Underground fire area land son-mother ship detection device based on intelligent optical fiber communication and explosion prediction

Abstract

The invention discloses an underground fire area land sub-carrier and mother-carrier detection device based on intelligent optical fiber communication and explosion prediction, and relates to the field of collaborative operation of coal mine emergency rescue cluster robots, and the underground fire area land sub-carrier and mother-carrier detection device comprises a communication system, a mother carrier and a sub-carrier unmanned aerial vehicle; the communication system comprises an intelligent winch device and an optical fiber, and is in signal connection with the ship body and the unmanned aerial vehicle through the optical fiber; the mother ship comprises a ship body and a deck, the ship body comprises a walking device and an energy supply device, a miniature gas chromatographic analysis module is carried in the deck, a plurality of sub-ship unmanned aerial vehicle parking platforms are arranged on the upper portion of the deck, and each sub-ship unmanned aerial vehicle parking platform comprises a charging module and an optical fiber intelligent scaling module; the optical fiber intelligent scaling module is connected with the unmanned aerial vehicle through an optical fiber, and the unmanned aerial vehicle collects gas for analysis or returns to a mother ship for analysis; according to the invention, the problems of unstable communication, incapability of realizing high-confidence judgment of an explosion source and insufficient endurance in traditional coal mine fire area detection are solved, and the gas detection precision and the explosion early warning capability in a high-risk environment are improved.

Description

Technical Field

[0001] This invention relates to the field of collaborative operation technology of cluster robots for emergency rescue in coal mines, and more specifically, to a land-based mother-daughter ship detection device for underground fire zones based on intelligent optical fiber communication and explosion prediction. Background Technology

[0002] In the coal mining industry, with increasing mining depth and intensity, the risk of underground fires is becoming increasingly prominent. During the "closure-treatment-unsealing" process in fire zone management: external power networks are cut off, and traditional detection equipment relies on short battery life, making it unsuitable for long-term continuous operation; high temperatures, high humidity, and electromagnetic interference underground lead to high packet loss rates in wireless communication, making it impossible to transmit critical fire zone data in real time; the confined space and rugged terrain of the fire zone result in low mobility for traditional tracked robots, and drones lack sufficient obstacle avoidance capabilities, leading to a high risk of collisions. How to detect gas and temperature hazards in these complex underground scenarios has become a significant safety challenge in coal mine rescue. However, existing technologies have significant limitations in such complex scenarios, specifically in the following aspects: Traditional detection methods face communication and data limitations. Current technologies primarily rely on manually carried equipment or single robots for underground fire zone detection, such as CN201810001234.5 "An Underground Coal Mine Detection Robot." These devices often employ wireless communication technologies like WiFi or ZigBee. However, as reported in the 2021 issue 5 of *Coal Mine Safety*, "Analysis of Wireless Communication Interference in Underground Mines," the high temperature, high humidity, and complex electromagnetic environment underground result in a wireless signal packet loss rate as high as 30%, making it impossible to transmit critical fire zone data in real time, such as CO concentration and temperature gradient. Furthermore, single robots are limited by their endurance and detection range, making it difficult to fully cover the fire zone. This often leads to errors in delineating closed areas exceeding ±5 meters. In 2020, a coal mine in Shanxi Province experienced a reignition accident due to misjudgment of the fire zone boundary.

[0003] Due to their advantages such as high mobility, flexible deployment, and ability to cover complex terrain, unmanned aerial vehicle (UAV) platforms are increasingly being used in reconnaissance missions in high-risk environments. However, traditional UAV data links based on wireless communication are unstable in fire or explosion sites with strong electromagnetic interference or severe signal obstruction, and are prone to data interruption and link damage, failing to meet the requirements for high-reliability, low-latency data backhaul in explosion prediction scenarios.

[0004] Furthermore, conventional explosion risk prediction methods often rely on single-type sensor data, such as infrared thermal imaging, high-temperature alarms, or fixed-point gas concentration threshold judgments, lacking qualitative and quantitative analysis methods for multi-source gas components. This hinders a deep understanding and high-confidence determination of the explosion source formation mechanism. Simultaneously, existing power supply mechanisms struggle to maintain continuous system operation after on-site power system failures and lack intelligent switching energy management strategies, limiting the autonomous operation capability and survival time of subsystems.

