Banana sieve fire prevention and control system and working method

Abstract

The invention provides a banana sieve fire prevention and control system and a working method. The banana sieve fire prevention and control system comprises: a fire source detection unit for detecting fire source information during operation of a banana sieve; the edge computing terminal is communicated with the fire source detection unit, receives the fire source information and can determine the position of an ignition point and judge the type of a fire behavior; the multi-mode directional fire extinguishing unit communicates with the edge computing terminal and comprises a jet assembly, and the jet assembly comprises at least two fire extinguishing media adaptive to different fire behaviors; the edge computing terminal can drive the multi-mode directional fire extinguishing unit to drive the spraying assembly to be aligned with an ignition point, and corresponding fire extinguishing media are controlled to be released according to fire behavior types. According to the invention, the injection assembly is adapted to different fire behaviors; the fire behavior of the electrical equipment releases inert gas, so that secondary electrical accidents are avoided; slight fire on the screen surface and at the joint releases water mist for cooling and fire control, so that waste is reduced; in a major fire, water mist is firstly used for cooling, then inert gas is used for isolating oxygen, and both rapid fire control and reburning inhibition are considered.

Description

Technical Field

[0001] This invention relates to the field of safety protection technology for mineral processing equipment, specifically to a banana screen fire prevention and control system and its operating method. More particularly, it relates to a banana screen fire prevention and control system and method based on dual-band thermal imaging and location-based fire suppression. Background Technology

[0002] Banana screens, commonly used mineral processing equipment in the mining and metallurgical industries, are prone to generating localized high temperatures exceeding 200°C during operation due to friction between the screen and materials, bearing overload, or electrical short circuits, posing a fire hazard. However, existing fire detection and extinguishing technologies have significant shortcomings, making it impossible to accurately control different types of fires involving banana screens.

[0003] Existing fire suppression solutions mostly employ a single sprinkler system, which can only spray a single medium through fixed nozzles, making it unsuitable for the diverse fire conditions required by banana screens. For example, indiscriminately spraying water mist can easily cause secondary short circuits in cases of fires involving electrical equipment such as motors and cables; using strong extinguishing media directly for minor fires on the screen surface and connections would be wasteful; and for major fires on the screen surface and connections, a single medium cannot simultaneously achieve rapid cooling and fire control while isolating oxygen to prevent reignition, resulting in low fire suppression efficiency.

[0004] A Chinese patent with publication number CN209657458U discloses a spark detection and prevention device and a hanging screen machine using the same. The hanging screen machine is equipped with a fire suppression system, which consists of an automatic fire suppression system and a manual fire suppression system, each equipped with a one-way valve. The fire suppression system also includes multiple high-pressure sprinklers distributed on the top and / or around the hanging screen machine.

[0005] However, the patent document CN209657458U has obvious shortcomings: its fire protection system cannot select the appropriate fire extinguishing medium according to different fire types; especially for major fires, it cannot achieve rapid fire control and effective suppression of fire reignition through the graded use of multiple media. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the purpose of this invention is to provide a banana screen fire prevention and control system and its working method.

[0007] A banana screen fire prevention system according to the present invention includes: The fire source detection unit is used to detect fire source information during the operation of the banana screen; An edge computing terminal is communicatively connected to the fire source detection unit and receives fire source information output by the fire source detection unit. The edge computing terminal can determine the location of the ignition point on the banana sieve based on the fire source information and judge the type of fire. A multimodal directional fire suppression unit, communicatively connected to the edge computing terminal, includes: a spray assembly, wherein the spray assembly includes at least two fire suppression media adapted to different fire types; The edge computing terminal can drive the multimodal directional fire extinguishing unit to aim the spray assembly at the fire location, and control the spray assembly to release the corresponding fire extinguishing medium according to the fire type.

[0008] Preferably, the spraying assembly includes: a high-pressure fine water mist module and an inert gas module; The high-pressure fine water mist module includes an ultra-high pressure pump and a high-pressure fine water mist nozzle. The pressure of the ultra-high pressure pump is not less than 10MPa, and the nozzle diameter of the high-pressure fine water mist nozzle is 0.3mm, with a droplet size of 50-100μm. The inert gas module includes an inert gas nozzle and a compressed nitrogen tank, wherein the nitrogen in the compressed nitrogen tank has a purity of >99.5%.

