Flame detector

Abstract

The flame detector comprises a polygonal shell, a mounting rod is arranged at the bottom of the polygonal shell, the polygonal shell is provided with an inner cavity, and an MCU module is connected with an artificial intelligence AI module, a three-band infrared sensor module, an ultraviolet sensor module, a camera module and a G CAT-module. A power supply module is connected with the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the GCAT-module, and the side faces of the transverse plate, the vertical plate and the inclined plate are fixedly connected with the inner wall of the polygonal shell. According to the invention, the power supply module, the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module are installed by using the polygonal shell; the composite flame detector with the photographing image recognition function has the advantages of being long in endurance time, free of false alarm, high in anti-interference capacity, short in response time and capable of automatically recognizing and judging fire behaviors on pictures.

Description

Technical Field

[0001] The invention relates to the field of forest surface flame detection, and in particular to a flame detector. Background Art

[0002] Forest fire is one of the most serious natural disasters. It occurs all year round and has the characteristics of wide occurrence, strong suddenness, great destructiveness, high risk, particularly difficult to deal with and extinguish, difficult to find geographical location, inconvenient transportation, wide scope, and serious environmental and ecological damage. It seriously endangers people's lives and property and forest resources, and even causes ecological disasters. Controlling forest fires is a global problem and has been highly valued by forestry, environmental protection and other relevant administrative agencies in various countries. How to achieve the standardization, scientificization and informatization of forest fire prevention work, and truly achieve early detection and early resolution of fire hidden dangers, intelligent forest monitoring has become a necessary means to eliminate forest fires. By monitoring the signals generated by the flames, it is determined whether there is a fire, so as to find the source of the fire and alarm immediately.

[0003] After searching, it was found that there are currently very few patents in China for composite flame detectors with red and ultraviolet photography in the field of deep forest fire detection, but they all have certain commonalities. First of all, the power supply problem. Most devices are powered by solar panels and backup lithium batteries, but basically the power charged by the solar panels cannot meet the daily power needs of the equipment, so the backup lithium battery must be used. However, the capacity of the lithium battery is limited and the power will be used up quickly, which greatly reduces the service life of the equipment; the equipment has weak anti-interference ability in strong light sources or in interference environments, and the frequency of false alarms is relatively high; some flames have slow alarm response speeds after detection, and the fire reporting time is relatively long, which may delay early disasters. Some detectors cannot even determine whether a fire has occurred after taking pictures, but they all need to be identified and judged manually, which is prone to missed judgments and misjudgments, ultimately affecting fire rescue. Summary of the invention

[0004] The problem to be solved by the present invention is to provide a flame detector in view of the above-mentioned deficiencies existing in the prior art, which can effectively solve the problems of short power supply endurance time and inability to automatically identify and judge the fire occurring in the picture.

[0005] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a flame detector, comprising a polygonal shell, a mounting rod at the bottom of the polygonal shell, the polygonal shell having an inner cavity, a horizontal plate, a vertical plate and an inclined plate arranged crosswise in the inner cavity, the horizontal plate, the vertical plate and the inclined plate divide the inner cavity into a plurality of sealed installation cavities, the polygonal shell is installed with a power supply module, an MCU module, an artificial intelligence AI module, a three-band infrared sensor module, an ultraviolet sensor module, a camera module and a G CAT-module, the power supply module, the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module are arranged in the plurality of the installation cavities, the MCU module is connected to the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module, and the power supply module is connected to the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module.

[0006] Furthermore, the power supply module includes a solar panel and a rechargeable lithium battery pack connected to each other, the rechargeable lithium battery pack is installed in the installation cavity and connected to the MCU module, artificial intelligence AI module, three-band infrared sensor module, ultraviolet sensor module, camera module and G CAT-module, and the solar panel is installed on the outer wall surface of the polygonal shell.

[0007] Furthermore, the solar panel includes a first solar panel and a second solar panel, the polygonal shell is a regular pentagonal shell, the first solar panel and the second solar panel are respectively mounted on two opposite inclined surfaces at the upper end of the regular pentagonal shell, the first solar panel and the second solar panel are both provided with a plurality of elastic rods, the ends of the elastic rods are provided with barbs, and the regular pentagonal shell is provided with through holes for the elastic rods to pass through.

[0008] Furthermore, the horizontal plate is provided with a groove, the groove is provided with a first recess and a second recess, the first recess and the second recess are respectively provided with a first ball and a second ball, the first ball and the second ball are respectively connected with a first elastic rope and a second elastic rope, the first elastic rope and the second elastic rope have built-in connecting wires, the two connecting wires are respectively connected to the first ball and the second ball, the two connecting wires are respectively connected to an early warning module, the early warning module has a buzzer, and the early warning module is activated when the ball and the second ball collide.

