Device and method for detecting air leakage of ventilation hose in plateau tunnel based on visual recognition

Abstract

The present invention belongs to the technical field of fluid tightness testing of structural components, and specifically relates to a device and method for detecting air leakage of a ventilation hose in a plateau tunnel based on visual recognition, wherein the device includes a support frame, a traction rope and a traction parachute, the support frame includes a central axis, a plurality of spring rods are radially arranged at both ends of the central axis, a support sheet is fixed to the top of the spring rod, the support sheets at the same end of the central axis are circumferentially distributed, a battery and a plurality of cameras are installed in the middle of the central axis, the cameras are evenly distributed along the circumference of the central axis, and the sum of the viewing angles of all cameras is not less than 360°; the traction parachute includes an umbrella frame and an umbrella surface, one end of the traction rope is connected to the central axis, and the other end is connected to the umbrella frame; and also includes a central control system for receiving camera photos and visually identifying whether there are pores in the photos, the central control system and the camera are powered by batteries, and the central control system is connected to an audible and visual alarm device. The present invention can detect air leakage of the ventilation hose safely and quickly.

Description

Technical Field

[0001] The present invention belongs to the technical field of fluid tightness testing of structural components, and specifically relates to a device and method for detecting air leakage of a ventilation hose in a plateau tunnel based on visual recognition. Background Art

[0002] Tunnel construction ventilation is a necessary way to exchange air with the outside world. It is an indispensable safeguard for providing fresh air to workers and maintaining the normal operation of electromechanical equipment. Most tunnels use duct pressure-type air supply to the face: an axial flow fan is installed outside the tunnel entrance, and the axial flow fan delivers fresh air to the tunnel excavation working face through the duct connected to it. For high-altitude plateau areas, in order to facilitate transportation and storage, a considerable number of air ducts currently use ventilation hoses that are smaller in size and lighter in weight. Such ventilation hoses are usually made of glass fiber cloth, nylon cloth or canvas into a tube shape. Some ventilation hoses are also equipped with spiral steel wire rings that can be folded or retracted.

[0003] The ventilation hose may be damaged and leak due to the folding or contraction of the ventilation hose, and the air duct may be damaged and leak during installation or use due to the operation of personnel or the extrusion of the template trolley. Damage and leakage of the air duct will reduce the efficiency of the ventilation system and increase the difficulty of operation in the tunnel, especially in high-altitude tunnels with low air pressure and low oxygen content, where air duct leakage may affect personnel safety.

[0004] For the detection of air leakage in air ducts, some methods are also disclosed in the relevant technology: Light detection method: the inspector needs to hold an incandescent lamp through the entire air duct and manually observe the air leakage position of the air duct, but for the hose, there is no bearing capacity, so it cannot be realized, and for the hypoxic environment of the plateau, the inspector enters the small and closed environment of the ventilation hose, which is very easy to be hypoxic, seriously threatening the personal safety of the inspector; in addition, the pressure in the air duct is controlled to 700Pa by a blower and a static pressure sensor, and then the air leakage sound of the air duct is manually observed, but the construction environment is noisy, which seriously affects the identification of the leakage sound; for example, a smoke generating device is installed in the air duct, and a fan is used to increase the static pressure in the branch pipe, and at the same time observe whether there is smoke leakage from the air duct wall, but for the environment of smoke and dust such as tunnel construction, this method is very limited. Therefore, a means of safely and quickly detecting air leakage in the ventilation hose is needed. Summary of the invention

[0005] The present invention intends to provide a plateau tunnel ventilation hose air leakage detection device based on visual recognition, so as to safely and quickly perform air leakage detection on the ventilation hose.

[0006] In order to achieve the above object, the present invention is achieved in the following ways:

[0007] A plateau tunnel ventilation hose air leakage detection device based on visual recognition includes a support frame, a traction rope and a traction parachute. The support frame includes a central axis. A plurality of spring rods are radially arranged at both ends of the central axis. A support sheet is fixed to the top of the spring rod. The support sheets at the same end of the central axis are distributed in a circle. A battery and a plurality of cameras are installed in the middle of the central axis. The cameras are evenly distributed along the circumference of the central axis. The sum of the viewing angles of all cameras is not less than 360°. The traction parachute includes an umbrella frame and an umbrella surface. One end of the traction rope is connected to the central axis, and the other end is connected to the umbrella frame.

