Gas inspection robot suitable for complex environment of coal mine and inspection method of gas inspection robot

Abstract

The invention relates to a gas inspection robot suitable for a complex coal mine environment and an inspection method thereof, and belongs to the technical field of coal mine safety monitoring. The robot comprises a multi-parameter sensor module which is used for synchronously detecting parameters in a working environment; the self-adaptive vibration isolation system is connected between the multi-parameter sensor module and the robot body and is used for suppressing the influence of vibration on the detection precision of the sensor in the moving process and providing a stable platform during fixed-point detection; the intelligent power supply management unit comprises a main power supply module and a self-power-generation standby module, and the intelligent power supply management unit is used for automatically switching a power supply mode from the main power supply module to the self-power-generation standby module when detecting that the gas concentration exceeds the limit; and the central control unit is electrically connected to the hanging rail type moving mechanism, the multi-parameter sensor module, the self-adaptive vibration isolation system and the intelligent power supply management unit, and is used for controlling the inspection operation of the robot and carrying out fusion processing and intelligent decision making on the collected multi-parameter data.

Description

Technical Field

[0001] This invention belongs to the field of coal mine safety monitoring technology, and relates to a gas inspection robot and its inspection method suitable for complex coal mine environments. Background Technology

[0002] Coal mine underground operations present a complex environment with various safety risks. Among these, methane gas (whose main component is methane) is the number one hazard threatening safe production in coal mines. When the concentration of methane gas in the roadway air reaches the explosive limit and encounters an ignition source, a catastrophic methane explosion will occur. Therefore, to prevent accidents, it is essential to conduct timely, accurate, and continuous monitoring of key environmental parameters in the underground environment, such as methane concentration, carbon monoxide, carbon dioxide, oxygen concentration, and temperature.

[0003] To meet this safety requirement, the coal mining industry currently mainly uses two monitoring methods: fixed sensors and manual inspections. Fixed sensors are typically installed at key locations such as coal faces and return airways, enabling real-time monitoring at fixed points. However, their deployment costs are high, their coverage is limited, and a large number of monitoring blind spots inevitably exist in the vast network of roadways. Manual inspections, on the other hand, involve gas inspectors carrying portable detectors for periodic patrols. While this method offers greater flexibility in coverage, it suffers from high labor intensity and low inspection efficiency. More importantly, this method directly exposes inspectors to potentially hazardous environments; in the event of a gas outburst or explosion, inspectors face extremely high risks to their lives.

[0004] To overcome the drawbacks of traditional methods, inspection robots have been introduced into the field of coal mine safety monitoring as an emerging technology in recent years. Inspection robots aim to replace manual inspections in an automated and intelligent manner, thereby improving monitoring efficiency and safety. However, in practical underground applications, existing inspection robots still face a series of unresolved technical challenges, specifically in the following aspects: (1) Problems with motion detection accuracy: The underground environment of coal mines is harsh, and the tracks or roads used for robot movement are often uneven. The mechanical vibrations generated by the robot during movement are directly transmitted to the high-precision sensors on board, which seriously interferes with their normal readings, resulting in inaccurate data collection and failure to accurately reflect the environmental conditions.

[0005] (2) Multi-parameter interference problem: In order to achieve comprehensive monitoring, inspection robots usually need to integrate multiple gas sensors. In the compact integration space, if the sampling gas path is not designed properly, cross-interference can easily occur between different gases, which will affect the detection accuracy of each parameter.

[0006] (3) Environmental adaptability and reliability issues: The underground environment is generally characterized by high dust, high humidity, and strong electromagnetic interference, which poses a severe challenge to the long-term stable operation of the robot's electronic components and mechanical structure. A particularly fatal flaw is that when the gas concentration exceeds the limit, the power must be cut off in the area according to safety regulations. At this time, the traditional inspection robot, which relies on external power supply, will immediately lose power and monitoring capabilities, creating a fatal monitoring "window" at the most dangerous moment when real-time monitoring is most needed.

[0007] (3) Insufficient intelligence: Most inspection robots currently only collect and transmit data, and their functions are relatively simple. They generally lack the ability to conduct in-depth analysis, trend prediction, and intelligent decision-making on the collected multidimensional data, making it difficult to achieve accurate early warning of potential risks.

[0008] (4) Explosion-proof safety requirements: As equipment that operates in an explosive gas environment, the entire inspection robot (including its circuits, motors, power supply system, etc.) must meet explosion-proof standards. This brings extremely high technical difficulty and cost to the overall design of the robot, especially the complex internal circuits and power supply system.

