, low-temperature features are extracted from the infrared thermal image data to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

[0069] In the specific implementation process, the first feature extraction module 402 is further used to determine the current dynamic percentile threshold based on the target percentile value of the relative temperature distribution of the effective region of the relative temperature field at the current time; determine the median and median absolute deviation of the temperature field based on the relative temperature distribution of the effective region of the relative temperature field at the current time; calculate the current statistical deviation low temperature threshold based on the median of the temperature field, the median absolute deviation, and a predetermined deviation coefficient; filter the current dynamic percentile threshold and the current statistical deviation low temperature threshold, and determine the minimum value of the current dynamic percentile threshold and the current statistical deviation low temperature threshold as the dynamic low temperature threshold.

[0070] In the specific implementation process, the first feature extraction module 402 is also used to screen the infrared thermal image data based on the dynamic low temperature threshold and the benchmark low temperature deviation to obtain candidate low temperature points; to count the number of candidate low temperature points at different times at the same location of the target engineering coal and rock mass sample to obtain the continuous low temperature points of the target engineering coal and rock mass sample; to perform spatial connectivity analysis based on the continuous low temperature points to obtain effective low temperature points; to perform neighborhood analysis based on the effective low temperature points to obtain low temperature clusters; and to extract features from the low temperature clusters to obtain infrared low temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

[0071] In the specific implementation process, the first feature extraction module 402 is also used to calculate the infrared radiation variance feature parameters based on the total number of effective pixels in the effective area, the relative temperature corresponding to each effective pixel, and the average relative temperature of the effective area; calculate the low-temperature point area ratio feature parameters based on the number of effective low-temperature points at different times and the total number of effective pixels in the effective area; calculate the maximum low-temperature cluster area ratio feature parameters based on the area of ​​each low-temperature cluster and the area of ​​the effective area; calculate the maximum low-temperature cluster centroid migration feature parameters based on the maximum low-temperature cluster centroid coordinates and the predetermined time window length at different times; determine the dispersion or aggregation state feature parameters of the low-temperature anomaly area at different times based on the number of low-temperature clusters at different times; and determine the low-temperature cluster boundary connectivity index based on the spatial overlap relationship and / or minimum distance between the maximum low-temperature cluster and the target boundary area; wherein, the infrared low-temperature cluster feature parameters include the infrared radiation variance feature parameters, the low-temperature point area ratio feature parameters, the maximum low-temperature cluster area ratio feature parameters, the maximum low-temperature cluster centroid migration feature parameters, the dispersion or aggregation state feature parameters, and the low-temperature cluster boundary connectivity index.

[0072] In specific implementation, the second feature extraction module 403 is specifically used to extract features from the acoustic emission data to obtain the acoustic emission event rate and the acoustic emission energy growth rate; and to extract features from the hydraulic data to obtain the water pressure change rate and / or the flow rate change rate; wherein, the acoustic emission feature parameters include the acoustic emission event rate and the acoustic emission energy growth rate; and the hydraulic feature parameters include the water pressure change rate and / or the flow rate change rate.

[0073] In the specific implementation process, the identification module 404 is specifically used to calculate the baseline mean and standard deviation of the continuous characteristic parameters, acoustic emission characteristic parameters, and hydraulic characteristic parameters in the infrared low-temperature cluster characteristic parameters, respectively, and to use an anomaly judgment function based on the baseline mean and standard deviation to judge the anomalies of the continuous characteristic parameters, acoustic emission characteristic parameters, and hydraulic characteristic parameters; to judge the anomalies of the dispersed or aggregated state characteristic parameters based on the degree of deviation of the dispersed or aggregated state characteristic parameters from the average number of low-temperature clusters in the baseline stage; to judge the anomalies of the low-temperature cluster boundary connectivity index based on the current value of the low-temperature cluster boundary connectivity index, and to determine that it has reached an abnormal state when the low-temperature cluster boundary connectivity index is equal to 1; and to identify the catastrophic precursors of the target engineering coal and rock mass sample based on the anomaly judgment results corresponding to different characteristic parameters, and to obtain the judgment results; wherein, the continuous characteristic parameters include the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, and the maximum low-temperature cluster centroid migration characteristic parameter.