[0008] It also includes a central control system for receiving camera photos and visually identifying whether there are pores in the photos. The central control system and the camera are powered by batteries, and the central control system is connected to an audible and visual alarm device.

[0009] The working principle and beneficial effects of this solution are as follows: the entire device is placed into the ventilation hose from the end of the ventilation hose close to the axial flow fan, the central axis is coaxial with the ventilation hose, the elasticity of the spring rod allows the support sheet to support the ventilation hose in a cylindrical shape, the traction parachute is located on the side of the support frame away from the axial flow fan, and the umbrella surface faces the support frame. After the axial flow fan blows, the wind passes through the support frame and blows onto the umbrella surface, so that the traction parachute is thrust and drives the support frame to move forward at a constant speed. During the movement, the inspection personnel only need to observe the raised position of the ventilation hose lifted by the support sheet to determine the position of the entire device in the ventilation hose. During the movement of the entire device, the ventilation hose in the middle area of ​​the central axis is supported and stretched due to the spring rods and support plates at both ends of the central axis. The camera on the central axis takes 360° photos of the area where the inner wall of the ventilation hose is stretched. With the traction of the traction parachute, the camera moves to the next area after taking pictures of one area, automatically taking pictures continuously, and transmitting the pictures to the central control system. The central control system performs visual recognition on the pictures. When the existence of a gap is recognized, the sound and light alarm device immediately sends out a sound and light alarm. The inspector can determine the section of the ventilation hose where the gap is by listening or observing the position of the sound and light alarm device. At this time, the inspector can mark it, stop blowing, and carefully check the ventilation hose at the marked position; or the inspector can wait until the entire ventilation hose inspection is completed and then carefully check the areas near all marked positions.

[0010] This solution can quickly detect air leaks through visual recognition, and inspectors can quickly locate the area of ​​air leaks through the sound and light alarm device, greatly reducing the area that requires manual inspection and verification by inspectors, and can significantly improve the accuracy and efficiency of air leak detection. The entire inspection process is fully automatic and does not require inspectors to enter the ventilation hose, avoiding risks to personal safety of inspectors.

[0011] Optionally, the umbrella frame is a circular frame, and two mounting seats are provided inside the umbrella frame, the two mounting seats are arranged opposite to each other, and the two mounting seats are connected to a rotating shaft for common rotation, the umbrella surface is a circular plate with a diameter smaller than the inner diameter of the umbrella frame, the umbrella surface is fixed on the rotating shaft, the umbrella surface is parallel to the rotating shaft, one end of the rotating shaft is connected to a first servo motor, and the first servo motor is electrically connected to the central control system. The rotation of the first servo motor is controlled by the central control system, thereby controlling the rotation of the rotating shaft, thereby adjusting the inclination angle of the umbrella surface relative to the umbrella frame, and then adjusting the size of the windward area of ​​the umbrella surface, realizing the adjustment of the traction size, and finally realizing the adjustment of the moving speed of the entire device in the ventilation hose.

[0012] Optionally, the umbrella frame includes an umbrella pole and a plurality of umbrella ribs, one end of the umbrella rib is connected to one end of the umbrella pole, the other end of the umbrella rib is a free end, the umbrella surface is a flexible cloth, the edge of the umbrella surface is evenly distributed with connection parts of the same number as the number of umbrella ribs, and the connection parts are fixedly connected to the free ends of the umbrella ribs; a rotating drum is rotatably connected to the umbrella pole, and the rotating drum is connected with the same number of flexible ropes as the number of umbrella ribs, and the flexible ropes are respectively fixedly connected to the free ends of the umbrella ribs, and a worm gear is coaxially fixedly connected to the rotating drum, and the worm gear is rotatably connected to the umbrella pole, and a second servo motor is fixed on the rotating drum, and the output end of the second servo motor is connected to a worm, and the worm gear meshes with the worm, and the second servo motor is electrically connected to the central control system. The central control system controls the rotation of the rotating drum, thereby controlling the winding length of the flexible rope on the rotating drum, thereby controlling the opening degree of the umbrella surface, and realizing the adjustment of the traction force. The greater the opening degree, the greater the traction force, and finally realizing the adjustment of the moving speed of the entire device in the ventilation hose.

