Intelligent tunnel gas monitoring device and method

Abstract

The present invention provides an intelligent tunnel gas monitoring device and method. The device includes a gas sensor, a cloud bridge serial port converter, a server, a controller, a laser radar, an IMU module, a communication module, a motor drive module, a motor, and a crawler walking mechanism. A tunnel map is generated using a SLAM algorithm based on collected point cloud data. An LSTM model is constructed and trained based on collected gas concentration data. A prediction model for gas concentration is constructed to achieve long- and short-term prediction of gas concentration. Generating a tunnel map enables automatic path planning for the device, autonomous navigation and mapping functions for GPS-free underground equipment, intelligent underground gas collection, cloud-based intelligent storage, and intelligent analysis and prediction. This provides guarantees for safe, green, and efficient mining in deep coal mines.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent mine safety monitoring, and specifically relates to a cloud-based intelligent monitoring device and method for tunnel gas. Background Art

[0002] Mine gas concentration monitoring is particularly important for gas disaster prevention and control. Traditional monitoring mainly relies on manually arranging a large number of measuring points and taking multiple measurements at each measuring point. The on-site construction process is cumbersome, the monitoring workload increases, the data processing time is long and the accuracy is low. In addition, the existing monitoring equipment has limitations such as high cost, poor anti-interference ability, cumbersome operation, high technical requirements, and small amount of measurement data. Based on the above situation, there is an urgent need for a better tunnel gas cloud-based intelligent monitoring equipment to achieve the purpose of accurately, quickly, intelligently and efficiently monitoring tunnel gas concentration. Summary of the Invention

[0003] Based on the above problems, the present invention proposes an intelligent tunnel gas monitoring device, including: a gas sensor, a cloud bridge serial port converter, a server, a controller, a laser radar, an IMU module, a communication module, a motor drive module, a motor, and a crawler walking mechanism; the server is wirelessly connected to the controller through the communication module, the gas sensor is electrically connected to the cloud bridge serial port converter, the cloud bridge serial port converter is electrically connected to the controller, the laser radar and the motor drive module are respectively electrically connected to the controller, the IMU module and the motor are respectively electrically connected to the motor drive module, and the crawler walking mechanism is driven by the motor.

[0004] The gas sensor is used to collect the gas concentration in the tunnel and transmit it to the controller;

[0005] The laser radar is used to collect point cloud data of the surrounding environment of the device during walking and transmit it to the controller;

[0006] The IMU module is used to collect acceleration and angular velocity signals of the device during walking and transmit them to the motor drive module;

[0007] The motor drive module is used to control the movement of the motor-driven crawler walking mechanism and is also used to transmit the received acceleration and angular velocity signals to the controller;

[0008] The controller is used to transmit the received signal to the server;

[0009] The communication module is used to send WiFi signals or Ethernet signals;

[0010] The server is used to predict the gas concentration value based on the trained LSTM model and plan the walking route based on the generated tunnel map.

[0011] The crawler-type walking mechanism includes a belt, gears, and a metal chassis; an explosion-proof housing is installed on the device, and the metal chassis located on one side of the mechanism is fixed to the bottom of the explosion-proof housing. The motor drives the driving gear to rotate, and the driving gear drives the four driven gears to rotate. The gears and belts are driven by meshing.

[0012] The gas sensor and cloud bridge serial port converter are respectively installed on the middle partition inside the cavity composed of the explosion-proof shell. The laser radar is installed on the top of the explosion-proof shell. The IMU module, communication module and motor drive module are respectively installed on the bottom of the explosion-proof shell.

[0013] Furthermore, in order to collect video information during the movement of the equipment, a camera is installed on the top of the explosion-proof casing. The camera is wirelessly connected to the server. RViz software is installed on the server. The camera switch is controlled by RViz software. When the camera is turned on, the camera collects video information in the tunnel in real time and uploads it to the server for visual display.

[0014] A method for monitoring gas concentration using the intelligent tunnel gas monitoring device comprises:

[0015] Step 1: The RViz software installed on the server controls the monitoring device to move in the monitored tunnel. The laser radar collects point cloud data of the tunnel's surrounding environment in real time, the gas sensor collects gas concentration data in real time, and the IMU module collects angular velocity and acceleration information during the movement.