[0074] In specific implementation, the identification module 404 is also used to determine that the current state is under a Level 1 warning and to issue a Level 1 warning when the anomaly judgment result is that any two of the following features among the infrared low-temperature cluster feature parameters are abnormal and the duration of the anomaly is greater than or equal to the length of the first preset time window: the infrared radiation variance feature parameter, the low-temperature point area ratio feature parameter, the maximum low-temperature cluster area ratio feature parameter, and the dispersion or aggregation state feature parameter. When the anomaly judgment result is that the low-temperature point area ratio feature parameter, the maximum low-temperature cluster area ratio feature parameter, the maximum low-temperature cluster centroid migration feature parameter, and the acoustic emission event rate and acoustic emission energy increase feature parameters among the infrared low-temperature cluster feature parameters are abnormal, the module 404 is also used to determine that the current state is under a Level 1 warning and to issue a Level 1 warning. If any three characteristic parameters in the growth rate are abnormal and the abnormal duration is greater than or equal to the length of the second preset time window, the current state is determined to be in a level two warning state, and a level two warning is issued. When the low-temperature cluster boundary connectivity index reaches an abnormal state, at least one characteristic parameter of the acoustic emission event rate and the acoustic emission energy growth rate is abnormal, the collected hydraulic characteristic parameters are abnormal, and at least one characteristic parameter of the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, the maximum low-temperature cluster centroid migration characteristic parameter, and the dispersion or aggregation state characteristic parameter is abnormal and continues synchronously within the third preset time window, the current state is determined to be in a level three warning state, and a level three warning is issued.

[0075] This application employs a joint determination method combining dynamic low-temperature threshold and baseline low-temperature deviation to avoid the problem of a fixed proportion of candidate low-temperature points naturally generated during the stable phase using a simple percentile threshold. This invention uses continuous frame persistence filtering and connected component area filtering to eliminate transient cold spots, isolated noise points, and edge artifacts, improving the stability of low-temperature anomaly identification. This invention organizes low-temperature points into low-temperature clusters and extracts... LTAF, LTCM, LTCI These indicators can characterize the evolution of low-temperature anomalies from point-like appearances to clustered aggregations, directional expansion, and boundary connectivity. This invention couples the evolution of infrared low-temperature clusters with acoustic emission, water pressure, and flow parameters within a unified time window, enabling the identification of flood precursors to no longer rely on a single temperature value or empirical observation. This invention retains existing... IRV and LTPSF The monitoring foundation is enhanced by dynamic thresholds, continuous screening, and low-temperature cluster parameters to improve its noise resistance and generalization ability.

[0076] Another embodiment of this application provides an electronic device, which can be a server. The electronic device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the processor executes the program of the electronic device, it implements the functions or steps of a server-side method for identifying precursors of coal and rock mass disasters in engineering.

[0077] In one embodiment, an electronic device is provided, which can be a client. The electronic device includes a processor, memory, a network interface, a display screen, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When the program of the electronic device is executed by the processor, it implements the functions or steps of a client-side method for identifying precursors of coal and rock mass disasters in engineering.

[0078] Another embodiment of this application provides an electronic device, including at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, performs the following method steps: Step 1: Collect infrared thermographic data, acoustic emission data, and hydraulic data of the target engineering coal and rock mass sample during the hydraulic coupling loading process using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. Step 2: Use the feature extraction method that combines dynamic low temperature threshold and baseline low temperature deviation to extract low temperature features from the infrared thermal image data to obtain infrared low temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. Step 3: Extract features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters; Step 4: Based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters, identify the catastrophic precursors of the target engineering coal and rock mass sample, and obtain the identification results.

[0079] The specific implementation process of the above method steps can be found in the embodiment of the above arbitrary engineering coal and rock mass disaster precursor identification method, which will not be repeated here.

[0080] This application employs a joint determination method combining dynamic low-temperature threshold and baseline low-temperature deviation to avoid the problem of a fixed proportion of candidate low-temperature points naturally generated during the stable phase using a simple percentile threshold. This invention uses continuous frame persistence filtering and connected component area filtering to eliminate transient cold spots, isolated noise points, and edge artifacts, improving the stability of low-temperature anomaly identification. This invention organizes low-temperature points into low-temperature clusters and extracts... LTAF , LTCM , LTCI These indicators can characterize the evolution of low-temperature anomalies from point-like appearances to clustered aggregations, directional expansion, and boundary connectivity. This invention couples the evolution of infrared low-temperature clusters with acoustic emission, water pressure, and flow parameters within a unified time window, enabling the identification of flood precursors to no longer rely on a single temperature value or empirical observation. This invention retains existing... IRV and LTPSF The monitoring foundation is enhanced by dynamic thresholds, continuous screening, and low-temperature cluster parameters to improve its noise resistance and generalization ability.