[0013] Optionally, a contact speed sensor is also included, which is installed on the support sheet, and the signal of the contact speed sensor is transmitted to the central control system, so as to collect the moving speed of the entire device in the ventilation hose, so as to adjust the frequency of camera shooting according to the moving speed, ensure the continuity of the shooting area, and avoid missing shooting areas.

[0014] Optionally, a sled is connected to the support sheet, and the length direction of the sled is parallel to the axis direction of the central axis, so that the entire device can move more smoothly in the ventilation hose.

[0015] Optionally, a counterweight is fixed on the side of the central axis, and the counterweight is located on the same side of the central axis. Under the action of the gravity of the counterweight, the center of gravity of the support frame is biased to one side, so that one side of the central axis is always facing downward, thereby preventing the support frame from rotating in the ventilation hose and affecting the shooting quality of the camera.

[0016] In another aspect of the present invention, a method for detecting air leakage in a plateau tunnel ventilation hose based on visual recognition is also provided, comprising:

[0017] Collection steps: The central control system obtains photos from the camera;

[0018] Identification step: pre-process the photo, identify the pre-processed photo, and determine whether there is a pore in the current photo based on the identification result. If there is a pore, control the sound and light alarm device to alarm.

[0019] This solution can quickly detect air leaks through visual recognition, and inspectors can quickly locate the area of ​​air leaks through the sound and light alarm device, greatly reducing the area that requires manual inspection and verification by inspectors, and can significantly improve the accuracy and efficiency of air leak detection. The entire inspection process is fully automatic and does not require inspectors to enter the ventilation hose, avoiding risks to personal safety of inspectors.

[0020] Optionally, in the acquisition step, the central control system further acquires the current moving speed from a contact speed sensor;

[0021] It also includes: a shooting adjustment step: the central control system calculates the moving distance L1 based on the current moving speed and the current shooting interval of the camera, and determines whether the moving distance L1 is greater than the pre-stored length L2 of the camera shooting area. If greater, the camera's shooting interval is recalculated based on the current moving speed and the length L2 of the camera shooting area, and the calculated shooting interval is used to update the camera's current shooting interval.

[0022] When L1 is greater than L2, it indicates that there are areas that are missed, and the shooting interval needs to be shortened to ensure that there are no missed areas between two adjacent photos taken by the same camera.

[0023] Optionally, in the acquisition step, the central control system also records the displacement time;

[0024] It also includes a positioning step, in which the central control system calculates the total displacement distance based on the moving speed and the displacement time;

[0025] The central control system also determines the position of the pore in the photo; the central control system also obtains the photographing area of ​​the camera corresponding to the photo, and determines the circumferential position of the pore in the ventilation hose based on the photographing area of ​​the camera and the position of the pore in the photo;

[0026] The central control system stores the circumferential position of the pores in the ventilation hose and the corresponding total displacement distance when the pores appear.

[0027] The circumferential position of the pores in the ventilation hose and the corresponding total displacement distance when the pores appear can help inspectors quickly determine the position of the ventilation hose where the pores are located.

[0028] Optionally, it also includes a speed adjustment step: the central control system determines whether the current moving speed is within a preset range. If not, the first servo motor or the second servo motor is controlled to rotate based on the current moving speed, and this step is repeated until the current moving speed is within the preset range.

[0029] By dynamically adjusting the rotation of the first servo motor or the second servo motor, the moving speed can be kept within a preset range, thereby avoiding blurring of photos due to excessive moving speed, or affecting detection efficiency due to excessive moving speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a schematic diagram of the structure of a plateau tunnel ventilation hose air leakage detection device based on visual recognition in Embodiment 1 of the present invention;

[0031] Figure 2 This is a flow chart of a method for detecting air leakage in a ventilation hose of a plateau tunnel based on visual recognition in Embodiment 1 of the present invention;

[0032] Figure 3 It is a structural schematic diagram of a plateau tunnel ventilation hose leakage detection device based on visual recognition in Example 2 of the present invention.