[0016] Step 2: Generate a road map using the SLAM algorithm based on the collected point cloud data;

[0017] Step 3: Build an LSTM model and train it using the collected gas concentration data;

[0018] Step 4: After setting the starting and ending points of the walk, collect the gas concentration values in the tunnel according to the generated tunnel map, and use the trained LSTM model to output the predicted gas concentration.

[0019] The step 3 comprises:

[0020] Step 3.1: Filter the collected gas concentration data and use the filtered gas concentration data as the sample data set;

[0021] Step 3.2: Build an LSTM model and train it using the sample dataset to obtain a trained LSTM model.

[0022] Step 3.3: Use the trained LSTM model to output gas concentration values within different prediction periods.

[0023] The beneficial effects of the present invention are:

[0024] The present invention proposes an intelligent tunnel gas monitoring device and method. By constructing a pre-stored model of gas concentration, long- and short-term gas concentration predictions are achieved. By generating a tunnel map, the device can automatically plan the path, realize autonomous navigation and mapping functions without GPS equipment underground, and intelligently collect underground gas, intelligently store it in the cloud, and intelligently analyze and predict it, providing guarantees for safe, green, and efficient mining in deep coal mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a system block diagram of the intelligent tunnel gas monitoring equipment of the present invention;

[0026] Figure 2 This is a flow chart of the gas concentration monitoring method using the tunnel gas intelligent monitoring equipment in the present invention;

[0027] Figure 3 This is a left view of the structure of the tunnel gas cloud intelligent monitoring equipment in the present invention;

[0028] Figure 4 This is a front view of the structure of the tunnel gas cloud intelligent monitoring equipment in the present invention;

[0029] Figure 5 This is a top view of the structure of the tunnel gas cloud intelligent monitoring equipment in the present invention;

[0030] Figure 6 The control principle diagram of the intelligent roadway gas monitoring device in the present invention; wherein (a) is the principle diagram for constructing a roadway map, and (b) is the principle diagram for collecting gas concentration;

[0031] Figure 7 This is a schematic diagram of the construction principle of the LSTM (Long Short-Term Memory) model in the present invention;

[0032] Figure 8 This is the electrical wiring diagram of the intelligent tunnel gas monitoring device of the present invention;

[0033] In the figure, 1. Equipment bottom panel; 2. Equipment front panel; 3. Equipment rear panel; 4. Equipment top panel; 5. Explosion-proof casing; 6. Equipment middle partition; 7. Bolt I; 8. Bolt II; 9. Drive panel; 10. Metal chassis; 11. Motor; 12. Inertial measurement unit (IMU module); 13. Metal gear; 14. Pentagonal track; 15. Data transmission line; 16. Power supply line; 17. Battery; 18. LiDAR; 19. Binocular camera; 20. Bolt III; 21. Raspberry Pi 4B control panel; 22. Gas sensor; 23. Cloud Bridge serial port converter; 24. Ground remote control station; 25. Communication module (WiFi / Ethernet). DETAILED DESCRIPTION

[0034] The invention will be further described below with reference to the accompanying drawings and specific implementation examples.

[0035] like Figure 1 As shown, an intelligent tunnel gas monitoring device includes: a gas sensor, a cloud bridge serial port converter, a server, a controller, a laser radar, an IMU module, a communication module, a motor drive module, a motor, and a crawler walking mechanism; the server is wirelessly connected to the controller through the communication module, the gas sensor is electrically connected to the cloud bridge serial port converter, the cloud bridge serial port converter is electrically connected to the controller, the laser radar and the motor drive module are respectively electrically connected to the controller, the IMU module and the motor are respectively electrically connected to the motor drive module, and the crawler walking mechanism is driven by the motor.

