Mine thermodynamic disaster detection and prevention integrated intelligent cooperative emergency method and system

Abstract

The invention relates to the technical field of mine safety monitoring and disaster prevention and control, and provides a mine thermodynamic disaster detection and prevention integrated intelligent cooperative emergency method and system aiming at the problems of detection and prevention separation, response lag and model staticization in the prior art. Establishing a disaster identification model by using a machine learning algorithm to realize dynamic identification and risk grading of coal spontaneous combustion and gas explosion; a cooperative emergency instruction is automatically generated according to the risk level, and an underground prevention and control device is linked based on the cooperative emergency instruction to perform operations such as gas extraction, ventilation adjustment and spray cooling, so that synchronous response of early warning and prevention and control is realized; the system has a self-learning optimization capability, and continuously corrects a model and a strategy through backtracking analysis. According to the invention, accurate early warning, quick response and dynamic optimization of disaster prevention and control are realized, and the intelligent level and intrinsic safety capability of mine safety management are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of mine safety monitoring and disaster prevention technology, specifically to an intelligent collaborative emergency response method and system for detecting and preventing thermal and dynamic disasters in mines. Background Technology

[0002] Currently, monitoring and prevention technologies for mine thermal disasters mostly focus on single disaster types, such as independently constructing coal spontaneous combustion marker gas detection or methane concentration monitoring networks. While these systems have achieved certain results in specific areas, they generally suffer from the problem of "separation of detection and prevention," meaning that disaster detection and prevention are disconnected in terms of technology and implementation, lacking the ability to coordinate analysis and joint response to the thermal disaster chain, making it difficult to assess the overall risk situation of the mine.

[0003] Furthermore, existing monitoring methods, such as bundled tubes, optical fibers, and gas sensors, often operate independently, creating data silos. Their early warning models are mostly based on simple threshold judgments, lacking the ability to deeply integrate and intelligently mine multi-source heterogeneous data. This results in low early warning accuracy, high false alarm rates, and reliance on manual judgment for emergency response. It also prevents rapid, intelligent decision-making and automated handling based on real-time risk in the early stages of disaster evolution, often missing the optimal intervention window. Therefore, developing an integrated method capable of multi-source information fusion, intelligent analysis, and coordinated exploration and prevention has become an urgent need to improve mine safety. Summary of the Invention

[0004] In order to solve at least one of the above-mentioned technical problems in the prior art, the present invention provides an intelligent collaborative emergency response method and system for detecting and preventing thermal and dynamic disasters in mines.

[0005] In a first aspect, the present invention provides an intelligent collaborative emergency response method integrating detection and prevention of thermal and dynamic disasters in mines, the method comprising: Historical mine thermal and dynamic disaster data are acquired and preprocessed to obtain multimodal fusion data. The historical mine thermal and dynamic disaster data includes at least infrared thermal imaging data and disaster gas composition data. A disaster identification model is established based on the multimodal fusion data using a hybrid algorithm structure of convolutional neural networks and random forests. The convolutional neural network is used to extract deep features from infrared thermal imaging data in the historical mine thermal disaster data; the random forest is used to process the disaster gas composition data in the historical mine thermal disaster data. Real-time thermal disaster data of mines is collected by infrared thermal imagers, gas analysis devices, gas concentration sensors and temperature and humidity sensors deployed underground in the mine. The real-time thermal disaster data of mines is input into the disaster identification model to dynamically identify thermal disasters and assess their risk levels, thereby obtaining the risk level. The risk level is compared with a preset warning threshold. When the risk level is not lower than the preset warning threshold, a collaborative emergency command is generated based on the risk level and transmitted to the downhole control device for execution.

[0006] In one optional implementation, the historical mine thermal and dynamic disaster data is preprocessed, including: The historical mine thermal and dynamic disaster data were cleaned and outliers and missing data were removed to obtain multi-source data; The multi-source data is synchronized in time using a time-series registration algorithm, and multimodal fusion is performed using a weight-based time-series feature fusion algorithm to obtain multimodal fused data.

[0007] In one optional embodiment, the gas analysis device is an infrared Fourier transform analyzer, used to detect the composition of mine gas and obtain hazardous gas composition data; the infrared thermal imager is used to collect R, G, B and temperature characteristic data during the coal oxidation and spontaneous combustion process.