[0033] The marks in the drawings of the specification include: ventilation hose 1, central axis 2, spring rod 3, support plate 4, sled board 5, pod 6, fixing ring 7, camera 8, LED light strip 9, traction rope 10, frame 11, mounting seat 12, rotating shaft 13, contact speed sensor 14, umbrella pole 15, umbrella rib 16, nylon cloth 17, nylon rope 18, rotating drum 19, slide 20, umbrella cover 21. DETAILED DESCRIPTION

[0034] The following is further described in detail through specific implementation methods:

[0035] Embodiment 1

[0036] This embodiment is basically Figure 1 As shown: a plateau tunnel ventilation hose air leakage detection device based on visual recognition, including a support frame, a traction rope 10 and a traction parachute, the support frame includes a central axis 2, and four spring rods 3 are radially arranged at both ends of the central axis 2. In this embodiment, the spring rod 3 is a gas spring, and a circular arc-shaped support sheet 4 is welded on the top of each spring rod 3, and a sled board 5 is welded in the middle of the outer side surface of each support sheet 4. The two sides of the sled board 5 are tilted inward by 15°, and the length direction of the sled board 5 is parallel to the axial direction of the central axis 2.

[0037] Two pods 6 are welded on the side of the central axis 2. The pods 6 are located on the same side of the central axis 2. A battery is installed in one of the pods 6, and a central control system is installed in the other pod 6. A fixing ring 7 and an alarm belt are coaxially installed on the central axis 2 between the pods 6. A buzzer and an LED light belt 9 (with a built-in ballast, installed in the pod 6 with the battery) are installed on the alarm belt. Four CMOS cameras 8 are evenly installed circumferentially on the fixing ring 7. The cameras 8 are connected to the central control system through a serial port. The viewing angles of the four cameras 8 cover the wall of the ventilation hose 1 circumferentially. The battery supplies power to the LED light belt 9, the buzzer, the camera 8 and the central control system. The pods 6 and the components inside the pods 6 constitute a counterweight. A contact speed sensor 14 is installed on the support sheet 4 near one side of the pod 6. The contact speed sensor 14 usually includes a cantilever. A roller is rotatably connected to the cantilever. The roller rotates by friction with the object to be measured. The moving speed of the object to be measured is detected by the rotation of the roller. The contact speed sensor 14 is electrically connected to the central control system.

[0038] The umbrella frame is a circular frame 11, one end of the flexible rope is fixed to the central axis 2, and the other end is divided into four strands and respectively fixed to four points on the circumference of the frame 11. Two mounting seats 12 are provided on the inner side of the umbrella frame, one of which has a cavity inside, and the two mounting seats 12 are arranged opposite to each other, and the two mounting seats 12 are connected to a rotating shaft 13 for common rotation, and the rotating shaft 13 passes through the center of the circle of the frame 11. The umbrella surface 21 is a circular plate with a diameter smaller than the inner diameter of the umbrella frame, and the umbrella surface 21 is fixed on the rotating shaft 13, and the rotating shaft 13 coincides with a diameter line of the umbrella surface 21. One end of the rotating shaft 13 extends into the cavity and is installed with a first worm gear, and a support is fixed on the cavity, and a first worm is rotatably connected to the support, and one end of the first worm is connected to the output end of the first servo motor, and the first worm gear meshes with the first worm, and the first servo motor is electrically connected to the central control system.

[0039] The specific implementation process is:

[0040] Remove the air inlet end of the ventilation hose 1 from the axial flow fan, press the spring rod 3 to retract the sled 5 and the support sheet 4 inward, put the traction parachute frame 11 into the ventilation hose 1 from the inlet end of the ventilation hose 1, and then put the entire support frame into the ventilation hose 1. The central axis 2 is coaxial with the ventilation hose 1. The elasticity of the spring rod 3 allows the support sheet 4 to support the ventilation hose 1 in a cylindrical shape. The side of the central axis 2 with the pod 6 will automatically rotate to a drooping position due to its heavy weight. Then fix the inlet end of the ventilation hose 1 back to the axial flow fan.