[0009] Furthermore, the depths of the first depression and the second depression are respectively one-quarter to one-third of the diameters of the first ball and the second ball.

[0010] Furthermore, the first ball and the second ball are both iron balls, and arc-shaped magnetic blocks are respectively provided on the first ball and the second ball. The two arc-shaped magnetic blocks are respectively connected to the first ball and the second ball, and lubricants are respectively provided between the arc-shaped magnetic blocks and the first ball and the second ball.

[0011] Compared with the prior art, the present invention has the following beneficial effects: the present invention utilizes the polygonal shell to install the power supply module, MCU module, artificial intelligence AI module, three-band infrared sensor module, ultraviolet sensor module, camera module and G CAT-module, and the composite flame detector based on multi-band infrared sensor, ultraviolet sensor and camera image recognition has the advantages of long battery life, no false alarm, strong anti-interference ability, fast response time, and the ability to automatically identify and judge the fire situation occurring in the picture.

[0012] Other advantages, objectives and features of the present invention will be embodied in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The figure is a schematic diagram of the structure of a flame detector.

[0014] Figure 2 for Figure 1 A partial enlarged schematic diagram.

[0015] Figure 3 It is a side view of the polygonal shell.

[0016] Figure 4 Schematic diagram of the ball structure.

[0017] Figure 5 This is a schematic diagram of the connections of various modules of a flame detector. DETAILED DESCRIPTION

[0018] In order to make the technical means, creative features, objectives and functions of the present invention clearer and easier to understand, the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:

[0019] The present invention proposes a flame detector, comprising a polygonal shell 1, a rod 2 is installed at the bottom of the polygonal shell 1, the polygonal shell 1 has an inner cavity 7, a horizontal plate 4, a vertical plate 3 and an inclined plate 5 are cross-arranged in the inner cavity 7, the horizontal plate 4, the vertical plate 3 and the inclined plate 5 divide the inner cavity 7 into a plurality of sealed installation cavities, the polygonal shell 1 is installed with a power supply module, an MCU module, an artificial intelligence AI module, a three-band infrared sensor module, an ultraviolet sensor module, a camera module and a 4G CAT-1 module, the power supply module, the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module are arranged in the plurality of the installation cavities, the MCU module is connected to the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module, the power supply module is connected to the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module, and the horizontal plate 4, the vertical plate 3 and the inclined plate 5 are fixedly connected to the inner wall of the polygonal shell 1 on the side.

[0020] The power supply module includes two 2W solar panels and a rechargeable lithium battery pack with a capacity of 13200mAH. The solar panel absorbs sunlight and uses the photoelectric effect to charge the lithium battery pack. The lithium battery pack adopts Panasonic's high-quality lithium-ion battery with high energy density, strong stability and reliability. Its cycle charging life can exceed 2,000 times, providing a longer operating time for the device. The power supply module mainly provides power for each module to ensure that each module can operate effectively. The MCU module adopts a 32-bit ultra-low power microcontroller. The controller adopts a powerful 32-bit RISC CPU. 32 is a register and a constant generator, which can maximize code efficiency and is used to control other hardware modules connected to it, including sensor data, power system management, etc.; The artificial intelligence AI module adopts the RV1103 chip and RISC-V architecture, with a computing power of 0.5-1.0TOPS. It is equipped with a high-performance CPU and GPU, which can effectively process complex AI algorithms and a large amount of data, and provide sufficient computing resources to support deep learning, machine learning and other tasks. It mainly judges the flames on the pictures taken by the camera, recognizes and diagnoses the pictures through mathematical models and software algorithms, and outputs true judgment results.

[0021] The judgment is performed on the image by mathematical model and software algorithm processing, which is specifically composed of four parts: model training: first, a labeled dataset containing flame samples is taken. The dataset consists of more than 5,000 flame images, including the bounding box coordinates and category labels of the flame area. The model is trained using these data and the parameters are adjusted to optimize the detection performance; feature extraction: the model extracts image features through a series of convolutional layers. These features can distinguish flames from other objects. The model is designed to efficiently capture the visual features of flames, such as color, shape, and dynamic characteristics; target detection: the model predicts multiple bounding boxes in the image, each of which is accompanied by a confidence score and category probability. For flame detection, the system will screen out boxes with high confidence and belonging to the flame category; post-processing: a self-developed algorithm is used to remove overlapping prediction boxes to ensure that each flame is marked only once. Based on flame image recognition and judgment, deep learning technology is used to provide an automated, efficient and reliable fire detection technology method; the three-band infrared sensor module adopts the Korean Jinlun three-band infrared sensor. This flame sensor uses advanced multi-infrared sensing technology. Three high-sensitivity infrared sensors detect three different infrared wavelengths respectively, and the detection results are analyzed and processed by their own microprocessors to effectively eliminate external interference signals. The processed signal is transmitted to the MCU module; the ultraviolet sensor module detects the flame emitting ultraviolet radiation of a specific wavelength. It adopts the Japanese Hamamatsu R2868 ultraviolet phototube. This phototube has a high signal-to-noise ratio, high reliability, fast response speed, and has the ability to detect extremely weak signals. The monitored ultraviolet data is sent to the MCU control module; the camera module uses a photosensitive chip, supports 5 million pixels, has high sensitivity, high signal-to-noise ratio and low illumination performance, can present a more delicate and more realistic night vision imaging effect, and better adapt to changes in ambient light. The camera sends the collected fire photos to the AI ​​module; the 4G CAT-1 module uses a 4GCAT-1 module for data transmission, which has the advantages of wide network coverage, high reliability, support for multiple IoT protocols, and low latency. It transmits the received data to the server platform through the 4G communication network.