[0009] In summary, existing technologies, whether traditional monitoring methods or rudimentary inspection robots, all have significant limitations. Therefore, there is an urgent need in this field to develop a technology that can overcome these technical challenges. Summary of the Invention

[0010] In view of this, the purpose of this invention is to provide a gas inspection robot and its inspection method suitable for the complex environment of coal mines, solving the problems of low detection accuracy, susceptibility to environmental interference, inability to work continuously under extreme conditions such as power outages, and low level of intelligence in existing inspection robots. This robot integrates high-precision detection, high reliability, high intelligence, and high safety, thereby improving the level of coal mine safety monitoring.

[0011] To achieve the above objectives, the present invention provides the following technical solution: Option 1: A gas inspection robot suitable for the complex environment of coal mines, comprising: The robot itself; A rail-mounted mobile mechanism, installed on the robot body, is used to drive the robot body to move along a preset track; A multi-parameter sensor module is installed on the robot body to synchronously detect parameters such as methane, carbon monoxide, carbon dioxide, oxygen concentration, and temperature in the working environment. An adaptive vibration isolation system is connected between the multi-parameter sensor module and the robot body to suppress the impact of vibration on the sensor detection accuracy during movement and to provide a stable platform during fixed-point detection. The intelligent power management unit includes a main power supply module and a self-generated backup module. The intelligent power management unit is used to automatically switch the power supply mode from the main power supply module to the self-generated backup module when the gas concentration exceeds the limit. The central control unit is electrically connected to the rail-mounted mobile mechanism, the multi-parameter sensor module, the adaptive vibration isolation system, and the intelligent power management unit. It is used to control the robot's inspection operations and to perform fusion processing and intelligent decision-making on the collected multi-parameter data.

[0012] Preferably, the adaptive vibration isolation system is a three-stage vibration reduction structure, comprising: A primary spring damping mechanism connected to the robot body is used to absorb high-frequency vibrations; A secondary magnetohydrodynamic damper connected to the primary spring damping mechanism is used to adaptively adjust the damping coefficient according to the vibration frequency to absorb low and medium frequency vibrations. The three-level MEMS active vibration isolation platform, which carries the multi-parameter sensor module, is used to perform high-precision attitude compensation through microelectromechanical systems to isolate minute vibrations.

[0013] Preferably, the multi-parameter sensor module has a gas flow channel separation structure inside to physically isolate the sampling gas paths of different gas sensors in order to avoid cross-interference between sensors.

[0014] Preferably, the central control unit includes an edge computing module and a UWB positioning module; The edge computing module is used to run multi-source information fusion algorithms and machine learning calibration algorithms, analyze and process sensor data in real time, intelligently compensate for detection results by combining environmental factors such as temperature and humidity, and combine with the location information provided by the UWB positioning module to realize intelligent inspection decision-making and accurate early warning of abnormal situations. The UWB positioning module is used to provide the robot with high-precision real-time location information, assisting in inspection path planning and data spatiotemporal stamping.

[0015] Preferably, the switching logic of the intelligent power management unit is as follows: when the central control unit determines that the methane concentration value exceeds the preset safety threshold, a switching command is triggered, and the switching from the main power supply module to the self-generated backup module is completed within 2 seconds to ensure the continuity of the monitoring function.

[0016] Preferably, the robot also includes a remote air extraction pipeline interface for connecting to a fixed air extraction pipeline in the mine to assist in monitoring the gas emission in a specific area.

[0017] Option 2: An inspection method for a gas inspection robot suitable for complex coal mine environments, comprising the following steps: S1: Under the command of the central control unit, the robot moves and inspects along a preset track via a rail-mounted mobile mechanism; S2: During the movement, the adaptive vibration isolation system monitors the vibration in real time and actively reduces the vibration to ensure the detection stability of the multi-parameter sensor module; S3: The multi-parameter sensor module synchronously collects data on methane, carbon monoxide, carbon dioxide, oxygen concentration, and temperature, and avoids data cross-interference through a gas flow channel separation structure; S4: The central control unit uses multi-source information fusion algorithms and machine learning calibration algorithms to perform real-time fusion and analysis of the collected multi-source data. It dynamically calibrates and compensates sensor readings based on environmental parameters such as temperature and dust, and generates inspection logs with spatiotemporal stamps by combining UWB positioning information. S5: The central control unit continuously monitors the gas concentration. When the concentration exceeds the preset threshold, it immediately controls the intelligent power management unit to switch to the self-generating backup module and issues an early warning message.