[0041] The central control system starts the LED light strip 9 to illuminate the inner wall of the hose (if the ambient light is sufficient and the ventilation hose 1 has a certain light transmittance, the LED light strip 9 does not need to be started). The central control system controls the first servo motor to rotate, driving the rotating shaft 13 to rotate until the umbrella surface 21 is parallel to the umbrella frame, and starts the axial flow fan. The wind flows through the support frame and blows onto the umbrella surface 21, so that the traction parachute is thrusted and drives the support frame to overcome the friction between the sled board 5 and the inner wall of the ventilation hose 1 and move forward. During the movement, the camera 8 on the central axis 2 takes 360° photos of the area where the inner wall of the ventilation hose 1 is stretched. With the traction of the traction parachute, the camera 8 moves to the next area after taking a photo of one area, automatically taking photos continuously, and transmitting the photos to the central control system. The central control system performs visual recognition on the taken photos. When a gap is identified, the central control system controls the first servo motor to rotate so that the umbrella surface 21 is perpendicular to the umbrella frame. If the thrust of the traction parachute is not enough to overcome the friction generated by the ventilation hose 1, the entire device stops, and the central control system controls the LED light strip 9 to flash and starts the buzzer to sound and light alarm. At this time, the inspection personnel only need to arrive at the area for a careful inspection.

[0042] During the movement of the entire device, the roller of the contact speed sensor 14 abuts against and rotates against the inner wall of the ventilation hose 1, and the moving speed of the entire device is detected in real time. The distance L1 moved by the camera 8 during the shooting interval is calculated based on the moving speed and the interval time of the camera 8 taking pictures. Then, the length L2 of the single shooting area of ​​the camera 8 is calculated based on the diameter of the ventilation hose 1 and the viewing angle of the camera 8. When L1 is greater than L2, it means that there is an area that is missed. The central control system immediately controls the first servo motor to rotate, so that the umbrella surface 21 and the umbrella frame form a certain angle, reducing the windward area of ​​the umbrella surface 21, thereby reducing the traction force, slowing down the entire device, reducing L1, and avoiding missed areas.

[0043] like Figure 2 As shown, this embodiment also provides a plateau tunnel ventilation hose air leakage detection method based on visual recognition, including:

[0044] Collection steps: The central control system obtains photos from the camera 8, obtains the current moving speed from the contact speed sensor 14, and records the displacement time; in this embodiment, the displacement time is counted from the start of the device. In this embodiment, the central control system is a server.

[0045] Shooting adjustment steps: The central control system calculates the moving distance L1 based on the current moving speed and the current shooting interval of the camera 8, and determines whether the moving distance L1 is greater than the length L2 of the pre-stored camera 8 shooting area. If greater, the shooting interval of the camera 8 is recalculated based on the current moving speed and the length L2 of the camera 8 shooting area, and the calculated shooting interval is used to update the current shooting interval of the camera 8. For example, when the distance L1 obtained by multiplying the current moving speed by the shooting interval is greater than the length L2 of the camera 8 shooting area, the shooting interval needs to be shortened. In this embodiment, in order to further avoid missing shooting areas and improve the recognition success rate (because the edge quality of the photo is lower than the center quality, there is a situation where the edge fails to be recognized), when calculating the shooting interval, it is also necessary to ensure that two consecutive photos have a certain overlapping area, and the length of the overlapping area is L3. L3 can be selected according to the actual situation. The larger L3 is, the higher the accuracy is, the larger the corresponding photo processing volume is, and the greater the calculation demand is.

[0046] Speed ​​adjustment step: The central control system determines whether the current moving speed is within the preset range. If not, the first servo motor is controlled to rotate based on the current moving speed, and this step is repeated until the current moving speed is within the preset range. By dynamically adjusting the rotation of the first servo motor, the speed can be kept within the preset range to avoid blurring of photos due to excessive speed or affecting detection efficiency due to excessively slow speed.

[0047] Identification step: pre-process the photo, identify the pre-processed photo, determine whether there is a pore on the current photo based on the identification result, and if there is a pore, control the sound and light alarm device to alarm. In this embodiment, the pre-processing is to sharpen the image. In other embodiments, grayscale can also be performed. In this embodiment, the convolutional neural network model is trained in advance with a training sample set containing photos of gaps, so that the recognition success rate of the convolutional neural network model meets the requirements, for example, the success rate is greater than 95%, and then the trained convolutional neural network model is used to identify whether there is a gap in the current photo, and the recognition result is output.

[0048] In the positioning step, the central control system calculates the total displacement distance based on the moving speed and displacement time; the central control system also determines the position of the pore in the photo; the central control system also obtains the photographing area of ​​the camera 8 corresponding to the photo, and determines the circumferential position of the pore in the ventilation hose 1 based on the photographing area of ​​the camera 8 and the position of the pore in the photo; the central control system stores the circumferential position of the pore in the ventilation hose 1 and the corresponding total displacement distance when the pore appears. The actual position of the pore displacement ventilation hose 1 can be determined by the circumferential position of the pore in the ventilation hose 1 and the corresponding total displacement distance when the pore appears.