[0005] To address the above problems, this invention proposes a solution. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a land-based mothership detection device for underground fire zones based on intelligent optical fiber communication and explosion prediction. This device effectively solves key technical bottlenecks in traditional coal mine fire zone detection, such as unstable communication, inability to achieve high-confidence determination of explosion sources, and insufficient endurance. It significantly improves the gas detection accuracy and explosion early warning capability in high-risk environments.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A land-based mothership detection device for underground fire zones based on intelligent fiber optic communication and explosion prediction includes a communication system, a hull, a deck, and a daughtership UAV. Its features include: the communication system is connected to the hull and UAV via fiber optic cables, and includes an intelligent winch device for dynamically adjusting the fiber optic length; the hull includes a walking mechanism and a dual-mode power supply device, which has a cable interface for connecting to an external power source and a detachable battery pack for switching power supply via the internal battery when external power is interrupted; the deck is located on the upper part of the hull, and includes an internally installed micro gas chromatography analysis module and an upper-mounted UAV landing platform, which includes a charging module and an intelligent fiber optic scaling module; the daughtership UAV carries a gas collection device and communicates with the hull via the intelligent fiber optic scaling module, and is equipped with a gas sensor for real-time analysis of the collected gas, transmitting the analysis results back to the mothership via fiber optic link or returning the gas sample to the landing platform; the mothership, through comprehensive judgment of the UAV analysis data, locates the fire source and predicts the risk level of the explosion.

[0008] In a preferred embodiment, the intelligent winch device includes a support rod, a rotating mechanism, and a winch. The winch is driven by a servo motor and automatically winds up and winds up optical fibers based on the positioning data of the UAV. The optical fibers include a first optical fiber and a second optical fiber.

[0009] In a preferred embodiment, the method for switching power supply via an internal battery when external power is interrupted involves: acquiring and normalizing mothership status data to obtain input variables, including fire zone proximity, cable temperature, and remaining battery power; defining the input variables as fuzzy sets and mapping them to fuzzy input items, and constructing a fuzzy rule base; for each rule, using its membership degree as the rule's activation strength, and performing fuzzy inference on the fuzzy input items and the fuzzy rule base using a Mamdani-type inference mechanism to obtain a fuzzy output result; defuzzifying the fuzzy output result using the centroid method to obtain a switching control signal; and controlling the dual-mode power supply device to switch to power supply via the detachable battery pack according to the switching control signal.

[0010] In a preferred embodiment, the collected gas is analyzed in real time, and the analysis results are transmitted back to the mothership via an optical fiber link. Specifically, the collected gas is introduced into a gas analysis module, which includes a spectral analysis unit and an electrochemical sensing unit. The spectral analysis unit performs absorption spectroscopy or Raman scattering analysis on the collected gas to obtain the spectral response characteristics of the gas components. The electrochemical sensing unit performs quantitative detection of the gas and outputs current-concentration mapping data. Based on the Bayesian weighted average method, the spectral response characteristics and current-concentration mapping data are cross-fused to obtain a high-confidence gas feature vector data package. The gas feature vector data package is connected to the optical fiber communication link through an optical fiber intelligent scaling module and transmitted back to the mothership's main control system in real time via a second optical fiber.

[0011] In a preferred embodiment, the sub-ship UAV further includes a wireless relay module, specifically: real-time monitoring of the fiber optic link status, including signal strength, delay, and link integrity; if a link anomaly is detected, the wireless relay module automatically determines it to be in a fault state and actively performs a physical disconnection of the fiber optic link; after the link is disconnected, the module enters an emergency mode, scans and identifies the nearest sub-ship UAV as the relay target, selects a node based on signal strength and synchronization information; establishes an emergency communication link with the selected sub-ship UAV using a dynamic routing protocol; encapsulates and identifies the analysis results data, and transmits it back to the sub-ship UAV through the established link, which then forwards it to the mothership's main control system.

[0012] In a preferred embodiment, the return of the gas sample to the parking platform involves the following steps: When the sub-ship UAV detects in real time that its battery level is below a preset threshold of 20%, a return-to-home mechanism is triggered; 3D obstacle information of the current environment is collected to construct a local obstacle grid map, and an obstacle avoidance path is planned using an improved A* algorithm combined with a dynamic window algorithm to generate an optimal return-to-home path; the sub-ship UAV is controlled to fly along the planned path, and the lidar perception data is refreshed in real time during flight to dynamically correct the path; after the sub-ship UAV flies to the parking platform, the magnetic contact points on its bottom are controlled to precisely dock with the charging module on the parking platform; after the charging module docking is completed, the miniature gas chromatography analysis module built into the platform is triggered to perform a secondary analysis of the gas sample collected by the sub-ship UAV; the secondary analysis results are compared with the initial analysis results during the flight mission for redundancy detection, and a system warning signal is issued based on the detection results.