[0008] In one optional implementation, generating a coordinated emergency response instruction based on the risk level includes: Based on a preset knowledge base, historical disaster cases stored in the preset knowledge base are matched according to the risk level, and collaborative emergency instructions are generated according to the matching results. The collaborative emergency instructions include at least one of the following: linkage control of gas extraction, ventilation regulation, spray cooling and explosion-proof isolation devices.

[0009] In one alternative implementation, it further includes: The execution results of the downhole prevention and control device are backtracked and analyzed. Based on the backtracking analysis results, the parameters of the disaster identification model and the collaborative emergency command are optimized through a self-learning algorithm to achieve dynamic updating and continuous self-adaptation of the parameters.

[0010] In one optional implementation, the self-learning algorithm includes: An adaptive algorithm based on gradient descent and reinforcement learning is used to iteratively retrain the disaster identification model and collaborative emergency commands using the execution results of the downhole control device.

[0011] Secondly, the present invention provides an integrated intelligent collaborative emergency response system for the detection and prevention of thermal and dynamic disasters in mines, capable of realizing an integrated intelligent collaborative emergency response method for the detection and prevention of thermal and dynamic disasters in mines as described in the first aspect and any optional embodiment, comprising: The data acquisition module is used to collect real-time mine thermal disaster data through the infrared thermal imager, gas analysis device, gas concentration and temperature and humidity sensor; The intelligent analysis module is used to perform fusion analysis on the real-time mine thermal and dynamic disaster data, and to use the disaster identification model to dynamically identify and assess the risk level of thermal and dynamic disasters, thereby obtaining the risk level. The collaborative control module is used to automatically generate collaborative emergency commands based on the risk level and coordinate with the downhole control device to execute them. The self-learning optimization module is used to perform retrospective analysis of system operation data and retrain the disaster identification model to achieve continuous optimization of disaster prediction and prevention strategies.

[0012] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves intelligent closed-loop management of the entire process of mine thermal and dynamic disasters, from detection and assessment to prevention, control, and emergency response, by integrating multimodal data, disaster identification models, and collaborative control mechanisms. By fusing multimodal information such as infrared thermal images, gas composition, and temperature gradients, the accuracy and timeliness of disaster feature identification are significantly improved. The disaster identification model based on machine learning algorithms can predict the occurrence trends of disasters such as spontaneous combustion of coal and gas explosions in mines, realizing a shift from passive monitoring to proactive early warning. This invention automatically generates differentiated prevention and control strategies based on risk levels and links gas extraction, ventilation regulation, spray cooling, and explosion-proof isolation devices through an integrated detection and prevention platform, forming a synchronous linkage mechanism for early warning, prevention, and emergency response, effectively improving disaster handling efficiency and safety response capabilities. Simultaneously, this invention possesses self-learning and continuous optimization functions, continuously correcting model parameters and control strategies during disaster evolution and handling, exhibiting dynamic adaptive and evolutionary optimization characteristics, thereby significantly improving the intelligence and inherent safety level of mine safety management. Attached Figure Description

[0013] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0014] Figure 1This is a flowchart illustrating the intelligent collaborative emergency response method for mine thermal and dynamic disaster detection and prevention according to an embodiment of the present invention. Figure 2 This is a structural block diagram of an intelligent collaborative emergency system for detecting and preventing thermal and dynamic disasters in mines, according to an embodiment of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Currently, several typical technical approaches have been developed in the field of mine safety monitoring and disaster prevention. However, these approaches still have significant limitations in dealing with complex and cascading disaster risks. (1) Single disaster mechanism research and traditional monitoring system: Such research and practice often focus on a single type of disaster, such as focusing on the detection of gas markers of spontaneous combustion of coal, or independently constructing a gas concentration monitoring network. Although these systems or methods have achieved certain results in their specific fields, they lack synergistic consideration and linkage analysis of the thermodynamic disaster chain (such as gas explosion caused by spontaneous combustion of coal), and cannot assess the overall risk situation of the mine.