[0081] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. Those skilled in the art can make various modifications or equivalent substitutions to this application within the scope and nature of this application, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for identifying precursors of coal and rock mass disasters in engineering projects, characterized in that, include: Infrared thermographic data, acoustic emission data, and hydraulic data are collected during the hydraulic coupling loading process on the target engineering coal and rock mass sample using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. The infrared thermal image data is subjected to low-temperature feature extraction using a feature extraction method that combines dynamic low-temperature threshold and baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. Feature extraction is performed on the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters; Based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters, the catastrophic precursors of the target engineering coal and rock mass sample are identified, and the identification results are obtained.

2. The method as described in claim 1, characterized in that, The feature extraction method, which utilizes a dynamic low-temperature threshold and a benchmark low-temperature deviation to jointly determine the low-temperature features of the infrared thermal image data, yields infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. Specifically, these parameters include: The relative temperature field of the infrared thermal image data is constructed based on a preset reference temperature; The dynamic low temperature threshold is obtained by performing calculations based on the relative temperature distribution of the relative temperature field. Obtain the average relative temperature field during the hydraulic coupling loading reference stage and determine the reference low temperature deviation; The infrared thermal image data is used to extract low-temperature features using the dynamic low-temperature threshold and the baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

3. The method as described in claim 2, characterized in that, The calculation and processing based on the relative temperature distribution of the relative temperature field to obtain the dynamic low temperature threshold specifically includes: The current dynamic percentile threshold is determined based on the target percentile value of the relative temperature distribution of the effective region of the relative temperature field at the current moment; The median and median absolute deviation of the temperature field are determined based on the relative temperature distribution of the effective region of the relative temperature field at the current moment. The current statistical deviation low temperature threshold is calculated based on the median of the temperature field, the median absolute deviation, and the predetermined deviation coefficient. The current dynamic percentile threshold and the current statistical deviation low temperature threshold are filtered, and the minimum value between the current dynamic percentile threshold and the current statistical deviation low temperature threshold is determined as the dynamic low temperature threshold.

4. The method as described in claim 3, characterized in that, The process of extracting low-temperature features from the infrared thermal image data using the dynamic low-temperature threshold and the baseline low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation specifically includes: Candidate low-temperature points are obtained by filtering the infrared thermal imaging data based on the dynamic low-temperature threshold and the baseline low-temperature deviation. The number of candidate low-temperature points at different times at the same location of the target engineering coal and rock mass sample is counted to obtain the continuous low-temperature points of the target engineering coal and rock mass sample. Based on the sustained low temperature points, spatial connectivity analysis is performed to obtain effective low temperature points; Based on the effective low temperature points, a neighborhood analysis is performed to obtain low temperature clusters; Feature extraction is performed on the low-temperature clusters to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation.

5. The method as described in claim 4, characterized in that, The step of extracting features from the low-temperature clusters to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation specifically includes: The infrared radiation variance characteristic parameters are calculated based on the total number of effective pixels in the effective area, the relative temperature of each effective pixel, and the average relative temperature of the effective area. The area ratio characteristic parameter of low temperature points is obtained by calculating the number of effective low temperature points at different times and the total number of effective pixels in the effective area. Based on the area of ​​each low-temperature cluster and the area of ​​the effective region, the characteristic parameter of the maximum low-temperature cluster area ratio is obtained. Based on the maximum low-temperature cluster centroid coordinates at different times and the predetermined time window length, the characteristic parameters of the maximum low-temperature cluster centroid migration amount are obtained. The dispersion or aggregation characteristics of the low-temperature anomaly region at different times are determined based on the number of low-temperature clusters at different times. Based on the spatial overlap relationship and / or minimum distance between the maximum low-temperature cluster and the target boundary region, the connectivity index of the low-temperature cluster boundary is determined; The infrared low-temperature cluster characteristic parameters include the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, the maximum low-temperature cluster centroid migration characteristic parameter, the dispersion or aggregation state characteristic parameter, and the low-temperature cluster boundary connectivity index.

6. The method as described in claim 1, characterized in that, The step of extracting features from the acoustic emission data and the hydraulic data to obtain acoustic emission feature parameters and hydraulic feature parameters specifically includes: Feature extraction is performed on the acoustic emission data to obtain the acoustic emission event rate and the acoustic emission energy growth rate. Feature extraction is performed on the hydraulic data to obtain the water pressure change rate and / or flow rate change rate; The acoustic emission characteristic parameters include the acoustic emission event rate and the acoustic emission energy growth rate. The hydraulic characteristic parameters include the water pressure change rate and / or the flow rate change rate.