[0018] Furthermore, in step S4, the central control unit uses multi-source information fusion algorithms and machine learning calibration algorithms to perform real-time fusion and analysis of the collected multi-source data, specifically including the following steps: S41: Data preprocessing and alignment: Clean the collected data from various sensors and convert it into a data stream with a uniform format and synchronized timestamps; Data cleaning involves using Kalman filtering or moving average filtering to perform preliminary smoothing on the time series data of each individual sensor, filtering out obvious abnormal peaks and extracting more stable basic signals. Format unification involves normalization or standardization to map data from different sensors to a unified numerical range. Timestamp synchronization: A high-precision clock source is integrated within the central control unit. When the robot is running, all data acquisition events—whether it's a reading from a gas sensor or a position update from the UWB positioning module—are forcibly assigned a unified timestamp accurate to the millisecond level. This is crucial for achieving data alignment in the time dimension.

[0019] S42: State estimation and spatiotemporal correlation: By using advanced algorithms, isolated data points are fused into an organically related whole to achieve accurate estimation of the environmental state; (1) Constructing a Spatiotemporal Data Object (SDO): After processing in step S41, the data of each timestamp is packaged into a structured "spatiotemporal data object", so that each data has attributes in four dimensions: time, space, environment and its own state; (2) Environmental interference compensation based on machine learning: using a pre-trained gradient boosting decision tree (GBDT) model or a small neural network model, learn and fit the complex nonlinear relationship between human-induced changes in temperature, vibration and other interference factors and sensor reading errors. (3) Constructing a dynamic spatiotemporal heat map: The robot continuously generates spatiotemporal data objects (SDOs) and updates them in real time to a dynamic three-dimensional grid map; this map virtualizes the mine tunnels into small cubic grids; Each grid cell stores the latest compensated environmental parameter values ​​for that spatial location. Furthermore, spatial interpolation algorithms such as Kriging interpolation can be used to estimate the values ​​of neighboring grid cells that were not directly measured, thus forming a continuous and smooth environmental situation map. S43: Situational Awareness and Decision Output: Based on the accurate environmental situation map generated in step S42, perform intelligent analysis, risk assessment, and make a final decision.

[0020] Furthermore, the method also includes a fixed-point detection mode: when the robot moves to a preset key detection point, the central control unit controls the magnetohydrodynamic damper to increase damping, locks the MEMS active vibration isolation platform, and switches the adaptive vibration isolation system to a high stability state to perform high-precision fixed-point detection.

[0021] The beneficial effects of this invention are as follows: (1) This scheme effectively solves the vibration interference problem in the moving state through a three-level adaptive vibration isolation system, and greatly improves the detection accuracy. The gas flow channel separation technology adopted avoids mutual interference of multi-parameter detection and ensures the reliability of data.

[0022] (2) The unique intelligent power management unit and self-generating system of this solution ensure that the robot can continue to work even in the most dangerous situations such as gas exceeding the limit and power outage, which greatly improves the inherent safety level of coal mines.

[0023] (3) This solution combines edge computing, machine learning and UWB positioning technology to achieve a leap from simple data collection to intelligent analysis, decision-making and early warning, and can detect security risks earlier and more accurately.

[0024] (4) This solution uses a rail-mounted robot to replace manual inspection, which fundamentally eliminates the safety risks of manual inspection, and at the same time realizes all-weather, full-coverage automated inspection, which is far more efficient than manual inspection.

[0025] (5) The overall design of the robot in this scheme strictly follows the coal mine safety regulations. The power switching and self-generating system are both designed to be explosion-proof, ensuring that they will not become the ignition source of gas under any working conditions.

[0026] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0027] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is an overall structural block diagram of the gas inspection robot in an embodiment of the present invention; Figure 2 This is a schematic diagram of the adaptive vibration isolation system in an embodiment of the present invention; Figure 3 This is a flowchart of the gas inspection method of the gas inspection robot in this embodiment of the invention.