[0009] Preferably, the types of fires include: motor fires, electrical control box fires, cable fires, major fires on the screen surface, major fires at connections, major fires at structural points, minor fires on the screen surface, minor fires at connections, and minor fires at structural points. When the fire type is a motor fire, an electrical control box fire, or a cable fire, the edge computing terminal controls the inert gas module to release inert gas toward the ignition point. When the fire type is a minor fire on the screen surface, a minor fire at the connection point, or a minor fire at the structure, the edge computing terminal controls the high-pressure fine water mist module to release fine water mist toward the fire point. The criteria for judging the three types of minor fires are: the fire point temperature is 20-50°C higher than the ambient temperature, the fire area is less than 10% of the screen surface area of ​​the banana screen, and the temperature rise rate of the fire point is 1-3°C / minute. When the fire type is a major fire on the screen surface, a major fire at the connection point, or a major fire at the structure, the edge computing terminal first controls the high-pressure fine water mist module to release fine water mist toward the ignition point, and then controls the inert gas module to release inert gas toward the ignition point. The criteria for judging the three types of major fires are: the ignition point temperature is more than 50°C higher than the ambient temperature, the ignition area is greater than or equal to 10% of the screen surface area of ​​the banana screen, and the temperature rise rate of the ignition point is greater than 3°C / minute.

[0010] Preferably, the fire source detection unit includes a dual-band infrared thermal imager, which integrates a mid-wave infrared imaging module and a long-wave infrared imaging module. The fire source information is a temperature field distribution map of the banana sieve surface scanned by a dual-band infrared thermal imager. The edge computing terminal integrates a three-dimensional coordinate mapping algorithm, which can convert the coordinates of the fire point detected by the dual-band infrared thermal imager into coordinates in the three-dimensional coordinate system of the banana sieve surface.

[0011] Preferably, the wavelength of the mid-wave infrared imaging module is 3-5μm, and the temperature detection range is 200-600℃; The long-wave infrared imaging module has a wavelength of 8-14μm and is used to compensate for the ambient temperature of the measurement results of the mid-wave infrared imaging module.

[0012] Preferably, it also includes: a vibration compensation pan-tilt unit, and a dual-band infrared thermal imager is deployed on the vibration compensation pan-tilt unit; The vibration compensation gimbal is equipped with an inertial navigation unit, which can counteract the image jitter of the dual-band infrared thermal imager caused by the vibration of the banana screen body in real time.

[0013] Preferably, it further includes: an audible and visual alarm, which is communicatively connected to the edge computing terminal; When the edge computing terminal determines that a fire has occurred, the audible and visual alarm can issue corresponding voice broadcasts and light reminders according to the type of fire.

[0014] Preferably, the multimodal directional fire extinguishing unit includes a multi-axis robotic arm and a guide rail, wherein the spraying assembly is disposed on the multi-axis robotic arm, and the multi-axis robotic arm is slidably connected to the guide rail and is capable of moving along the guide rail.

[0015] According to the present invention, a method for preventing and controlling banana screen fires, based on the aforementioned banana screen fire prevention and control system, includes the following steps: S1: The fire source detection unit detects fire source information during the operation of the banana screen and transmits the fire source information to the edge computing terminal; S2: After receiving the fire source information using the edge computing terminal, the location of the ignition point on the banana screen is determined through its integrated fire source positioning function, and the type of fire is judged. S3: Using an edge computing terminal, based on the determined location of the fire point, the multi-axis robotic arm is driven to align the spray assembly with the fire point. At the same time, the appropriate fire extinguishing medium module in the spray assembly is selected according to the type of fire, and the fire extinguishing medium module is controlled to release the fire extinguishing medium.