[0036] The gas sensor is used to collect the gas concentration in the tunnel and transmit it to the controller;

[0037] The laser radar is used to collect point cloud data of the surrounding environment of the device during walking and transmit it to the controller;

[0038] The IMU module is used to collect acceleration and angular velocity signals of the device during walking and transmit them to the motor drive module;

[0039] The motor drive module is used to control the movement of the motor-driven crawler walking mechanism and is also used to transmit the received acceleration and angular velocity signals to the controller;

[0040] The controller is used to transmit the received signal to the server;

[0041] The communication module is used to send WiFi signals or Ethernet signals;

[0042] The server is used to predict the gas concentration value based on the trained LSTM model and plan the walking route based on the generated tunnel map.

[0043] The crawler-type walking mechanism includes a belt, gears, and a metal chassis; an explosion-proof housing is installed on the device, and the metal chassis located on one side of the mechanism is fixed to the bottom of the explosion-proof housing. The motor drives the driving gear to rotate, and the driving gear drives the four driven gears to rotate. The gears and belts are driven by meshing.

[0044] The gas sensor and cloud bridge serial port converter are respectively installed on the middle partition inside the cavity composed of the explosion-proof shell. The laser radar is installed on the top of the explosion-proof shell. The IMU module, communication module and motor drive module are respectively installed on the bottom of the explosion-proof shell.

[0045] In order to collect video information during the movement of the equipment, a camera is installed on the top of the explosion-proof casing. The camera is wirelessly connected to the server. The RViz software is installed on the server. The camera switch is controlled by the RViz software. When the camera is turned on, the camera collects video information in the tunnel in real time and uploads it to the server for visual display.

[0046] In this embodiment, the server is a computer, and RViz software is installed on the server. The RViz software is used to generate a map of the lane based on the point cloud data collected by the laser radar. The controller is a Raspberry Pi (model H-20-Header 20X2). The RViz software is used to build a ground remote control station for path planning. The starting and ending points of the walk are set through the ground remote control station, and the running trajectory of the equipment is displayed. A monitoring interface for monitoring real-time data acquisition values is also programmed on the server; the motor model is TSSOP-14_74HC00PW; the motor drive module (model STM32); the gas sensor (model JXBS-4001); the communication module (model SIM900A); the IMU module (model IMU PA-IMU-03D); the cloud bridge serial port converter model is the JXYH-7001 series Ethernet Guanqiao Cloud Box, which can realize two-way seamless conversion between Ethernet and RS485 interfaces. The Ethernet Guanqiao Cloud Box can organically combine RS485 devices with computers or servers through Ethernet. The specific electrical wiring diagram is as follows: Figure 8 shown.

[0047] like Figure 6As shown in (a), the ground remote control station is connected to the Raspberry Pi 4B control panel via WIFI / LAN to send commands to it for control. The Raspberry Pi 4B control panel controls the lidar and STM32 motor drive control module through commands. The IMU module collects acceleration and angular velocity information and transmits it to the STM32 drive control module. The STM32 motor drive control module controls the servo and motor movements. The lidar collects lane environment data and transmits it to the Raspberry Pi 4B control panel. The Raspberry Pi 4B control panel feeds its information back to the ground remote control station. The lane map construction visualization is realized by inputting commands through the Rviz software in the ground remote control station.

[0048] like Figure 6 As shown in (b), the gas sensor collects the gas concentration in the tunnel, uses the Yunqiao serial port converter to convert the RS485 communication signal into a network communication signal, and automatically stores the data in the cloud server.