[0013] In a preferred embodiment, the secondary analysis results are compared with the initial analysis results during the flight mission for redundancy detection. Based on the detection results, a system warning signal is issued. Specifically, the following steps are taken: the chromatogram datasets after the initial analysis and the secondary chromatogram datasets after the secondary analysis are obtained and normalized; the sample chromatogram difference is calculated using Kullback-Leibler divergence between the processed chromatogram datasets and the secondary chromatogram datasets; the sample chromatogram difference is compared with a preset redundancy tolerance threshold. If the sample chromatogram difference is greater than the preset redundancy tolerance threshold, it is determined that the current secondary gas sample and the sample collected during the initial flight mission have significant differences in chromatographic characteristics, a sample deviation risk marker is generated, and the marker is entered into the system alarm queue.

[0014] In a preferred embodiment, the mothership locates the ignition source and predicts the risk level of the explosion by comprehensively judging the data analyzed by the UAV. Specifically, the process involves: acquiring the data transmitted back from the UAV; performing ignition source clustering analysis using a multi-scale high-temperature anomaly extraction and spatial density clustering algorithm to obtain a set of candidate ignition sources and their spatial coordinates; and constructing an explosion risk assessment function model in the edge computing node using the UAV. The transmitted data includes gas concentration data, gas component spectrum heterogeneity data, and ignition source-related parameter data. A risk value is calculated based on the explosion risk assessment function model, and the risk value is compared with a preset explosion threshold. Based on the comparison results, a dynamic hierarchical threshold interval is adaptively generated using Bayesian update and sliding window methods. The current risk value is mapped to the corresponding risk level according to the threshold interval. The spatial coordinates and time label of the work surface are acquired, and a lightweight clustering algorithm is used to perform local area risk clustering analysis in conjunction with the risk level, outputting a spatiotemporal evolution trend map of the risk level and the coordinates of the center point of the spatiotemporal cluster of high-incidence explosions. Based on the spatiotemporal evolution trend map and the coordinates of the center point of the spatiotemporal cluster with high explosion incidence, executable control suggestions and operating procedures are generated for different levels.

[0015] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This land-based mother-daughter ship, utilizing swarm robotics technology, overcomes safety challenges such as power outages, communication failures, detection blind spots, and delayed risk assessments caused by gas and temperature hazards in enclosed fire-affected areas. It forms a real-time feedback closed loop from detection and analysis to treatment, significantly improving the scientific rigor and safety of the entire fire zone handling process. It provides an innovative solution for the safe and efficient handling of underground coal mine fire zones, effectively reducing the risk of fire zone accidents and ensuring safe coal mine production and the safety of personnel. Attached Figure Description

[0016] Figure 1 A schematic diagram of the land-based letter ship provided in this application embodiment. Figure 2 A schematic diagram of the ship's hull provided for an embodiment of this application. Figure 3 A schematic diagram of the miniature gas chromatography analysis module provided in the embodiments of this application. Figure 4 A schematic diagram of the deck provided for an embodiment of this application. Figure 5 A schematic diagram of a drone provided for an embodiment of this application. In the diagram: 1. Communication system; 101. Intelligent winch device; 101a. Support rod; 101b. Rotating mechanism; 101c. Winch; 102. First optical fiber; 2. Hull; 201. Tracked walking device; 201a. Cable interface; 201b. Detachable battery pack; 3. Deck; 301. Parking platform; 301a. Charging module; 301b. Miniature gas chromatography analysis module; 301c. Fiber optic intelligent scaling module; 301d. Second optical fiber; 4. Several sub-ship UAVs; 401. Sub-ship UAV gas analysis module; 401b. Wireless relay module. Detailed Implementation

[0017] 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, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] Reference Figure 1The land-based mothership consists of a communication system (1), a hull (2), a deck (3), and several sub-ship UAVs (4). The communication system (1) is connected to the hull (2) and the UAVs (4) via optical fibers (102, 301d). The communication system (1) includes an intelligent winch device (101), which is used to dynamically adjust the length of the optical fiber. The intelligent winch device includes a support rod (101a), a rotating mechanism (101b), and a winch (101c). The winch (101c) is driven by a servo motor and automatically winds up and winds up the optical fiber according to the positioning data of the UAVs (4). The optical fiber includes a first optical fiber (102) and a second optical fiber (301d).