[0016] Preferably, in S1, the fire source detection unit is a dual-band infrared thermal imager. The dual-band infrared thermal imager scans the sieve surface of the banana sieve at a frequency of 10Hz, generates a temperature field distribution map in real time, and transmits the temperature field distribution map information to the edge computing terminal. In S2, the edge computing terminal reads the temperature field distribution map information and filters transient high-temperature noise points through the integrated AI fire identification algorithm. The AI ​​fire identification algorithm includes: determining points with a temperature rise rate > 20℃ / s detected in more than 3 consecutive frames as temperature rise points; if the total area growth rate of temperature rise points is greater than zero, a fire is determined to have occurred, and the temperature rise point is the ignition point. After a fire is determined, the edge computing terminal immediately issues a power cut-off command and transmits a start signal to the audible and visual alarm. In addition, the edge computing terminal uses a three-dimensional coordinate mapping algorithm to match the three-dimensional spatial coordinates of the ignition point detected by the dual-band infrared thermal imager with the built-in coordinates of the multi-axis robotic arm to determine the position of the ignition point in the multi-axis robotic arm coordinate system. In S3, when the fire type is a major fire on the screen surface, a major fire at the connection point, or a major fire at the structure, a graded fire suppression system is adopted. The graded fire suppression system is specifically as follows: Level 1 response, within 1 second, the high-pressure fine water mist module is controlled to spray water mist for local cooling; Level 2 blocking, within 5 seconds, the inert gas module is controlled to release nitrogen gas, thereby isolating oxygen.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention, by setting up at least two spray components adapted to different fire types in a multimodal directional fire suppression unit, and by using an edge computing terminal to achieve precise location of the ignition point and determination of the fire type, can release corresponding fire suppression media for different types of fires. This can not only avoid secondary accidents caused by mismatched fire suppression media, but also accurately select the appropriate fire suppression media and fire suppression strategy for minor and major fires respectively. In minor fires, it can reduce the waste of fire suppression media, and in major fires, it can balance rapid fire control and effective suppression of reignition.

[0018] 2. This invention uses a three-dimensional coordinate mapping algorithm to solve the positioning deviation of the ignition point caused by the tilting and vibration of the vibrating screen, so that the spraying component can accurately track the ignition point on the banana screen, avoiding blind spraying that causes equipment corrosion, material agglomeration, or even secondary electrical accidents.

[0019] 3. This invention improves the accuracy of temperature field detection in dusty environments by using a dual-band infrared thermal imager that integrates mid-wave and long-wave infrared measurements, thus avoiding misjudgment or missed detection of fires caused by mineral dust obstruction. Attached Figure Description

[0020] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a diagram showing the overall system layout in an embodiment of the present invention; Figure 2 This is a structural diagram of the multimodal fire extinguishing unit in an embodiment of the present invention.

[0021] As shown in the figure: Detailed Implementation

[0022] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0023] This invention provides a dedicated fire extinguishing system for banana screens, integrating non-contact thermal imaging monitoring, fire location algorithms, and intelligent sprinkler functions. It is suitable for fire prevention and emergency response in high-dust and high-vibration environments such as mines and metallurgy, and can solve the fire prevention and control problems under the complex working conditions of banana screens, achieving precise fire extinguishing, reducing equipment damage, and improving safety.

[0024] The core components and functions of this system are as follows: The dual-band infrared thermal imager 1 integrates mid-wave infrared and long-wave infrared imaging functions. The mid-wave infrared wavelength is 3-5μm, used to detect high temperatures of 200-600°C; the long-wave infrared wavelength is 8-14μm, used to compensate for environmental temperature fluctuations in the measurement results of the mid-wave infrared imaging module. The compensation principle is as follows: in the working environment of the banana sieve, mineral dust can cause localized temperature fluctuations, affecting the detection accuracy of the mid-wave infrared wavelength. The long-wave infrared sensor can collect the ambient reference temperature around the sieve in real time. The edge computing terminal 3 uses the ambient temperature data collected by the long-wave infrared sensor to correct the fire source temperature detected by the mid-wave infrared sensor, thereby eliminating environmental temperature interference, accurately identifying the real fire source, and avoiding misjudgment or missed detection of fires.

[0025] The dual-band infrared thermal imager 1 can improve dust penetration and anti-dust interference capabilities, and meets the IP67 protection level and has a temperature range of -30°C to 80°C. It can be fixedly installed 1.5 to 2m above the banana sieve surface, and its monitoring range can cover the entire banana sieve.

[0026] The dual-band infrared thermal imager 1 can be deployed on a vibration-compensating gimbal 2 for mounting on a fixed surface. The vibration-compensating gimbal 2 can be equipped with an inertial navigation unit (IMU) that meets IP67 protection standards and has a temperature range of -30°C to 80°C, and can compensate for the thermal image jitter caused by the vibration of the banana screen in real time.