[0017] (2) Decentralized monitoring technologies and non-intelligent early warning models: In existing technologies, various monitoring methods such as bundled tube monitoring, fiber optic temperature measurement, and gas sensing usually operate independently, resulting in strict data barriers. Their early warning models are mostly based on simple threshold judgments or static statistical laws, and lack the ability to deeply integrate and intelligently mine multi-source heterogeneous monitoring data. They are unable to capture the cross-modal and nonlinear characteristics of disaster precursors, resulting in low early warning accuracy and high false alarm rate.

[0018] (3) Delayed response mechanism and emergency mode of separation of "detection and prevention": Current emergency response mostly relies on manual judgment and segmented handling process. Disaster detection ("detection") and disaster prevention ("prevention") are disconnected from each other in terms of technical system and execution process. This mode lacks the ability to make rapid intelligent decisions and automated handling based on real-time risks in the early stage of disaster evolution, and cannot achieve coordinated emergency response of "detection and prevention integration", often missing the best intervention opportunity.

[0019] Specifically, although in-depth research has been conducted on coal spontaneous combustion mechanisms, gas explosion suppression, and the division of the "three zones" in goaf areas, and technologies such as in-situ spectral beam tube monitoring have been developed, the complex geological conditions of coal mines in my country mean that major thermal and dynamic disasters still occur frequently. This profoundly exposes the lack of core capabilities in the existing technological system regarding multi-source information fusion, intelligent judgment and decision-making, and collaborative exploration and prevention.

[0020] Therefore, developing a collaborative emergency response method and system that can deeply integrate multimodal monitoring data, realize intelligent disaster assessment, and connect the "exploration" and "prevention" links to achieve early identification, early warning, and early handling of thermal and dynamic disasters in mines has become an urgent technical requirement for improving the level of coal mine safety assurance. This is also the core problem that this invention aims to solve.

[0021] According to an embodiment of the present invention, an embodiment of an intelligent collaborative emergency response method integrating detection and prevention of thermal and dynamic disasters in mines is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0022] This embodiment provides an intelligent collaborative emergency response method integrating detection and prevention of thermal and dynamic disasters in mines. Figure 1 This is a flowchart of an intelligent collaborative emergency response method for mine thermal and dynamic disaster detection and prevention according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: S1: Acquire historical mine thermal and dynamic disaster data, and preprocess the historical mine thermal and dynamic disaster data to obtain multimodal fusion data. The historical mine thermal and dynamic disaster data includes at least infrared thermal imaging data and disaster gas composition data.

[0023] Optionally, the historical mine thermal and dynamic disaster data is preprocessed, including: cleaning the historical mine thermal and dynamic disaster data and removing outliers and missing points to obtain multi-source data; synchronizing the multi-source data using a time-series registration algorithm, and performing multi-modal fusion using a weight-based time-series feature fusion algorithm to obtain multi-modal fused data.

[0024] In this embodiment, historical mine thermal disaster data is initially cleaned by edge computing nodes to remove outliers and missing data, resulting in multi-source data; then, a time-series registration algorithm is used to achieve time synchronization of the multi-source data.

[0025] In this embodiment, outlier removal is achieved by: The criteria identify and delete extreme outliers in temperature and gas concentration in historical mine thermal disaster data, and use adjacent frame interpolation to fill in missing frames in infrared thermal image data; time registration: based on the mine synchronous clock (BeiDou time synchronization or underground industrial clock), historical mine thermal disaster data from different acquisition devices are aligned to a unified time axis with a registration accuracy of ≤1 second.

[0026] In this embodiment, after obtaining multi-source data, numerical data such as temperature, gas concentration, and wind speed can be standardized using Z-score.

[0027] In this embodiment, a time-series feature fusion algorithm based on weight allocation is used for multimodal fusion. For example, in the early stage of coal spontaneous combustion, fusion weights of 0.45, 0.35 and 0.20 are dynamically allocated to the infrared temperature gradient, gas component change rate and temperature and humidity change.

[0028] S2: A disaster identification model is established based on the multimodal fusion data using a hybrid algorithm structure of convolutional neural network and random forest. The convolutional neural network is used to extract deep features from the infrared thermal imaging data in the historical mine thermal disaster data; the random forest is used to process the disaster gas composition data in the historical mine thermal disaster data.