[0028] Reference numerals: 1-Robot body, 2-Rail-mounted mobile mechanism, 3-Multi-parameter sensor module, 4-Adaptive vibration isolation system, 401-Spring vibration damping mechanism, 402-Magnetofluid damper, 403-MEMS active vibration damping platform, 5-Intelligent power management unit, 6-Central control unit, 7-Wireless communication module, 8-UWB positioning module. Detailed Implementation

[0029] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0030] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0031] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0032] Please see Figure 1 This invention provides a gas inspection robot suitable for the complex environment of coal mines, comprising a robot body 1, a rail-mounted mobile mechanism 2, a multi-parameter sensor module 3, an adaptive vibration isolation system 4, an intelligent power management unit 5, and a central control unit 6. The robot is also equipped with a wireless communication module 7 and a UWB positioning module 8.

[0033] The rail-mounted mobile mechanism 2 is installed on the top of the robot body 100 and consists of an explosion-proof motor, a transmission device and guide wheels, enabling it to move stably along the track laid on the top or side wall of the mine roadway.

[0034] The multi-parameter sensor module 3 is an integrated detection unit that internally includes at least a methane sensor, a carbon monoxide sensor, a carbon dioxide sensor, an oxygen sensor, and a temperature sensor. To avoid cross-interference, the module is designed with an independent miniature air pump and separate sampling channels to ensure the purity of each gas sample. Furthermore, this module can also integrate a TMR magnetic sensing chip to monitor the current status of the robot itself and surrounding equipment, serving as an auxiliary data source for equipment health diagnostics.

[0035] Please combine Figure 2 The adaptive vibration isolation system 4 is crucial for ensuring detection accuracy; it connects the robot body 1 and the multi-parameter sensor module 3. This system consists of a three-stage vibration reduction structure: The primary spring damping mechanism 401, as the basic damping layer, mainly absorbs high-frequency, small-amplitude vibrations caused by track joints, etc.

[0036] Secondary magnetohydrodynamic damper 402: The central control unit 6 adjusts the magnetic field strength applied to the magnetohydrodynamic fluid in real time based on vibration information fed back from the accelerometer, thereby changing its damping coefficient. This enables the robot to adaptively absorb various low- and medium-frequency vibrations. During fixed-point detection, it can be set to the maximum damping state to achieve a quasi-rigid connection and provide a stable platform.

[0037] The Level 3 MEMS active vibration damping platform 403 is the highest precision damping layer, directly supporting the multi-sensor module 3. The platform senses minute vibrations through built-in micro gyroscopes and accelerometers, and drives micro actuators to perform reverse compensation, achieving complete isolation of residual micro-vibrations.

[0038] The intelligent power management unit 5 is responsible for powering the entire robot system. It includes a main power supply module connected to an external power source (such as track power) and an independent self-generated backup module (such as a built-in high-energy explosion-proof lithium battery pack or a triboelectric generator). The central control unit 6 continuously monitors the methane concentration. When the concentration exceeds the nationally stipulated 0.5% (or a lower threshold defined by the mine), the central control unit 6 immediately issues a switching command to the power management unit 5. The power management unit 5 cuts off the main power supply within 2 seconds and smoothly switches to the self-generated backup module, while simultaneously sending the highest-level gas over-limit and power failure alarm to the ground control center via the wireless communication module 7.

[0039] The central control unit 6 employs a high-performance embedded processor with powerful edge computing capabilities. It runs a complex multi-source information fusion algorithm. This algorithm not only integrates five environmental parameters but also combines high-precision location information (accuracy down to the centimeter level) and timestamps provided by the UWB positioning module 8 to generate a three-dimensional environmental parameter cloud map. Furthermore, the machine learning model embedded in the unit can perform real-time calibration and compensation of the sensor's raw readings based on historical data and current environmental parameters (such as temperature, humidity, and air pressure), greatly improving the accuracy of detection.

[0040] Specific implementation of the multi-source information fusion algorithm in the central control unit: The multi-source information fusion algorithm used in the central control unit is a hierarchical and progressive intelligent processing framework. Its core objective is to transform raw data from different sensors, different dimensions, and containing noise into accurate, reliable, and actionable information about the mine environment. The algorithm framework is mainly divided into three layers: data preprocessing and alignment layer, state estimation and spatiotemporal correlation layer, and situational awareness and decision output layer.

[0041] The first layer is the data preprocessing and alignment layer. This layer is the foundation for all subsequent advanced analyses. It is mainly responsible for cleaning the raw, heterogeneous sensor data and transforming it into a data stream with a uniform format and synchronized time.