[0027] Meanwhile, to further optimize the accuracy of ignition point positioning, angular velocity sensors and acceleration sensors can be fixedly installed on the banana screen body. The angle sensor is used to collect the tilt angle of the screen body in real time, and the acceleration sensor is used to collect the vibration acceleration of the screen body in real time. These two types of data constitute the dynamic compensation data of the screen body. Both the angle sensor and the acceleration sensor can communicate with the edge computing terminal 3 via Modbus TCP wireless communication 6, transmitting the collected dynamic compensation data of the screen body to the edge computing terminal 3 in real time. The edge computing terminal 3 can calculate the vibration trajectory of the vibrating screen through vibration acceleration.

[0028] Edge computing terminal 3 can have locally deployed lightweight AI models, integrating AI fire detection algorithms and 3D coordinate mapping algorithms. The AI ​​fire detection algorithm can distinguish between normal mechanical friction hotspots and combustion characteristics, and the judgment criteria can include flame spread pattern and temperature rise rate.

[0029] The three-dimensional coordinate mapping algorithm includes: first, establishing a three-dimensional coordinate system for the banana screen surface using laser calibration equipment; then, creating a 3D geometric model of the screen surface on this coordinate system as the reference for coordinate mapping. Key parameters include: the screen surface's length, width, tilt angle, vibration amplitude, vibration frequency, vibration horizontality, and vibration verticality. The screen surface's length and width are standard dimensions and can be fixed values. The tilt angle is measured using an angle sensor. The vibration amplitude and frequency can be measured using an accelerometer. The vibration horizontality and vibration verticality can be assumed to be constants or measured using a displacement sensor.

[0030] Secondly, the coordinates of the ignition point (three-dimensional spatial point P=(x,y,z)) in the thermal imaging image detected by the dual-band infrared thermal imager 1 are projected to the pixel coordinate system (p=(u,v)), and combined with the above-mentioned dynamic compensation data of the sieve body and the coordinates of the coordinate system of the multi-axis robotic arm 404.

[0031] The method of combining the dynamic compensation data of the screen body includes: when converting the coordinates of the ignition point in the thermal imaging image to the coordinates of the multi-axis robotic arm 404, the edge computing terminal 3 calls the tilt angle data of the screen body transmitted by the angle sensor to correct the spatial position of the ignition point in the three-dimensional coordinate system of the screen surface; at the same time, it calls the vibration trajectory data transmitted by the accelerometer to predict the position of the screen body vibration at the next moment, and superimposes the predicted position into the coordinate transformation process of the multi-axis robotic arm 404 to achieve accurate mapping of the ignition point in the coordinate system of the multi-axis robotic arm 404 (error ≤ 5cm). By using a locally deployed lightweight AI model, the edge computing terminal 3 can reduce the network dependence of this banana screen fire prevention and control system, and the fire response time is < 0.5s.

[0032] The core of the multimodal directional fire suppression unit 4 is a multi-axis robotic arm 404, which has its own positioning coordinates. The multi-axis robotic arm 404 can be a six-axis robotic arm. The spray assembly is mounted on the multi-axis robotic arm 404, and includes a high-pressure fine water mist module and an inert gas module. The multimodal directional fire suppression unit 4 is also equipped with a guide rail 403. The multi-axis robotic arm 404 is slidably connected to the guide rail 403 and can move along the guide rail 403. The guide rail increases the range of motion of the multi-axis robotic arm 404, providing more spray angles for the spray assembly and improving fire suppression accuracy.

[0033] The aforementioned high-pressure fine water mist module includes a water storage tank, an ultra-high-pressure pump, and a high-pressure fine water mist nozzle 402. The ultra-high-pressure pump has a pressure of not less than 10 MPa, and the nozzle diameter of the high-pressure fine water mist nozzle 402 is 0.3 mm, with a droplet size of 50–100 μm. The fine water mist driven by ultra-high pressure exhibits excellent atomization effect, rapidly absorbs heat for cooling, and the small droplet size reduces the impact of high-dust environments in mines, ensuring high fire extinguishing efficiency.

[0034] The inert gas module includes an inert gas nozzle 401 and a compressed nitrogen tank. The inert gas module uses a compressed nitrogen tank with a purity >99.5%. High-purity nitrogen can quickly remove oxygen from the vicinity of the fire source, inhibit combustion, and leaves no impurities after extinguishing the fire. It will not corrode electrical equipment contacts or contaminate the ore, reducing equipment and production maintenance costs after gas release.