[0019] Reference Figure 3 The communication system (1) includes an intelligent winch device (101) and a dual-mode communication link; the support rod (101a) is welded with Q345B steel, with a bending strength ≥470MPa, and the bottom is fixed to the hull (2) by a flange; the rotating mechanism (101b) has a built-in harmonic reducer, supports 360° horizontal rotation, and sets a reasonable rotation angular velocity; the winch (101c) is driven by a servo motor, and winds the first optical fiber (102) and the second optical fiber (103), with an optical fiber diameter of 0.5mm, a tensile strength ≥1000N, a maximum telescopic distance of 500 meters, and an error of ±0.5 meters controlled by encoder feedback.

[0020] Reference Figure 2 The hull (2) includes a walking device (201) and a dual-mode power supply device (201a, 201b). The hull (2) adopts a tracked walking device (201) with a track width of 300mm. The ground pressure is monitored in real time by a ground pressure sensor to ensure it is ≤0.1MPa. The maximum climbing angle is 30°, which is suitable for rugged terrain with a slope of ≤25° in the well. The dual-mode power supply device has a cable interface (201a) for connecting to an external power source and a detachable battery pack (201b). The cable interface (201a) is located at the stern of the hull, with an input voltage of 380V±5%, and is connected to an external power source through an IP67 waterproof plug. The detachable battery pack (201b) has a capacity of 100kWh and an output voltage of 24V±2%. It is composed of lithium iron phosphate battery modules connected in series. Based on a rated power of 12kW and considering the actual complex conditions in the well, the endurance is calculated to be ≥4 hours. It is used to switch power supply through the internal battery when the external power supply is interrupted.

[0021] The method for switching power supply via an internal battery when external power is interrupted is as follows: The mothership status data is acquired and normalized to obtain input variables, including the fire zone proximity, cable temperature, and remaining battery power. Define fuzzy sets for the input variables and map them to fuzzy input items, and construct a fuzzy rule base; The definition of a fuzzy set is as follows: The cable temperature is classified as low, medium, and high; the proximity of the fire zone is classified as far, medium, and near; and the remaining battery power is classified as sufficient, medium, and insufficient. The construction of the fuzzy rule base specifically involves: If the "Proximity to Fire Zone" is "Near", the "Cable Temperature" is "High", and the "Battery Remaining Power" is "Sufficient", then the "Switch Mode" is "Switch Immediately". If the "proximity to fire zone" is "medium", the "cable temperature" is "medium", and the "remaining battery power" is "medium", then the "switching mode" is "pre-switching preparation". If the "proximity to fire zone" is "far", then the "switch mode" is "maintain current". For each rule, its membership degree is used as the activation strength of the rule, and the fuzzy input items and fuzzy rule base are subjected to fuzzy inference through a Mamdani-type inference mechanism to obtain fuzzy output results; The fuzzy output result is defuzzified using the centroid method to obtain the switching control signal. The dual-mode power supply device is switched to removable battery pack power supply according to the switching control signal.

[0022] It should be noted that the fuzzy rule base was constructed based on the safety analysis of cable power supply and energy security requirements when the ship enters the fire zone, and was obtained by extracting from expert experience or training based on historical event data.

[0023] Reference Figure 4 The deck (3) is 3m×2m in size and is welded from high-temperature resistant steel plates. Six UAV parking platforms (301) are evenly distributed on its surface, and each platform is equipped with a magnetic locking device. The deck (3) is located on the upper part of the hull (2). The deck (3) includes a micro gas chromatography analysis module (103) installed inside and a UAV parking platform (301) installed on the upper part. The UAV parking platform (301) includes a charging module (301a) and a fiber optic intelligent scaling module (301c).

[0024] Each parking platform (301) includes a charging module (301a) and an optical fiber intelligent scaling module (301c). The charging module (301a) includes magnetic contact with a contact resistance of ≤0.1Ω and a charging current of 30A. Based on the sub-ship UAV with a battery capacity of 6kWh, the single charging time is calculated to be ≤15 minutes. The micro gas chromatography analysis module (301b) is calibrated with standard gas to achieve a detection accuracy of ±0.1% and an extremely short response time. The optical fiber intelligent scaling module (301c) consists of a servo motor and a drive winch with a telescoping rate of 0.5m / s and a maximum optical fiber weight of 10kg. The length of the optical fiber is dynamically adjusted based on the real-time coordinates of the sub-ship UAV (4) (updated once per second).