[0022] The power supply module includes a solar panel and a rechargeable lithium battery pack connected to each other. The rechargeable lithium battery pack is installed in the installation cavity and connected to the MCU module, artificial intelligence AI module, three-band infrared sensor module, ultraviolet sensor module, camera module and 4G CAT-1 module. The solar panel is installed on the outer wall surface of the polygonal shell 1.

[0023] The power supply module is connected to each unit module to supply power to each unit module. The signal output end of the three-band infrared sensor module is connected to the MCU module. The signal output end of the ultraviolet sensor module is connected to the MCU module. The output end of the picture taken by the camera module is connected to the AI ​​module. The AI ​​module is connected to the MCU control module. The CAT-1 module is connected to the AI ​​module and the MCU control module. The anti-theft module is connected to the MCU control module. The output end of the main control MCU module is connected to an LED indicator. The flashing of LED indicators of different colors represents different types of alarm data. The flashing of the red LED indicator indicates fire data collection, and the flashing of the blue LED indicator indicates uploading of alarm information.

[0024] The infrared sensor module includes three sensors of different wavebands, the working wavelength of the sensor of the first waveband is 3.6um, the working wavelength of the sensor of the second waveband is 4.4um, and the working wavelength of the sensor of the third waveband is 5.0um. When the sensor detects the flame signal of the 3.6um waveband, the red LED indicator of the warning information connected to the output end of the MCU module starts to flash, and the sensor starts to collect data, and transmits the collected data to the MCU module. The MCU module compares the data with the preset threshold. If the data matches the set threshold, the main control MCU module drives the camera to take pictures. The taken pictures are first pre-processed inside the camera, and then transmitted to the AI ​​module for a series of algorithm processing and judgment. If it is judged that there is a fire in the picture, the picture data will be transmitted to the 4G CAT-1 module. The 4G CAT-1 module transmits the alarm picture data to the platform server through the wireless communication network. While transmitting the picture data, the data transmission blue LED indicator starts to flash, and stops flashing to indicate that the alarm picture data transmission is over, and the entire alarm process is completed. Such composite detection can more accurately identify fires. When the sensor detects a flame signal in the 4.4um band, the red LED indicator connected to the output of the MCU module starts flashing, and the sensor starts to collect data and transmits the collected data to the MCU module. The MCU module compares the data with the preset threshold. If the data matches the set threshold, the main control MCU module drives the camera to take photos. The photos taken are first pre-processed inside the camera, and then sent to the AI ​​module for a series of algorithm processing and judgment. If it is judged that there is a fire in the picture, the picture data will be sent to the 4G CAT-1 module. The 4G CAT-1 module transmits the alarm picture data to the platform server through the wireless communication network. While transmitting the picture data, the blue LED indicator of data transmission starts flashing. Stopping flashing indicates that the alarm picture data transmission is over and the entire alarm process is completed. This composite detection can identify fires more accurately. When the sensor detects a flame signal in the 5.0um band, the red LED indicator of the warning information connected to the output end of the MCU module begins to flash, and the sensor begins to collect data and transmits the collected data to the MCU module. The MCU module compares the data with the preset threshold. If the data matches the set threshold, the main control MCU module drives the camera to take pictures. The taken pictures are first pre-processed inside the camera, and then sent to the AI ​​module for a series of algorithm processing and judgment. If it is judged that there is a fire in the picture, the picture data will be sent to the 4G CAT-1 module. The 4G CAT-1 module transmits the alarm picture data to the platform server through the wireless communication network. While transmitting the picture data, the blue LED indicator of data transmission begins to flash. Stopping flashing indicates that the alarm picture data transmission is completed and the entire alarm process is completed. This composite detection can identify fires more accurately.