[0035] The multi-axis robotic arm 404 can automatically adjust the joint angles according to the coordinates of the fire point, thereby controlling the angle of the spray components. It can also adaptively adjust the flow rate of the spray components to match different fire extinguishing needs. In the early stages of a fire, water mist can be used for cooling. If the fire continues for more than a preset time, inert gas can be switched on to suppress reignition.

[0036] The audible and visual alarm 5 can be installed near the banana screen, 2 meters above the ground, in an unobstructed location. It features high-powered audio broadcasting and high-intensity light, unaffected by dust or noise. It has a built-in voice preset function, capable of issuing corresponding voice announcements based on different fire types. These fire types include: motor fire, electrical control box fire, cable fire, major fire on the screen surface, major fire at connections, major fire at structural points, minor fire on the screen surface, minor fire at connections, and minor fire at structural points. The aforementioned connections refer to the joints between different structural components of the banana screen, primarily used for fixing or transmitting power, such as the flexible connection between the reducer and the vibrating screen, and the connection between the reducer and the motor. The aforementioned structural points refer to the main structural components of the banana screen, mostly the core load-bearing or functional components of the banana screen equipment, such as the screen box, side plates, crossbeams, support frames, and other structural parts, as well as vibrator support springs and shock absorbers.

[0037] Pre-programmed sounds may include: fire announcements from motors, control boxes, cables, screen surfaces, connections, and structures.

[0038] The criteria for determining whether a minor fire on the screen surface, at a connection point, or at a structural point is established are: the ignition point temperature is 20-50°C higher than the ambient temperature, the ignition area is less than 10% of the screen surface area of ​​the banana screen, and the temperature rise rate of the ignition point is 1-3°C / minute.

[0039] The criteria for determining whether a major fire is occurring on the screen surface, at a connection point, or at a structural point are: the temperature at the ignition point is more than 50°C higher than the ambient temperature, the ignition area is greater than or equal to 10% of the screen surface area of ​​the banana screen, and the temperature rise rate at the ignition point is greater than 3°C / minute.

[0040] Modbus TCP wireless communication 6 is used to provide a standardized interconnection method for various devices in the system (dual-band infrared thermal imager 1, edge computing terminal 3, audible and visual alarm 5, etc.), supporting real-time and reliable data transmission and command interaction.

[0041] The fire prevention and control process and firefighting strategy of this system are as follows, and the specific fire handling procedures are shown in Table 1.

[0042] Table 1 Fire Prevention and Control Strategies

[0043] Specific work steps: S1: The dual-band infrared thermal imager 1 scans the screen surface at a frequency of 10Hz and generates a temperature field distribution map in real time; and transmits the temperature field distribution map information to the edge computing terminal 3 through Modbus TCP wireless communication 6; at the same time, the angle sensor and the acceleration sensor collect the tilt angle and vibration acceleration data of the screen body respectively; the above temperature field data and screen body dynamic data are all transmitted to the edge computing terminal 3 through Modbus TCP wireless communication 6.

[0044] S2: Edge computing terminal 3 first corrects the fire source temperature detection value of mid-wave infrared by using the ambient temperature data of long-wave infrared to eliminate environmental interference; then it uses AI fire identification algorithm to filter transient high temperature noise (such as sparks from ore impact). The AI ​​fire identification algorithm includes: determining the ignition point with a temperature rise rate > 20℃ / s detected in more than 3 consecutive frames as a temperature rise point. If the total area growth rate of the temperature rise point is greater than zero, it is determined that a fire has occurred. The temperature rise point is the ignition point.

[0045] Upon determining that a fire has occurred, the system immediately issues a command to cut off the power to the banana screen; simultaneously, it immediately transmits a control signal to the audible and visual alarm 6 to activate the alarm. The audible and visual alarm 6 provides corresponding voice and light alerts for different types of fires.

[0046] Edge computing terminal 3 uses a three-dimensional coordinate mapping algorithm to map the three-dimensional spatial coordinates of the ignition point to the built-in coordinates of the multi-axis robotic arm 404, thus determining the position of the ignition point within the coordinate system of the multi-axis robotic arm 404. During this process, dynamic compensation of the ignition point coordinates can be performed by combining the screen tilt angle and vibration trajectory data, improving the positioning accuracy of the ignition point within the coordinate system of the multi-axis robotic arm 404. This avoids ignition point positioning deviations caused by dynamic screen surfaces, preventing inaccurate spraying by the spraying components, which could lead to equipment corrosion, material agglomeration, and secondary electrical accidents.