[0025] The sub-ship UAV (4) includes a gas collection device (401) and is connected to the ship (2) via the fiber optic intelligent scaling module (301c). It is equipped with a gas sensor for real-time analysis of the collected gas and transmits the analysis results back to the mother ship via a fiber optic link or carries the gas sample back to the docking platform (301). The explosion-proof shell of the sub-ship UAV (4) conforms to the GB3836-2010 standard, has a built-in high-temperature resistant chip, and is equipped with multiple gas sensors (CO, CH4, O2 detection accuracy ±0.1%). The process of performing real-time analysis of the collected gas and transmitting the analysis results back to the mothership via fiber optic link specifically includes: The collected gas is introduced into a gas analysis module (401), which includes a spectral analysis unit and an electrochemical sensing unit; The sub-ship UAV is equipped with a gas sampling device. It initiates a gas sampling mission within a preset cruise path or designated area, introducing the ambient gas to be tested into a sampling chamber via a miniature sampling pump. This sampling chamber has a sealed structure and is equipped with a constant temperature and humidity control unit to ensure the stability of the gas composition. The spectral analysis unit includes a miniature spectrometer, a light source module, and a sample gas channel. The light source module projects a highly stable excitation beam into the gas analysis channel. The light source module includes: in absorption spectroscopy analysis scenarios, a broadband continuous spectrum light source or a tunable laser is used to emit incident light covering a specific wavelength band; in Raman scattering analysis scenarios, a narrow linewidth laser (preferably 532nm or 785nm) is used to output a highly coherent laser beam to excite the target molecules to undergo inelastic scattering. The excitation beam is collimated into a parallel beam by a collimating lens system before entering the sample gas channel and interacting with the gas molecules. The micro spectrometer receives the transmitted (absorption spectrum) or scattered (Raman spectrum) spectral signals, preprocesses the spectral signals to obtain a spectral response vector, which contains the characteristic band position information, light intensity distribution information and their variation trend of each gas component; the spectral response vector is then subjected to pattern matching analysis with a preset spectral feature database, and the least squares matching method is used to complete gas identification and component labeling to obtain spectral response features.

[0026] The gas is analyzed by absorption spectrum or Raman scattering using a spectral analysis unit to obtain the spectral response characteristics of the gas components. The gas is quantitatively detected by an electrochemical sensing unit, and current-concentration mapping data is output. Based on the Bayesian weighted average method, spectral response characteristics and current-concentration mapping data are cross-fused to obtain a high-confidence gas feature vector data package. The gas feature vector data packet is connected to the optical fiber communication link through the optical fiber intelligent scaling module and transmitted back to the mother ship's main control system in real time through the second optical fiber.

[0027] It should be noted that the electrochemical sensing unit includes multiple target gas electrode sensors, a temperature compensation module, and a potentiometer, used for quantitative detection of gases such as CO, O3, and H2S.

[0028] It should be noted that the high-confidence gas feature vector data package refers to a multi-dimensional quantitative description vector formed by Bayesian weighted fusion of spectral response features and electrochemical detection signals. It is used to accurately characterize the component features, concentration estimates and confidence weights of a specific gas or gas mixture, and has real-time, reliable and compressed transmission capabilities.

[0029] Reference Figure 5Communication link switching logic: Under normal conditions, the sub-ship UAV (4) transmits data to the ship (2) in real time through optical fiber, with a packet loss rate of ≤0.5%; when the optical fiber is accidentally tangled or broken, the sub-ship UAV (4) actively disconnects the optical fiber and immediately starts the wireless relay module (401b). The sub-ship UAV (4) also includes a wireless relay module (401b), specifically: Real-time monitoring of fiber optic link status, including signal strength, delay, and link integrity; If a link anomaly is detected, the wireless repeater module automatically determines it to be in a fault state and actively performs a physical disconnection of the fiber optic link. The detected link anomaly specifically includes: The received signal power is lower than a preset first threshold and continues for more than a set time window; The packet loss rate exceeds the preset second threshold, and this continues for more than N times; Data packet response timeout, meaning that no ACK response has been received for more than the set period; The link status monitoring register keeps returning illegal status codes or CRC check errors. After the link is lost, the module enters emergency mode, scans and identifies the nearest sub-ship UAV as a relay target, and selects a node based on signal strength and synchronization information; Establish a stable emergency communication link with the selected sub-ship UAV using a dynamic routing protocol; The analysis results are packaged and labeled, and then transmitted back to the sub-ship UAV through the established link, from where they are forwarded to the mother ship's main control system.