[0049] like Figures 3-5As shown, the device bottom panel 1, the device front panel 2, the device rear panel 3, the device top panel 4, the explosion-proof shell 5, the device middle partition 6, the bolts I7, and the bolts II8 constitute the device shell of the monitoring device. The drive panel 9 is fixed to the metal chassis 10, and they are connected by welding. The motor 11 is fixed to the outside of the metal chassis 10 by bolts II8 to provide power for the equipment. The metal gear 13 is fixed to the outside of the metal chassis 10 by bolts II8. The anti-slip pentagonal track 14 cooperates with the metal gear 13, and the two form a gear-belt connection to achieve normal walking on complex ground in the tunnel. The nine-axis gyroscope 12 is fixed to the drive control panel 9, and they are welded. The drive control panel 9 is connected to the motor 11 through a data transmission line 15, and the battery 17 is connected to the drive control panel 9 through a power supply line 16 to provide power for the drive control panel. The LiDAR 18 is secured to the top panel 4 of the device via bolts III20. The binocular camera 19 is also secured to the top panel 4 via bolts III20. The LiDAR 18 is connected to the Raspberry Pi 4B control panel 21 via data transmission cable 15 to collect point cloud data of the roadway and construct a map. The binocular camera 19 is also connected to the Raspberry Pi 4B control panel 21 via data transmission cable 15 to obtain image information of the roadway. The LiDAR measures the propagation time of a light pulse hitting an object and reflecting back to the receiver. Based on the known speed of light, this propagation time is converted into the distance to the object measured by the LiDAR, which then forms a point cloud. A filter is used to remove point cloud data with severe motion distortion or areas with distances exceeding a threshold. The point cloud data is divided into units, and the data within a unit is histogrammed to obtain the least squares optimal solution to obtain the interpolation position of the data. Graph optimization is performed between units, and each unit is treated as a node to estimate its relative position with other nodes. High-precision LiDAR SLAM mapping is then performed in segments, and the starting and end points of the map are marked on the constructed map to ensure that the map construction conforms to reality and is not distorted or fragmented.

[0050] The gas sensor 22 is secured to the central partition 6 of the device via bolts III20. The cloud bridge serial converter 23 is also secured to the central partition 6 via bolts III20. The cloud bridge serial converter 23 is connected to the gas sensor 22 via a data transmission line 15, converting the sensor's RS485 signal into a network signal. The cloud bridge serial converter 23 is also connected to the Raspberry Pi 4B control panel 21 via data transmission line 15, transmitting the network signal to the cloud database of the ground remote control station 24, which is built on a computer. This data is then analyzed and displayed using intelligent algorithms for gas data. The communication module 25 is installed in the Raspberry Pi 4B control panel, with the two connected by welding. The data transmission line 15 connects the sensor and the control panel, further transmitting signals to ensure the device collects data properly.

[0051] like Figure 2As shown, a method for monitoring gas concentration using the tunnel gas intelligent monitoring device includes:

[0052] Step 1: The RViz software installed on the server controls the monitoring device to move in the monitored tunnel. The laser radar collects point cloud data of the tunnel's surrounding environment in real time, the gas sensor collects gas concentration data in real time, and the IMU module collects angular velocity and acceleration information during the movement.

[0053] Step 2: Generate a road map using the SLAM algorithm based on the collected point cloud data; SLAM stands for simultaneous localization and mapping.

[0054] Step 3: Build an LSTM model and train it using the collected gas concentration data, such as Figure 7 Shown; including:

[0055] Step 3.1: Filter the collected gas concentration data and use the filtered gas concentration data as the sample data set;

[0056] Step 3.2: Build an LSTM model and train it using the sample dataset to obtain a trained LSTM model.

[0057] Step 3.3: Use the trained LSTM model to output gas concentration values within different prediction periods;

[0058] Step 4: After setting the starting and ending points of the walk, collect the gas concentration values in the tunnel according to the generated tunnel map, and use the trained LSTM model to output the predicted gas concentration.

[0059] The use process of the monitoring device is as follows:

[0060] S10: Check the equipment installation and whether the equipment can operate normally;

[0061] S20: Connect the motor drive module and gas sensor in the device to the Raspberry Pi 4B control module through the communication module (WIFI / Ethernet);

[0062] S30: The ground remote control station is connected to the Raspberry Pi 4B control module through the communication module (WIFI / Ethernet) to ensure the normal operation of the device;

[0063] S40: Turn on the drive module and the gas data processing module (gas sensor) to collect and transmit data;

[0064] S50: The LiDAR sensor, fixed to the top of the equipment housing, transmits a light pulse that hits an object and reflects back to the receiver. Based on the known speed of light, this propagation time is converted into the distance to the object measured by the LiDAR. The data is then converted into point cloud data. The ground remote control station uses the SLAM algorithm to process the data and construct an initial map of the underground tunnel.

[0065] S60: The equipment uses the tunnel SLAM positioning and mapping module to build an initial map. The operator uses the ground remote control module to mark the equipment's gas collection path. The equipment then conducts autonomous gas collection according to the planned path.