[0029] In this embodiment, a disaster identification model is constructed using a hybrid algorithm of convolutional neural network and random forest. The convolutional neural network is dedicated to extracting deep features from infrared thermal images, while the random forest processes disaster gas composition data from historical mine thermal disaster data, such as gas change rate and ventilation parameters. Based on multi-dimensional features such as infrared temperature gradient, carbon monoxide growth rate, and gas concentration fluctuation, the model outputs dynamic identification results and risk levels for thermal disasters such as coal spontaneous combustion and gas explosion in real time.

[0030] In this embodiment, the constructed disaster identification model is trained by dividing the multimodal fusion data into training, validation, and test sets in a 7:2:1 ratio. The cross-entropy loss function is used to calculate the model prediction error. The model is iteratively trained using the Adam optimizer (learning rate 0.001, decay coefficient 0.9). The validation set is used to adjust hyperparameters such as the number of convolutional kernels in the convolutional neural network and the number of decision trees in the random forest until the disaster identification model achieves a risk level identification accuracy of ≥95% on the test set.

[0031] S3: Real-time mine thermal disaster data is collected by infrared thermal imagers, gas analysis devices, gas concentration sensors, and temperature and humidity sensors deployed underground in the mine. The real-time mine thermal disaster data is then input into the disaster identification model to dynamically identify and assess the risk level of the thermal disaster, thereby obtaining the risk level.

[0032] Optionally, the gas analysis device is an infrared Fourier transform analyzer, used to detect the composition of mine gas and obtain hazardous gas composition data; the infrared thermal imager is used to collect R, G, B and temperature characteristic data during the coal oxidation and spontaneous combustion process.

[0033] In this embodiment, real-time mine thermal disaster data is collected using a dual-mode acquisition method of "fixed sensing + mobile inspection": Fixed sensing acquisition: Infrared thermal imagers are pre-embedded in the goaf of the mine (one per 50 meters, with a field of view of 60°), and a set of gas concentration sensors (capable of simultaneously detecting CO, CO2, and CH4 concentrations, with detection accuracies of ±1ppm, ±0.1%, and ±0.01%, respectively) are deployed every 30 meters in the return airway. Temperature and humidity sensors and wind speed sensors are deployed at the working face (collecting data once per minute). All fixed equipment transmits data in real time through the underground industrial bus. Mobile inspection acquisition: A mine explosion-proof inspection robot (equipped with a portable infrared Fourier transform analyzer and a high-definition camera) is used to inspect the mine every 2 hours along a preset path (covering the blind spots of fixed sensors). It collects deep coal temperature, local gas concentration, and coal oxidation discoloration image data. The inspection data is transmitted back to the ground processing terminal via a wireless mesh network.

[0034] In this embodiment, before the real-time mine thermal and dynamic disaster data is input into the disaster identification model, it can also be preprocessed in real time through edge computing nodes: smoothing filtering is performed on the instantaneous jump data of fixed sensors (such as gas concentration fluctuating by more than 50% within 1 second), and distortion correction is performed on the image data of mobile inspection to ensure that the real-time data and the historical multimodal fusion data are consistent in format and dimension.

[0035] In this embodiment, the risk level is set to at least three levels: low risk, medium risk, and high risk. For the low risk level, the threshold conditions can be set as follows: coal temperature ≤ 30℃, CO concentration ≤ 10ppm, CH4 concentration ≤ 0.5%, and no obvious infrared temperature anomaly area. For the medium risk level, the threshold conditions can be set as follows: coal temperature 30-60℃ and CO concentration 10-50ppm (corresponding to the initial stage of coal spontaneous combustion) or CH4 concentration 0.5%-1.0% and wind speed ≤ 0.5m / s. For the high risk level, the threshold conditions can be set as follows: coal temperature > 60℃ or CO concentration > 50ppm (corresponding to the active period of coal spontaneous combustion); CH4 concentration ≥ 1.0% (corresponding to the risk of gas explosion); and simultaneously, coal temperature > 40℃ and CH4 concentration > 0.8% (corresponding to the risk of coal spontaneous combustion-gas coupled disaster). When the risk level is lower than the corresponding preset threshold, the current state is maintained and monitoring continues. If the risk level is equal to or higher than the corresponding preset threshold, a subsequent coordinated emergency command is generated.