[0042] (1) High-precision timestamp synchronization: The central control unit integrates a high-precision clock source. When the robot is running, all data acquisition events—whether it's a reading from a gas sensor or a position update from the UWB positioning module—are forcibly assigned a uniform, millisecond-accurate timestamp. This is crucial for achieving data alignment in the time dimension.

[0043] (2) Data denoising and filtering: Raw sensor data inevitably contains noise such as electromagnetic interference and transient fluctuations. The algorithm first uses a Kalman filter or a moving average filter to perform preliminary smoothing on the time series data of each individual sensor (such as methane, temperature, etc.), filtering out obvious abnormal peaks and extracting a more stable basic signal.

[0044] (3) Data standardization: Because different sensors have vastly different dimensions and numerical ranges (e.g., methane is measured in ppm, oxygen in %, and temperature in °C), all data will be normalized or standardized to map to a uniform numerical range (e.g., 0 to 1 or -1 to 1) to facilitate subsequent processing by machine learning models.

[0045] The second layer is the state estimation and spatiotemporal correlation layer. This layer is the core of information fusion. Through advanced algorithms, isolated data points are fused into an organically connected whole, thereby achieving accurate estimation of the environmental state.

[0046] (1) Constructing a Spatiotemporal Data Object (SDO): At each time step k The data processed in the first layer is integrated to form a spatiotemporal data object. :

[0047] in, yes m Normalized reading vectors of each sensor, For the first i One sensor (such as) i ∈{CH4,CO,O2, T ,...}) at time The original measurement value, The sensor after normalization i The measured value, In time step k The moment; For robots at all times The three-dimensional spatial position vector; For robots at all times The state vector, where For vibration level, For speed.

[0048] (2) Dynamic state estimation based on Kalman filtering: To obtain a smoother and more accurate state estimate, we define the system's state vector, for example, for methane concentration and its rate of change: ,in For the system at time k The true state vector (to be estimated).

[0049] Kalman filtering is performed iteratively through two steps: prediction and update. predict:

[0050]

[0051] renew:

[0052]

[0053]

[0054] in, It is the state transition matrix. It is the actual measured value (i.e.) ), and These are the covariance matrices of process noise and measurement noise, respectively. It is the observation matrix. For at any time k For the state vector The posterior estimate. The final output. The first component is the optimal estimate that combines historical information and current measurements.

[0055] (3) Environmental disturbance compensation based on machine learning: Problem: Sensor readings can be affected by temperature, humidity, air pressure, and even the robot's own vibrations. For example, an increase in temperature may cause a fixed deviation in the readings of a certain gas sensor.

[0056] Implementation: The central control unit embeds a pre-trained gradient boosting decision tree (GBDT) model or a small neural network model.

[0057] Model inputs: raw sensor readings, environmental parameters, and robot state parameters.

[0058] Model output: Compensated sensor readings that are closer to the true values.

[0059] Training process: The model was trained offline in a laboratory environment through numerous calibration experiments. With known real gas concentrations, disturbance factors such as temperature and vibration were artificially altered, allowing the model to learn and fit the complex nonlinear relationship between these disturbance factors and sensor reading errors.

[0060] This process can be viewed as a nonlinear mapping function. Its mathematical form is:

[0061] in, It is a sensor after environmental interference compensation. i The final estimate, It's a temperature reading. It refers to the vibration level. (Function) Implemented by a pre-trained machine learning model (such as GBDT or a neural network), whose parameters This correction value was obtained through offline calibration data. For simplicity, the sensor values ​​in subsequent formulas refer to this correction value.

[0062] (4) Construction of dynamic spatiotemporal heat map: The robot continuously generates SDOs and updates them in real time to a dynamic 3D grid map. This map virtualizes the mine tunnels as a series of small cubic grids.

[0063] Each grid cell stores the latest compensated environmental parameter values ​​for that spatial location. Using spatial interpolation algorithms such as Kriging interpolation, neighboring grid cells that were not directly measured can be estimated, thus forming a continuous, smooth gas concentration distribution map (or other parameter distribution map). This not only shows information at the "points" but also reveals the distribution pattern over a "surface."

[0064] The third layer is the situational awareness and decision-making output layer. This is the highest layer of the algorithm. Based on the accurate environmental situation map generated by the second layer, it performs intelligent analysis, risk assessment, and makes the final decision.

[0065] (1) Multidimensional risk factor association analysis: Define a risk assessment function This function is calculated using a set of logical rules and weights. Define the indicator function. Its value is 1 when the condition is true, and 0 otherwise.