[0047] S3: The servo motor drives the multi-axis robotic arm 404 to align with the coordinates of the fire point. The multi-axis robotic arm 404 can move along the guide rail 403 to the vicinity of the fire point and start the preset fire extinguishing program. The fire extinguishing program is stored in the edge computing terminal 3.

[0048] S4: The fire extinguishing procedure can perform fire extinguishing operations according to the type of fire: For fires involving electrical equipment, such as motors, cables, or electrical control boxes, the inert gas module is controlled to release inert gas toward the fire source; for minor fires on the screen surface, joints, or structures, the high-pressure fine water mist module is controlled to release fine water mist toward the fire source; for major fires on the screen surface, joints, or structures, a graded fire extinguishing strategy is adopted, including: Level 1 response: Within 1 second, the high-pressure fine water mist module is controlled to spray water mist locally for cooling; Level 2 blocking: Within 5 seconds, the inert gas module is controlled to release nitrogen to isolate oxygen and simultaneously cut off the power supply to the screen body.

[0049] Since the dual-band infrared thermal imager 1 is installed on a fixed surface, the positions of motors, cables, electrical control boxes, screens, connections, and structures can be pre-set in the edge computing terminal 3 during debugging. Thus, when a fire point is detected, the location of the fire can be determined based on the three-dimensional coordinates of the fire point, thereby determining the type of fire and selecting the corresponding fire extinguishing strategy.

[0050] In the description of this application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application 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, they should not be construed as limitations on this application.

[0051] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A banana fire screening system, characterized in that, The application relates to a banana screen fire extinguishing system, which comprises the following parts: a fire source detection unit for detecting fire source information during the operation of the banana screen; an edge computing terminal (3) in communication connection with the fire source detection unit and receiving the fire source information output by the fire source detection unit, wherein the edge computing terminal (3) can determine the fire point position on the banana screen according to the fire source information and judge the fire type; a multi-modal directional fire extinguishing unit (4) in communication connection with the edge computing terminal (3), which comprises a spraying assembly, wherein the spraying assembly comprises at least two kinds of fire extinguishing media suitable for different fire types; the edge computing terminal (3) can drive the multi-modal directional fire extinguishing unit (4) to aim at the fire point position and control the spraying assembly to release the corresponding fire extinguishing medium according to the fire type. the spraying assembly comprises a high-pressure fine water mist module and an inert gas module; 2. The banana fire prevention and control system of claim 1, wherein, the high-pressure fine water mist module comprises an ultrahigh-pressure pump and a high-pressure fine water mist nozzle (402), wherein the pressure of the ultrahigh-pressure pump is not less than 10 MPa, the jet hole diameter of the high-pressure fine water mist nozzle (402) is 0.3 mm, and the spray droplet particle size is 50-100 mu m; the inert gas module comprises an inert gas nozzle (401) and a compressed nitrogen tank, and the purity of nitrogen in the compressed nitrogen tank is greater than 99.5%. the fire types include motor fire, electric control box fire, cable fire, screen surface major fire, connection major fire, structure major fire, screen surface slight fire, connection slight fire and structure slight fire; 3. The banana fire prevention and control system of claim 2, wherein, when the fire type is motor fire, electric control box fire or cable fire, the edge computing terminal (3) controls the inert gas module to release inert gas towards the fire point position; when the fire type is screen surface slight fire, connection slight fire or structure slight fire, the edge computing terminal (3) controls the high-pressure fine water mist module to release fine water mist towards the fire point position, and the judgment basis for the three kinds of slight fires is that the temperature of the fire point is 20-50 DEG C higher than the ambient temperature, the fire area is less than 10% of the screen surface area of the banana screen, and the temperature rising rate of the fire point is 1-3 DEG C / min; when the fire type is screen surface major fire, connection major fire or structure major fire, the edge computing terminal (3) first controls the high-pressure fine water mist module to release fine water mist towards the fire point position, and then controls the inert gas module to release inert gas towards the fire point position, and the judgment basis for the three kinds of major fires is that the temperature of the fire point is more than 50 DEG C higher than the ambient temperature, the fire area is greater than or equal to 10% of the screen surface area of the banana screen, and the temperature rising rate of the fire point is greater than 3 DEG C / min. the fire source detection unit comprises a dual-band infrared thermal imager (1), wherein the dual-band infrared thermal imager (1) integrates a middle-wave infrared imaging module and a long-wave infrared imaging module; 4. The banana screen fire prevention system of claim 1, wherein, the fire source information is a banana screen screen surface temperature field distribution diagram scanned by the dual-band infrared thermal imager (1); the edge computing terminal (3) integrates a three-dimensional coordinate mapping algorithm and can convert the fire point coordinates detected by the dual-band infrared thermal imager (1) into the coordinates in the three-dimensional coordinate system of the banana screen screen surface. ​ 5. The banana screen fire prevention system of claim 4, wherein, The wavelength of the middle wave infrared imaging module is 3-5 μm, and the temperature detection range is 200-600 ℃; The wavelength of the long wave infrared imaging module is 8-14 μm, and is used for environmental temperature compensation of the measurement result of the middle wave infrared imaging module.