[0049] In other embodiments, a terminal device for use by inspectors may also be included. The terminal device is wirelessly connected to the central control system, so that inspectors can use the terminal device to obtain the circumferential position of the pore in the ventilation hose 1 and the corresponding total displacement distance when the pore appears, and quickly reach the position of the ventilation hose 1 where the pore is located.

[0050] Embodiment 2

[0051] The difference between the system of this embodiment and the first embodiment is that: Figure 3 As shown:

[0052] The umbrella frame includes an umbrella pole 15 and eight umbrella ribs 16. One end of the umbrella rib 16 is hinged to one end of the umbrella pole 15, and the other end of the umbrella rib 16 is a free end. Two of the umbrella ribs 16 are fixed to the free ends of the umbrella ribs 16. The umbrella cover 21 is an octagonal nylon cloth 17. There is a connecting hole on each of the eight corners of the umbrella cover 21. The umbrella rib 16 passes through the connecting hole and is tied and fixed to the free end of the umbrella rib 16 through the connecting hole with a cable tie. The umbrella pole 15 is rotatably connected to a rotating drum 19. There are eight nylon ropes 18 on the rotating drum 19. One end of the nylon ropes 18 are fastened to the rotating drum 19, and the other end of the nylon ropes 18 are fixedly connected to the free ends of the umbrella ribs 16. A second worm gear is coaxially fixedly connected to one end surface of the rotating drum 19, and the second worm gear is rotatably connected to the umbrella pole 15. A second servo motor is fixed to the rotating drum 19, and the output end of the second servo motor is connected to the second worm through a bevel gear, and the second worm gear is meshed with the second worm, and the second servo motor is electrically connected to the central control system. The traction rope 10 is connected to the umbrella pole 15, and the end of the umbrella pole 15 away from the hinged end of the umbrella rib 16 is facing away from the support frame.

[0053] When the wind blows toward the umbrella surface 21, the umbrella surface 21 and the ribs 16 are propped up. The ribs 16 with the slides 20 are heavier and automatically deflect to the bottom, so as to avoid direct friction between the free ends of the ribs 16 and the inner surface of the ventilation hose 1. The wind blows the umbrella surface 21 to generate traction to pull the entire device to move. When the contact speed sensor 14 detects that the speed is too fast, the central control system controls the second servo motor to rotate, thereby driving the rotating drum 19 to rotate, so that the nylon rope 18 is wound around the rotating drum 19 for a certain length, thereby driving the free ends of the ribs 16 to shrink inward, reducing the area of ​​the windward surface, and thus achieving speed reduction; when the nylon rope 18 is completely wound around the rotating drum 19, the umbrella surface 21 is completely retracted, and the traction of the entire device is reduced to a minimum and is not enough to overcome the friction of the ventilation hose 1, and the entire device stops moving.

[0054] The method of this embodiment is different from that of the first embodiment in that, in the speed adjustment step, the central control system determines whether the current moving speed is within the preset range. If not, the second servo motor is controlled to rotate based on the current moving speed, and this step is repeated until the current moving speed is within the preset range. By dynamically adjusting the rotation of the second servo motor, the speed can be kept within the preset range, avoiding blurring of photos due to excessive speed or affecting detection efficiency due to excessive slow speed.

[0055] The above are only embodiments of the present invention. The invention is not limited to the fields involved in this implementation case. The common sense such as the known specific structures and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field to which the invention belongs before the application date or the priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to explain the content of the claims.