[0066] For example, the risk of gas accumulation. It can be represented as:

[0067] in, These are the weights of different risk factors.

[0068] (2) Trend prediction: The development trend of the situation is assessed by calculating the time and spatial gradients of key parameters.

[0069] Time gradient:

[0070] Spatial gradient:

[0071] The gradient is calculated from the dynamic spatiotemporal map using numerical methods such as finite difference.

[0072] (3) Autonomous decision-making and task adjustment: Decision function: The robot's final action It is a decision function D based on risk assessment and trend analysis. The output of ).

[0073]

[0074] This function can be a piecewise function, formally represented as:

[0075] in, The decision action taken by the robot at time k; This includes power switching, highest-level alarms, etc. To send a warning to the ground station; It involves adjusting the path and encrypting the inspection. Continue with routine inspections. This is the preset threshold.

[0076] Please see Figure 3 The gas inspection method of the robot in this embodiment is as follows: Step S1: Startup and Initialization. The robot is powered on, the system performs a self-test, and receives inspection tasks from the ground control center.

[0077] Step S2: The mobile inspection robot starts the rail-mounted mobile mechanism 2 and moves along the preset path.

[0078] Step S3: Real-time vibration reduction and data acquisition. During the movement, the adaptive vibration isolation system 4 continues to operate, isolating vibrations. At the same time, the multi-parameter sensor module 3 continuously and synchronously acquires various environmental parameters.

[0079] Step S4: Data Fusion and Intelligent Analysis. The central control unit 6 processes the collected data in real time at the edge, performing fusion, positioning, and calibration, and comparing it with safety thresholds.

[0080] Step S5: Methane Concentration Determination. Determine if the current methane concentration exceeds the limit. If not, return to Step S2 to continue the inspection.

[0081] Step S6: Emergency Response. If the methane concentration exceeds the limit, immediately execute the emergency response: trigger the intelligent power management unit 5 to switch to the self-generating backup module.

[0082] Step S7: Send the highest level alarm. Send the highest level alarm, containing location, concentration, and time, to the ground station and downhole personnel (via the broadcast system) through the wireless communication module 7.

[0083] Step S8: Continuous Monitoring and Task Completion. With backup power, the robot continues to monitor in place or in a designated safe area until the danger is over or a new instruction is received, at which point the inspection task ends.

[0084] During the inspection process, if the task requires high-precision detection at a critical location (such as the return airway or sealed wall in a coal mining face), the robot will execute a fixed-point detection mode. At this time, the moving mechanism stops, and the adaptive vibration isolation system switches to a high-stability locked state to perform sampling and data analysis for a longer period of time to obtain accurate environmental parameters at that point.

[0085] In summary, through the collaborative innovation of mechanical structure, sensor technology, power management, and intelligent algorithms, this invention has constructed a robotic system capable of high-precision, high-reliability, and uninterrupted intelligent inspection in complex and even extreme environments of coal mines, providing strong technical support for safe coal mine production.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A gas inspection robot suitable for complex coal mine environments, comprising a robot body and a rail-mounted moving mechanism, wherein the rail-mounted moving mechanism is mounted on the robot body and is used to drive the robot body to move along a preset track; characterized in that, The robot also includes: A multi-parameter sensor module is installed on the robot body to simultaneously detect the concentrations of methane, carbon monoxide, carbon dioxide, oxygen, and temperature parameters in the working environment. An adaptive vibration isolation system is connected between the multi-parameter sensor module and the robot body to suppress the impact of vibration on the sensor detection accuracy during movement and to provide a stable platform during fixed-point detection. The intelligent power management unit includes a main power supply module and a self-generated backup module. The intelligent power management unit is used to automatically switch the power supply mode from the main power supply module to the self-generated backup module when the gas concentration exceeds the limit. The central control unit is electrically connected to the rail-mounted mobile mechanism, the multi-parameter sensor module, the adaptive vibration isolation system, and the intelligent power management unit. It is used to control the robot's inspection operations and to perform fusion processing and intelligent decision-making on the collected multi-parameter data.

2. The gas inspection robot according to claim 1, characterized in that, The adaptive vibration isolation system is a three-stage vibration reduction structure, including: A primary spring damping mechanism connected to the robot body is used to absorb high-frequency vibrations; A secondary magnetohydrodynamic damper connected to the primary spring damping mechanism is used to adaptively adjust the damping coefficient according to the vibration frequency to absorb low and medium frequency vibrations. The three-level MEMS active vibration isolation platform, which carries the multi-parameter sensor module, is used to perform high-precision attitude compensation through microelectromechanical systems to isolate minute vibrations.