[0066] S70: The gas sensor installed inside the equipment housing cavity is connected to the Raspberry Pi 4B control module through a communication module. The sensor uses a carrier catalytic element. The underground gas environment concentration can be directly collected and monitored by the gas sensor to realize display and alarm functions. The collected data is converted from RS485 signals to network signals through the cloud bridge signal converter and stored in the gas cloud database. The cloud gas data is analyzed and displayed by intelligent algorithms and charts are drawn through the gas big data platform. Operators can view it through the ground remote control module.

[0067] S80: After the equipment has completed the collection, it monitors whether the measurement data is valid. If the data is valid, the equipment returns autonomously through the ground remote control module. When the equipment returns to the ground, all modules of the equipment are shut down and inspected.

Claims

1. An intelligent tunnel gas monitoring device, characterized in that: include: Gas sensor, cloud bridge serial port converter, server, controller, laser radar, IMU module, communication module, motor drive module, motor, crawler walking mechanism; the server is wirelessly connected to the controller through the communication module, the gas sensor is electrically connected to the cloud bridge serial port converter, the cloud bridge serial port converter is electrically connected to the controller, the laser radar and the motor drive module are respectively electrically connected to the controller, the IMU module and the motor are respectively electrically connected to the motor drive module, and the crawler walking mechanism is driven by the motor; The gas sensor is used to collect the gas concentration in the tunnel and transmit it to the controller; The laser radar is used to collect point cloud data of the surrounding environment of the device during walking and transmit it to the controller; The IMU module is used to collect acceleration and angular velocity signals of the device during walking and transmit them to the motor drive module; The motor drive module is used to control the movement of the motor-driven crawler walking mechanism and is also used to transmit the received acceleration and angular velocity signals to the controller; The controller is used to transmit the received signal to the server; The communication module is used to send WiFi signals or Ethernet signals; The server is used to predict the gas concentration value based on the trained LSTM model and plan the walking route based on the generated tunnel map; The crawler-type walking mechanism includes a belt, gears, and a metal chassis; the gears include a driving gear and a driven gear; an explosion-proof housing is installed on the device, and the metal chassis located on one side of the mechanism is fixed to the bottom of the explosion-proof housing. The driving gear is driven by the motor to rotate, and the driving gear drives the four driven gears to rotate, and the gears and belts are driven by meshing; the belt adopts a non-slip pentagonal crawler, which is used to form a gear-belt connection with the gears; The gas sensor and cloud bridge serial port converter are respectively installed on the middle partition inside the cavity composed of the explosion-proof shell. The laser radar is installed on the top of the explosion-proof shell. The IMU module, communication module and motor drive module are respectively installed on the bottom of the explosion-proof shell.

2. The intelligent tunnel gas monitoring device according to claim 1, characterized in that: In order to collect video information during the movement of the equipment, a camera is installed on the top of the explosion-proof casing. The camera is wirelessly connected to the server. The RViz software is installed on the server. The camera switch is controlled by the RViz software. When the camera is turned on, the camera collects video information in the tunnel in real time and uploads it to the server for visual display.

3. A method for monitoring gas concentration using the intelligent tunnel gas monitoring device according to any one of claims 1 to 2, characterized in that: include: Step 1: The RViz software installed on the server controls the monitoring device to move in the monitored tunnel. The laser radar collects point cloud data of the tunnel's surrounding environment in real time, the gas sensor collects gas concentration data in real time, and the IMU module collects angular velocity and acceleration information during the movement in real time. Step 2: Generate a road map using the SLAM algorithm based on the collected point cloud data; Step 3: Build an LSTM model and train it using the collected gas concentration data; Step 3.1: Filter the collected gas concentration data and use the filtered gas concentration data as the sample data set; Step 3.2: Build an LSTM model and train it using the sample dataset to obtain a trained LSTM model. Step 3.3: Use the trained LSTM model to output gas concentration values within different prediction periods; Step 4: After setting the starting and ending points of the walk, collect the gas concentration values in the tunnel according to the generated tunnel map, and use the trained LSTM model to output the predicted gas concentration.

Images