7. The method as described in claim 5, characterized in that, The process of identifying pre-disaster precursors of the target engineering coal and rock mass sample based on the infrared low-temperature cluster characteristic parameters, the acoustic emission characteristic parameters, and the hydraulic characteristic parameters, and obtaining the identification results, specifically includes: The baseline mean and standard deviation are calculated for the continuous characteristic parameters, acoustic emission characteristic parameters and hydraulic characteristic parameters in the infrared low temperature cluster characteristic parameters, respectively, and anomaly judgment functions based on the baseline mean and standard deviation are used to judge the anomalies of the continuous characteristic parameters, acoustic emission characteristic parameters and hydraulic characteristic parameters. Based on the degree of deviation of the dispersion or aggregation state characteristic parameters from the average number of low-temperature clusters in the baseline stage, anomaly judgment is made on the dispersion or aggregation state characteristic parameters. Based on the current value of the low-temperature cluster boundary connectivity index, an anomaly judgment is made on the low-temperature cluster boundary connectivity index. When the low-temperature cluster boundary connectivity index is equal to 1, it is determined that it has reached an abnormal state. Based on the anomaly judgment results corresponding to different characteristic parameters, the catastrophic precursors of the target engineering coal and rock mass sample are identified, and the identification results are obtained. The continuous characteristic parameters include the infrared radiation variance characteristic parameter, the low-temperature point area ratio characteristic parameter, the maximum low-temperature cluster area ratio characteristic parameter, and the maximum low-temperature cluster centroid migration characteristic parameter.

8. The method as described in claim 7, characterized in that, The method of identifying pre-disaster precursors of the target engineering coal and rock mass sample based on anomaly judgment results corresponding to different characteristic parameters, and obtaining the identification results, specifically includes: When the anomaly judgment result is that any two of the infrared radiation variance characteristic parameter, the low temperature point area ratio characteristic parameter, the maximum low temperature cluster area ratio characteristic parameter, and the dispersion or aggregation state characteristic parameter are abnormal and the anomaly duration is greater than or equal to the length of the first preset time window, it is determined that the current state is in a level one warning state and a level one warning is issued. When the anomaly judgment result is that any three of the following three characteristic parameters are abnormal: the low temperature point area ratio characteristic parameter, the maximum low temperature cluster area ratio characteristic parameter, the maximum low temperature cluster centroid migration characteristic parameter, and the acoustic emission event rate and acoustic emission energy growth rate characteristic parameters, and the anomaly duration is greater than or equal to the length of the second preset time window, it is determined that the current state is in a level two warning state, and a level two warning is issued. When the connectivity index of the low-temperature cluster boundary reaches an abnormal state, at least one of the characteristic parameters of the acoustic emission event rate and the acoustic emission energy growth rate is abnormal, the collected hydraulic characteristic parameters are abnormal, and at least one of the characteristic parameters of the infrared radiation variance, the low-temperature point area ratio, the maximum low-temperature cluster area ratio, the maximum low-temperature cluster centroid migration, and the dispersion or aggregation state is abnormal, and the above abnormalities continue synchronously within a third preset time window; it is determined that the current state is in a level three warning state, and a level three warning is issued.

9. A device for identifying precursors of coal and rock mass disasters in engineering projects, characterized in that, include: The acquisition module is used to acquire infrared thermal image data, acoustic emission data and hydraulic data during the hydraulic coupling loading process of the target engineering coal and rock mass sample using a preset hydraulic coupling loading device. The hydraulic data includes water pressure data and / or flow rate data. The first feature extraction module is used to extract low-temperature features from the infrared thermal image data using a feature extraction method that combines dynamic low-temperature threshold and benchmark low-temperature deviation to obtain infrared low-temperature cluster feature parameters characterizing the spatiotemporal evolution of infrared radiation. The second feature extraction module is used to extract features from the acoustic emission data and the hydraulic data respectively to obtain acoustic emission feature parameters and hydraulic feature parameters. The identification module is used to identify the catastrophic precursors of the target engineering coal and rock mass sample based on the infrared low-temperature cluster feature parameters, the acoustic emission feature parameters, and the hydraulic feature parameters, and to obtain the identification results.

10. An electronic device, characterized in that, It includes at least a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the method for identifying precursors of engineering coal and rock mass disasters as described in any one of claims 1-8.

Images