[0030] The process of returning the carried gas sample to the parking platform specifically involves: When the sub-ship UAV detects in real time that its battery level is below a preset threshold of 20%, the return-to-home mechanism is triggered; Collect 3D obstacle information of the current environment, construct a local obstacle grid map, and use an improved A* algorithm combined with a dynamic window algorithm to plan obstacle avoidance path and generate the optimal return path. The sub-ship UAV is controlled to fly along the planned path, and the lidar perception data is refreshed in real time during the flight, and the path is dynamically corrected. Once the sub-ship UAV flies to the parking platform, it controls the magnetic contacts on its bottom to precisely dock with the charging module on the parking platform; After the charging module docking is completed, the platform’s built-in micro gas chromatography analysis module is triggered to perform secondary analysis on the gas samples collected by the sub-ship UAV. The secondary analysis results are compared with the initial analysis results during the flight mission for redundancy detection, and a system warning signal is issued based on the detection results.

[0031] Furthermore, the process of performing redundancy detection between the secondary analysis results and the initial analysis results during the flight mission, and issuing a system warning signal based on the detection results, specifically involves: Obtain the chromatogram dataset after the first analysis and the secondary chromatogram dataset after the second analysis, and perform normalization processing; The difference between the processed chromatographic dataset and the secondary chromatographic dataset was calculated using Kullback-Leibler divergence. The specific formula for calculating the difference in the sample spectra is as follows:

[0032] In the formula, For the spectral differences of the samples, This is a normalized vector of the chromatogram dataset after the initial analysis. This is the normalized vector of the secondary analysis chromatogram. For vector dimensions, For the first Normalized intensity values ​​of the first sample in each spectral band. For the first Normalized intensity values ​​of secondary samples in each spectral band; The sample spectrum difference is compared with a preset redundancy tolerance threshold. If the sample spectrum difference is greater than the preset redundancy tolerance threshold, it is determined that the current secondary gas sample and the sample collected in the first flight mission have significant differences in chromatographic characteristics. A sample deviation risk marker is generated and entered into the system alarm queue, and the following information is output: alarm timestamp, sample number, first and second spectrum summary, difference value, deviation level label. At the same time, the current alarm record is pushed to the platform data chain management module to start the subsequent anomaly verification and sub-ship UAV resampling mechanism or path tracing task.

[0033] It should be noted that KL divergence measures the "loss" or "gain" of information between the first and second detection samples, reflecting the distribution changes of gas components over retention time. The spectral difference calculation method based on KL divergence has the systematic advantages of accurate quantification, robust processing, intelligent alarm, and scalability, which significantly enhances the data reliability judgment and intelligent alarm capabilities of the sub-ship UAV in the process of performing complex tasks.

[0034] Furthermore, the mothership, through comprehensive analysis of data from the drones, locates the source of the fire and predicts the risk level of an explosion.

[0035] In this embodiment, the monitoring and feedback module carried by the mother ship can integrate sample data collected from the fire zone in real time, generate an explosion risk level based on the mining explosion triangle theory through edge computing, and synchronize it to the ground control center. The multi-unit network collaborative operation covers an area three times that of traditional methods, with high fire source positioning accuracy and small error in delineating enclosed areas. It solves the pain points of traditional technologies such as large blind spots, communication lag, and limited data, providing full-process, high-precision intelligent support for coal mine fire zone management.

[0036] The mothership, through comprehensive analysis of data from unmanned aerial vehicles (UAVs), located the source of the fire and predicted the risk level of the explosion. Specifically: The data transmitted back from the UAV is acquired, and a multi-scale high-temperature anomaly extraction and spatial density clustering algorithm is used to perform fire source clustering analysis to obtain a set of candidate fire source points and their spatial location coordinates. The UAV constructs an explosion risk assessment function model in the edge computing node. The transmitted data includes gas concentration data, gas component spectrum heterogeneity data, and ignition source intensity. The risk value is calculated based on the explosion risk assessment function model, and then compared with the preset explosion threshold. Based on the comparison results, a dynamic hierarchical threshold range is adaptively generated using Bayesian update and sliding window methods. The current risk value is mapped to the corresponding risk level based on the threshold range; Level 1 Risk (High Risk): ; Level 2 Risk (Medium): ; Level 3 risk (controllable): ; Level 4 Risk (Safety): .

[0037] in Furthermore, it adaptively updates based on the latest edge node data.

[0038] Obtain the spatial coordinates and time labels of the work area, and combine them with the risk level to perform local area risk clustering analysis using a lightweight clustering algorithm. Output the spatiotemporal evolution trend map of the risk level and the coordinates of the center point of the spatiotemporal cluster with high explosion incidence. Based on the spatiotemporal evolution trend map and the coordinates of the center point of the spatiotemporal cluster with high explosion incidence, executable control recommendations and operating procedures are generated for different levels, including: Automatically issue mandatory power outage and evacuation orders to Level 1 risk areas; The automatic ventilation and risk reassessment mechanism is triggered for Level 2 risk areas; Level 3 risk areas are marked as areas under continuous observation; Normal operating procedures have been resumed for Level 4 risk areas.