[0036] S4: Compare the risk level with the preset warning threshold. When the risk level is not lower than the preset warning threshold, generate a collaborative emergency command based on the risk level and transmit the collaborative emergency command to the downhole control device for execution.

[0037] Optionally, generating collaborative emergency instructions based on the risk level includes: matching historical disaster cases stored in the preset knowledge base according to the risk level, and generating collaborative emergency instructions based on the matching results, wherein the collaborative emergency instructions include at least one of the following: linkage control of gas extraction, ventilation regulation, spray cooling and explosion-proof isolation devices.

[0038] In this embodiment, a dual-driven logic of "rule base + case base" from a preset knowledge base is used to generate differentiated collaborative emergency commands. The rule base drives the generation of pre-stored basic prevention and control rules bound to risk levels. For example, when the risk level is low, the basic rule is "start the corresponding area spray cooling system (spray volume 50L / min), adjust the opening of the return airway door to 80%, and increase the flow rate of the gas extraction pump by 20%"; when the risk level is medium, the basic rule is "start the local ventilation fan (increase the wind speed to 1.0m / s), shut down unnecessary electrical equipment in the area, and turn on the gas dilution device"; when the risk level is high, the basic rule is "cut off the power supply to the corresponding area working face, start the explosion-proof airtight door closing procedure, and turn on the mine emergency refuge chamber oxygen supply system". For example, for medium-risk disasters of gas accumulation type, the "extraction-ventilation joint control" strategy is generated first; for low-risk disasters of spontaneous combustion type, the "cooling-isolation-inerting" combined strategy is triggered first.

[0039] Case-based approach: Retrieve historical disaster cases stored in the preset knowledge base that have "risk level, disaster type, and regional conditions consistent with the current situation," and extract "experience in adjusting prevention and control command parameters" from the cases to revise basic rules. For example, if a certain area in a historical case is at a medium risk level, the prevention and control effect of spray cooling + increasing inert gas injection (nitrogen flow rate 10m³ / min) is improved by 40%, then add a supplementary command "start nitrogen injection system (flow rate 10m³ / min)" to the current collaborative emergency command.

[0040] Optionally, it also includes: performing backtracking analysis based on the execution results of the downhole control device, and optimizing the parameters of the disaster identification model and the collaborative emergency command through a self-learning algorithm based on the backtracking analysis results, so as to achieve dynamic updating and continuous self-adaptation of the parameters.

[0041] Optionally, the self-learning algorithm includes: using an adaptive algorithm based on gradient descent and reinforcement learning to iteratively retrain the disaster identification model and collaborative emergency commands using the execution results of the downhole control device.

[0042] In this embodiment, after each disaster prevention and control operation, the content of the collaborative emergency command, execution process data, risk level change curve, and final prevention and control results are recorded. The recorded data is then added to a preset knowledge base, and a self-learning algorithm is used to optimize the feature weights of the multimodal fusion data (e.g., if multiple cases show that "inert gas concentration" has a significant impact on the prevention and control of spontaneous combustion of coal, its weight is increased from 0.1 to 0.2 during fusion). At the same time, the disaster identification model is iteratively updated (the recorded data is added to the training set, validation set, and test set in a 7:2:1 ratio to retrain the disaster identification model and ensure that the accuracy of the disaster identification model is consistently ≥95%).

[0043] This embodiment also provides an intelligent collaborative emergency response system integrating mine thermal disaster detection and prevention. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0044] This embodiment provides an integrated intelligent collaborative emergency response system for detecting and preventing thermal and dynamic disasters in mines, such as... Figure 2 As shown, it includes: The data acquisition module is used to collect real-time mine thermal disaster data through the infrared thermal imager, gas analysis device, gas concentration and temperature and humidity sensor; The intelligent analysis module is used to perform fusion analysis on the real-time mine thermal and dynamic disaster data, and to use the disaster identification model to dynamically identify and assess the risk level of thermal and dynamic disasters, thereby obtaining the risk level. The collaborative control module is used to automatically generate collaborative emergency commands based on the risk level and coordinate with the downhole control device to execute them. The self-learning optimization module is used to perform retrospective analysis of system operation data and retrain the disaster identification model to achieve continuous optimization of disaster prediction and prevention strategies.