Claims

1. A plateau tunnel ventilation hose air leakage detection device based on visual recognition, characterized in that: It includes a support frame, a traction rope and a traction parachute. The support frame includes a central axis. A plurality of spring rods are radially arranged at both ends of the central axis. Support plates are fixedly connected to the tops of the spring rods. The support plates at the same end of the central axis are distributed in a circle. A battery and a plurality of cameras are installed in the middle of the central axis. The cameras are evenly distributed along the circumference of the central axis. The sum of the viewing angles of all the cameras is not less than 360°. The traction parachute includes an umbrella frame and an umbrella surface. One end of the traction rope is connected to the central axis, and the other end is connected to the umbrella frame. It also includes a central control system for receiving camera photos and visually identifying whether there are pores in the photos, the central control system and the camera are powered by a battery, and the central control system is connected to an audible and visual alarm device; The umbrella frame is a circular frame, and two mounting seats are arranged inside the umbrella frame. The two mounting seats are arranged opposite to each other, and the two mounting seats are connected to a rotating shaft for common rotation. The umbrella surface is a circular plate with a diameter smaller than the inner diameter of the umbrella frame. The umbrella surface is fixed on the rotating shaft, and the umbrella surface is parallel to the rotating shaft. One end of the rotating shaft is connected to a first servo motor, and the first servo motor is electrically connected to the central control system. Or the umbrella frame includes an umbrella pole and a plurality of umbrella ribs, one end of the umbrella rib is connected to one end of the umbrella pole, and the other end of the umbrella rib is a free end, the umbrella cover is a flexible cloth, and the edge of the umbrella cover is evenly provided with connecting parts with the same number as the number of umbrella ribs, and the connecting parts are fixedly connected to the free ends of the umbrella ribs; a rotating drum is rotatably connected to the umbrella pole, and the rotating drum is connected with the same number of flexible ropes as the number of umbrella ribs, and the flexible ropes are respectively fixedly connected to the free ends of the umbrella ribs, and a worm gear is coaxially fixedly connected to the rotating drum, and the worm gear is rotatably connected to the umbrella pole, and a second servo motor is fixed on the rotating drum, and the output end of the second servo motor is connected to a worm, and the worm gear is meshed with the worm, and the second servo motor is electrically connected to the central control system.

2. The plateau tunnel ventilation hose air leakage detection device based on visual recognition according to claim 1 is characterized by: It also includes a contact speed sensor, which is installed on the supporting sheet and transmits the signal of the contact speed sensor to the central control system.

3. The plateau tunnel ventilation hose air leakage detection device based on visual recognition according to claim 2 is characterized by: A sled is also connected to the support sheet, and the length direction of the sled is parallel to the axis direction of the central axis.

4. The plateau tunnel ventilation hose air leakage detection device based on visual recognition according to claim 3 is characterized by: A counterweight is fixed on the side of the central axis, and the counterweight is located on the same side of the central axis.

5. A method for detecting air leakage in ventilation hoses in plateau tunnels based on visual recognition, using the detection device described in claim 1, characterized in that: include: Collection steps: The central control system obtains photos from the camera; Identification step: pre-process the photo, identify the pre-processed photo, and determine whether there is a pore in the current photo based on the identification result. If there is a pore, control the sound and light alarm device to alarm.

6. The method for detecting air leakage of ventilation hoses in plateau tunnels based on visual recognition according to claim 5 is characterized in that: In the acquisition step, the central control system also obtains the current moving speed from the contact speed sensor; It also includes: a shooting adjustment step: the central control system calculates the moving distance L1 based on the current moving speed and the current shooting interval of the camera, and determines whether the moving distance L1 is greater than the pre-stored length L2 of the camera shooting area. If greater, the camera's shooting interval is recalculated based on the current moving speed and the length L2 of the camera shooting area, and the calculated shooting interval is used to update the camera's current shooting interval.

7. The method for detecting air leakage of ventilation hoses in plateau tunnels based on visual recognition according to claim 6 is characterized in that: In the acquisition step, the central control system also records the displacement time; It also includes a positioning step, in which the central control system calculates the total displacement distance based on the moving speed and the displacement time; The central control system also determines the position of the pore in the photo; the central control system also obtains the photographing area of ​​the camera corresponding to the photo, and determines the circumferential position of the pore in the ventilation hose based on the photographing area of ​​the camera and the position of the pore in the photo; The central control system stores the circumferential position of the pores in the ventilation hose and the corresponding total displacement distance when the pores appear.

8. The method for detecting air leakage of ventilation hoses in plateau tunnels based on visual recognition according to claim 7 is characterized in that: It also includes a speed adjustment step: the central control system determines whether the current moving speed is within a preset range. If not, the first servo motor or the second servo motor is controlled to rotate based on the current moving speed, and this step is repeated until the current moving speed is within the preset range.

Images