[0039] The specific calculation formula for the gas component spectral isomer data is as follows:

[0040] In the formula, This is gas component spectral isomer data. For the first Measured concentrations of the gaseous components, This is the reference concentration of the gaseous component under standard safety conditions. This represents the characteristic ratio of the component to the reaction-related gases. This is the normal reference value for the characteristic ratio. This represents the rate of change of the gas component within the time window. This represents the highest rate of change for this component in history. This can be used as an indicator of spectral shift or heterogeneity. This represents the maximum permissible range for the gas spectrum shift or structural drift.

[0041] The specific calculation formula for the explosion risk assessment function model is as follows:

[0042] In the formula, This is the risk value. For gas concentration data, For ignition source intensity, This is gas component spectral isomer data. , , The risk coefficient is obtained by training the weights based on the explosive triangle theory.

[0043] It should be noted that gas component spectral isomer data refers to a set of gas chemical characteristic data formed by high-precision qualitative and quantitative detection of various components in a gas mixture. Its core is: not only to detect the concentration of conventional gases (such as CH4, O2, CO, etc.), but also to identify the molecular structural isomerism, spectroscopic characteristics, ratio relationships and their changing trends of specific components, thereby revealing potential hidden danger mechanisms such as thermal reactions, spontaneous combustion, mutations, and leaks.

[0044] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0045] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0046] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0047] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0049] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A land-based mother-daughter ship detection device for underground fire zones based on intelligent fiber optic communication and explosion prediction, comprising a communication system, a hull, a deck, and a daughter ship UAV, characterized in that: The communication system is connected to the ship's hull and UAV via optical fiber. The communication system includes an intelligent winch device, which is used to dynamically adjust the length of the optical fiber. The hull includes a walking device and a dual-mode power supply device. The dual-mode power supply device has a cable interface for connecting to an external power source and a detachable battery pack, which is used to switch power supply through the internal battery when the external power supply is interrupted. The deck is located on the upper part of the ship's hull. The deck includes a micro gas chromatography analysis module installed inside and a drone parking platform installed on the upper part. The drone parking platform includes a charging module and a fiber optic intelligent scaling module. The sub-ship UAV includes a gas collection device and is connected to the ship via the fiber optic intelligent scaling module. It is also equipped with a gas sensor for real-time analysis of the collected gas and to transmit the analysis results back to the mother ship via fiber optic link or to carry the gas sample back to the docking platform. The mothership uses a comprehensive analysis of data from drones to locate the source of the fire and predict the risk level of an explosion.

2. The underground fire zone land-based mother-daughter ship detection device based on intelligent optical fiber communication and explosion prediction according to claim 1, characterized in that, The intelligent winch device includes a support rod, a rotating mechanism, and a winch. The winch is driven by a servo motor and automatically winds up and winds up optical fibers based on the positioning data of the UAV. The optical fibers include a first optical fiber and a second optical fiber.

3. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 1, characterized in that, The method for switching power supply via an internal battery when external power is interrupted is as follows: The mothership status data is acquired and normalized to obtain input variables, including the fire zone proximity, cable temperature, and remaining battery power. Define fuzzy sets for the input variables and map them to fuzzy input items, and construct a fuzzy rule base; For each rule, its membership degree is used as the activation strength of the rule, and the fuzzy input items and fuzzy rule base are subjected to fuzzy inference through a Mamdani-type inference mechanism to obtain fuzzy output results; The fuzzy output result is defuzzified using the centroid method to obtain the switching control signal. The dual-mode power supply device is switched to removable battery pack power supply according to the switching control signal.

4. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 1, characterized in that, The process of performing real-time analysis of the collected gas and transmitting the analysis results back to the mothership via fiber optic link specifically includes: The collected gas is introduced into a gas analysis module, which includes a spectral analysis unit and an electrochemical sensing unit. The gas is analyzed by absorption spectrum or Raman scattering using a spectral analysis unit to obtain the spectral response characteristics of the gas components. The gas is quantitatively detected by an electrochemical sensing unit, and current-concentration mapping data is output. Based on the Bayesian weighted average method, spectral response characteristics and current-concentration mapping data are cross-fused to obtain a high-confidence gas feature vector data package. The gas feature vector data packet is connected to the optical fiber communication link through the optical fiber intelligent scaling module and transmitted back to the mother ship's main control system in real time through the second optical fiber.

5. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 1, characterized in that, The sub-ship UAV also includes a wireless relay module, specifically: Real-time monitoring of fiber optic link status, including signal strength, delay, and link integrity; If a link anomaly is detected, the wireless repeater module automatically determines it to be in a fault state and actively performs a physical disconnection of the fiber optic link. After the link is lost, the module enters emergency mode, scans and identifies the nearest sub-ship UAV as a relay target, and selects a node based on signal strength and synchronization information; An emergency communication link is established with the selected sub-ship UAV using a dynamic routing protocol; The analysis results are packaged and labeled, and then transmitted back to the sub-ship UAV through the established link, from where they are forwarded to the mother ship's main control system.

6. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 5, characterized in that, The process of returning the carried gas sample to the parking platform specifically involves: When the drone detected that its battery level was below a preset threshold of 20% in real time, the return-to-home mechanism was triggered. Collect 3D obstacle information of the current environment to construct a local obstacle grid map, and use an improved A* algorithm combined with a dynamic window algorithm to plan obstacle avoidance path and generate the optimal return path. The sub-ship UAV is controlled to fly along the planned path, and the lidar perception data is refreshed in real time during the flight, and the path is dynamically corrected. Once the sub-ship UAV flies to the parking platform, it controls the magnetic contacts on its bottom to precisely dock with the charging module on the parking platform; After the charging module docking is completed, the platform’s built-in micro gas chromatography analysis module is triggered to perform secondary analysis on the gas samples collected by the sub-ship UAV. The secondary analysis results are compared with the initial analysis results during the flight mission for redundancy detection, and a system warning signal is issued based on the detection results.

7. The underground fire zone land-based mother-daughter ship detection device based on intelligent optical fiber communication and explosion prediction according to claim 6, characterized in that, The process of performing redundancy detection between the secondary analysis results and the initial analysis results during the flight mission, and issuing a system warning signal based on the detection results, specifically involves: Obtain the chromatogram dataset after the first analysis and the secondary chromatogram dataset after the second analysis, and perform normalization processing; The difference between the processed chromatographic dataset and the secondary chromatographic dataset was calculated using Kullback-Leibler divergence. The sample spectrum difference is compared with a preset redundancy tolerance threshold. If the sample spectrum difference is greater than the preset redundancy tolerance threshold, it is determined that the current secondary gas sample and the sample collected in the first flight mission have significant differences in chromatographic characteristics. A sample deviation risk marker is generated and entered into the system alarm queue.

8. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 1, characterized in that, The mothership, through comprehensive analysis of data from unmanned aerial vehicles (UAVs), located the source of the fire and predicted the risk level of the explosion. Specifically: Data transmitted back from the UAV was acquired, and fire source clustering analysis was performed using multi-scale high-temperature anomaly extraction and spatial density clustering algorithms to obtain a set of candidate fire source points and their spatial coordinates. An explosion risk assessment function model was then constructed. The risk value is calculated based on the explosion risk assessment function model, and then compared with the preset explosion threshold. Based on the comparison results, a dynamic hierarchical threshold range is adaptively generated using Bayesian update and sliding window methods. The current risk value is mapped to the corresponding risk level based on the threshold range; Obtain the spatial coordinates and time labels of the work area, and combine them with the risk level to perform local area risk clustering analysis using a lightweight clustering algorithm. Output the spatiotemporal evolution trend map of the risk level and the coordinates of the center point of the spatiotemporal cluster with high explosion incidence. Based on the spatiotemporal evolution trend map and the coordinates of the center point of the spatiotemporal cluster with high explosion incidence, executable control suggestions and operating procedures are generated for different levels.

9. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 8, characterized in that, The specific calculation formula for the explosion risk assessment function model is as follows: In the formula, This is the risk value. For gas concentration data, For ignition source intensity, This is gas component spectral isomer data. , , This represents the risk coefficient.

10. The underground fire zone land-based mothership detection device based on intelligent optical fiber communication and explosion prediction according to claim 7, characterized in that, This includes the sample spectral difference, the specific calculation formula of which is as follows: In the formula, For the spectral differences of the samples, This is a normalized vector of the chromatogram dataset after the initial analysis. This is the normalized vector of the secondary analysis chromatogram. For vector dimensions, For the first Normalized intensity values ​​of the first sample in each spectral band. For the first Normalized intensity values ​​of secondary samples in each spectral band.

Images