6. The banana screen fire prevention system of claim 4, wherein, Further comprising: A vibration compensation holder (2), wherein the dual-band infrared thermal imager (1) is arranged on the vibration compensation holder (2); The vibration compensation holder (2) is provided with an inertial navigation unit, and can offset the image jitter of the dual-band infrared thermal imager (1) caused by the vibration of the banana screen body in real time.

7. The banana screen fire prevention system of claim 1, wherein, Further comprising: An audible and visual alarm (5) in communication connection with the edge computing terminal (3); When the edge computing terminal (3) determines that a fire occurs, the audible and visual alarm (5) can issue corresponding voice broadcast and light reminder according to the fire type.

8. The banana screen fire prevention system of claim 1, wherein, The multi-modal directional fire extinguishing unit (4) comprises a multi-axis mechanical arm (404) and a guide rail (403), wherein the spraying assembly is arranged on the multi-axis mechanical arm (404), and the multi-axis mechanical arm (404) is slidingly connected to the guide rail (403) and can move along the guide rail (403).

9. A method for preventing and controlling fire in a banana screen, characterized by, The banana screen fire prevention and control system according to any one of claims 1 to 8 comprises the following steps: S1: detecting fire information in the operation process of the banana screen by using a fire source detection unit, and transmitting the fire information to the edge computing terminal (3); S2: determining the position of the ignition point on the banana screen and judging the fire type by using the fire source positioning function integrated in the edge computing terminal (3) after receiving the fire information; S3: driving the multi-axis mechanical arm (404) to align the spraying assembly to the position of the ignition point according to the determined position of the ignition point, and selecting the corresponding fire extinguishing medium module in the spraying assembly according to the fire type, and controlling the fire extinguishing medium module to release the fire extinguishing medium.

10. The method of claim 9, wherein the banana screen fire prevention and control work method is characterized by, In S1, the fire source detection unit is a dual-band infrared thermal imager (1), which scans the screen surface of the banana screen at a frequency of 10 Hz, generates a temperature field distribution map in real time, and transmits the temperature field distribution map information to the edge computing terminal (3); In S2, the edge computing terminal (3) reads the temperature field distribution map information, filters transient high-temperature noise points by using an AI fire identification algorithm integrated therein, the AI fire identification algorithm comprises: determining a point with a temperature rise rate > 20 ℃ / s as a temperature rise point if more than 3 consecutive frames are detected, and determining that a fire occurs if the total area growth rate of the temperature rise point is greater than zero, the temperature rise point is the ignition point, after determining that a fire occurs, the edge computing terminal (3) immediately issues a power-off instruction, and transmits a start signal to the audible and visual alarm (5), in addition, the edge computing terminal (3) determines the position of the ignition point in the coordinate system of the multi-axis mechanical arm (404) by using a three-dimensional coordinate mapping algorithm, and corresponds the three-dimensional space coordinates of the ignition point detected by the dual-band infrared thermal imager (1) to the coordinates of the multi-axis mechanical arm (404). In S3, when the fire type is a screen major fire, a connection major fire or a structure major fire, hierarchical fire extinguishing is adopted, and the hierarchical fire extinguishing is specifically: first-level response, controlling the high-pressure water mist module to spray water mist for local cooling within 1s; second-level blocking, controlling the inert gas module to release nitrogen within 5s, so as to isolate oxygen.

Images