[0045] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0046] The intelligent collaborative emergency system for mine thermal disaster detection and prevention in this embodiment is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0047] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0048] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A smart collaborative emergency response method integrating detection and prevention of thermal and dynamic disasters in mines, characterized in that, The method includes: Historical mine thermal and dynamic disaster data are acquired and preprocessed to obtain multimodal fusion data. The historical mine thermal and dynamic disaster data includes at least infrared thermal imaging data and disaster gas composition data. A disaster identification model is established based on the multimodal fusion data using a hybrid algorithm structure of convolutional neural networks and random forests. The convolutional neural network is used to extract deep features from infrared thermal imaging data in the historical mine thermal disaster data; the random forest is used to process the disaster gas composition data in the historical mine thermal disaster data. Real-time thermal disaster data of mines is collected by infrared thermal imagers, gas analysis devices, gas concentration sensors and temperature and humidity sensors deployed underground in the mine. The real-time thermal disaster data of mines is input into the disaster identification model to dynamically identify thermal disasters and assess their risk levels, thereby obtaining the risk level. The risk level is compared with a preset warning threshold. When the risk level is not lower than the preset warning threshold, a collaborative emergency command is generated based on the risk level and transmitted to the downhole control device for execution.

2. The intelligent collaborative emergency response method for detecting and preventing thermal and dynamic disasters in mines according to claim 1, characterized in that, The historical mine thermal and dynamic disaster data are preprocessed, including: The historical mine thermal and dynamic disaster data were cleaned and outliers and missing data were removed to obtain multi-source data; The multi-source data is synchronized in time using a time-series registration algorithm, and multimodal fusion is performed using a weight-based time-series feature fusion algorithm to obtain multimodal fused data.

3. The intelligent collaborative emergency response method for detecting and preventing thermal and dynamic disasters in mines according to claim 1, characterized in that, The gas analysis device is an infrared Fourier transform analyzer, used to detect the composition of mine gas and obtain hazardous gas composition data; the infrared thermal imager is used to collect R, G, B and temperature characteristic data during the coal oxidation and spontaneous combustion process.

4. The intelligent collaborative emergency response method for detecting and preventing thermal and dynamic disasters in mines according to claim 1, characterized in that, Generate collaborative emergency response instructions based on the risk level, including: Based on a preset knowledge base, historical disaster cases stored in the preset knowledge base are matched according to the risk level, and collaborative emergency instructions are generated according to the matching results. The collaborative emergency instructions include at least one of the following: linkage control of gas extraction, ventilation regulation, spray cooling and explosion-proof isolation devices.

5. The intelligent collaborative emergency response method for detecting and preventing thermal and dynamic disasters in mines according to claim 1, characterized in that, Also includes: The execution results of the downhole prevention and control device are backtracked and analyzed. Based on the backtracking analysis results, the parameters of the disaster identification model and the collaborative emergency command are optimized through a self-learning algorithm to achieve dynamic updating and continuous self-adaptation of the parameters.

6. The intelligent collaborative emergency response method for detecting and preventing thermal and dynamic disasters in mines according to claim 5, characterized in that, The self-learning algorithm includes: An adaptive algorithm based on gradient descent and reinforcement learning is used to iteratively retrain the disaster identification model and collaborative emergency commands using the execution results of the downhole control device.

7. A mine thermal disaster detection and prevention integrated intelligent collaborative emergency response system, used to implement the mine thermal disaster detection and prevention integrated intelligent collaborative emergency response method according to any one of claims 1-6, characterized in that, include: The data acquisition module is used to collect real-time mine thermal disaster data through the infrared thermal imager, gas analysis device, gas concentration and temperature and humidity sensor; The intelligent analysis module is used to perform fusion analysis on the real-time mine thermal and dynamic disaster data, and to use the disaster identification model to dynamically identify and assess the risk level of thermal and dynamic disasters, thereby obtaining the risk level. The collaborative control module is used to automatically generate collaborative emergency commands based on the risk level and coordinate with the downhole control device to execute them. The self-learning optimization module is used to perform retrospective analysis of system operation data and retrain the disaster identification model to achieve continuous optimization of disaster prediction and prevention strategies.

Images