3. The gas inspection robot according to claim 1, characterized in that, The multi-parameter sensor module has a gas flow channel separation structure inside, which physically isolates the sampling gas paths of different gas sensors to avoid cross-interference between sensors.

4. The gas inspection robot according to claim 1, characterized in that, The central control unit includes an edge computing module and a UWB positioning module; The edge computing module is used to run multi-source information fusion algorithms and machine learning calibration algorithms, analyze and process sensor data in real time, intelligently compensate for detection results in combination with environmental factors, and combine with the location information provided by the UWB positioning module to realize intelligent inspection decision-making and accurate early warning of abnormal situations. The UWB positioning module is used to provide the robot with high-precision real-time location information, assisting in inspection path planning and data spatiotemporal stamping.

5. The gas inspection robot according to claim 1, characterized in that, The switching logic of the intelligent power management unit is as follows: when the central control unit determines that the methane concentration value exceeds the preset safety threshold, it triggers a switching command and completes the switching from the main power supply module to the self-generated backup module within 2 seconds to ensure the continuity of the monitoring function.

6. The gas inspection robot according to claim 1, characterized in that, The robot also includes a remote air extraction pipeline interface for connecting to a fixed air extraction pipeline in the mine to assist in monitoring the gas emission in a specific area.

7. An inspection method for a gas inspection robot applied to any one of claims 1 to 6, characterized in that, The method includes the following steps: S1: Under the command of the central control unit, the robot moves and inspects along a preset track via a rail-mounted mobile mechanism; S2: During the movement, the adaptive vibration isolation system monitors the vibration in real time and actively reduces the vibration to ensure the detection stability of the multi-parameter sensor module; S3: The multi-parameter sensor module synchronously collects data on the concentration and temperature of methane, carbon monoxide, carbon dioxide, oxygen, and avoids data cross-interference through a gas flow channel separation structure; S4: The central control unit uses multi-source information fusion algorithms and machine learning calibration algorithms to perform real-time fusion and analysis of the collected multi-source data, dynamically calibrates and compensates sensor readings according to environmental parameters, and generates inspection logs with spatiotemporal stamps by combining UWB positioning information. S5: The central control unit continuously monitors the gas concentration. When the concentration exceeds the preset threshold, it immediately controls the intelligent power management unit to switch to the self-generating backup module and issues an early warning message.

8. The inspection method according to claim 7, characterized in that, In step S4, the central control unit uses multi-source information fusion algorithms and machine learning calibration algorithms to perform real-time fusion and analysis of the collected multi-source data, specifically including the following steps: S41: Data preprocessing and alignment: Clean the collected data from various sensors and convert it into a data stream with a uniform format and synchronized timestamps; Data cleaning involves using Kalman filtering or moving average filtering to perform preliminary smoothing on the time series data of each individual sensor, filtering out abnormal spikes; Format unification involves normalization or standardization to map data from different sensors to a unified numerical range. S42: State estimation and spatiotemporal correlation: By using advanced algorithms, isolated data points are fused into an organically related whole to achieve accurate estimation of the environmental state; (1) Constructing spatiotemporal data objects: After processing in step S41, the data of each timestamp is packaged into a structured "spatiotemporal data object", so that each data has four dimensions of attributes: time, space, environment and its own state; (2) Environmental interference compensation based on machine learning: using a pre-trained gradient boosting decision tree model or a small neural network model to learn and fit the complex nonlinear relationship between interference factors and sensor reading errors; (3) Constructing a dynamic spatiotemporal heat map: The robot updates the continuously generated spatiotemporal data objects into a dynamic three-dimensional grid map in real time; and estimates the adjacent grids that are not directly measured through spatial interpolation algorithms, thereby forming a continuous and smooth environmental situation map; S43: Situational Awareness and Decision Output: Based on the environmental situation map generated in step S42, perform intelligent analysis, risk assessment, and make a final decision.

9. The inspection method according to claim 7, characterized in that, The method also includes a fixed-point detection mode: when the robot moves to a preset key detection point, the central control unit controls the magnetohydrodynamic damper to increase damping, locks the MEMS active vibration isolation platform, and switches the adaptive vibration isolation system to a high stability state to perform high-precision fixed-point detection.

Images