[0025] The wavelength range of the ultraviolet flame radiation signal detected by the ultraviolet sensor is 180nm~260nm. When the sensor detects the flame signal in the 180nm~260nm band, the red LED indicator of the warning information connected to the output end of the MCU module starts to flash, and the sensor starts to collect data and transmits the collected data to the MCU module. The MCU module compares the data with the preset threshold value. If the data matches the set threshold value, the main control MCU module drives the camera to take pictures. The taken pictures are first pre-processed and judged inside the camera, and then transmitted to the AI ​​module for a series of algorithm processing. If it is judged that there is a fire in the picture, the picture data will be transmitted to the 4G CAT-1 module. The 4G CAT-1 module transmits the alarm picture data to the platform server through wireless communication. While transmitting the picture data, the data transmission blue LED indicator starts to flash, and stops flashing to indicate that the alarm picture data transmission is completed, and the entire alarm process is completed. Such composite detection can identify fire more accurately.

[0026] The solar panel comprises a first solar panel 6 and a second solar panel 12. The polygonal shell 1 is a regular pentagonal shell. The first solar panel 6 and the second solar panel 12 are respectively mounted on two opposite inclined surfaces at the upper end of the regular pentagonal shell. The first solar panel 6 and the second solar panel 12 are both provided with a plurality of elastic rods 8. The ends of the elastic rods 8 are provided with hooks 9. The regular pentagonal shell is provided with through holes 10 for the elastic rods 8 to pass through. The second solar panel 12 is longer than the first solar panel 6. When the first solar panel 6 and the second solar panel 12 are in a staggered position at the top, the second solar panel 12 extends from the right inclined surface of the regular pentagonal shell and is mounted on the end of the first solar panel 6, so that rainwater can only flow down from the inclined surface of the regular pentagonal shell on rainy days, thereby avoiding being located in the The right inclined surface of the regular pentagonal shell leaks water; the multiple installation cavities include an upper left cavity and an upper right cavity divided by the horizontal plate 4 and the vertical plate 3, and a lower left cavity and a lower right cavity located below the horizontal plate 4. The inclined plate 5 is respectively connected to the intersection of the horizontal plate and the vertical plate 3 and divides the lower left cavity and the lower right cavity into a first cavity, a second cavity, a third cavity, and a fourth cavity. Rainwater enters the upper left cavity along the through hole 10 located at the left inclined surface 11 corresponding to the first solar panel 6 to form accumulated water, which is helpful for cooling the water storage of a flame detector. The upper left cavity, the upper right cavity, the lower left cavity and the lower right cavity respectively form a closed space with the regular pentagonal shell. The MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module, the power supply module, the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the G CAT-module are respectively located.

[0027] The transverse plate 4 is provided with a groove 13, and a first depression 14 and a second depression 15 are provided in the groove 13, and a first ball 16 and a second ball 17 are provided in the first depression 14 and the second depression 15 respectively, and the first ball 16 and the second ball 17 are connected with a first elastic cord 18 and a second elastic cord 19 respectively, and the first elastic cord 18 and the second elastic cord 19 have built-in connecting wires, and the two connecting wires are respectively connected to the first ball 16 and the second ball 17, and the two connecting wires are respectively connected to an early warning module, and the early warning module has a buzzer, and the early warning module is activated when the ball 16 and the second ball 17 collide; the first ball 16 and the second ball 17 collide and return to the first depression 14 and the second depression 15 respectively through the elastic cord 18 and the second elastic cord 19, and the first ball 16 and the second ball 17 are just located in the first depression 14 and the second depression 15 when the first elastic cord 18 and the second elastic cord 19 are in a natural state.

[0028] The depths of the first depression 14 and the second depression 15 are respectively one-quarter to one-third of the diameters of the first ball 16 and the second ball 17; when the polygonal shell 1 is tilted to a certain angle, any one of the first ball 16 or the second ball 17 rolls out from the first depression 14 and the second depression 15 respectively and then collides to trigger the early warning module. The first ball 16 and the second ball 17 are both iron balls, which will contact each other during the collision to activate the early warning module. The first ball 16 and the second ball 17 are respectively provided with arc-shaped magnetic blocks 18 and are connected thereto. The two arc-shaped magnetic blocks 21 are respectively connected to the first ball 16 and the second ball 17. Lubricants are arranged between the first ball 16 and the second ball 17, and the arc-shaped magnetic block 18 and the first ball 16 and the second ball 17 respectively. Since the arc-shaped magnetic block 21 is magnetic, the iron ball is always connected to the arc-shaped magnetic block 21 and slides around the ball under the action of the lubricant, so that it can be connected without affecting the rolling; the groove 13 is located at the intersection of the horizontal plate 4 and the vertical plate 3 and is not connected to the upper left cavity. The vertical plate 3 has a notch 20 located above the groove 13, and the first ball 16 and the second ball 17 are always located between the groove 13 and the notch 20.

[0029] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A flame detector, characterized in that: The invention comprises a polygonal shell (1), a rod (2) is installed at the bottom of the polygonal shell (1), the polygonal shell (1) has an inner cavity (7), a horizontal plate (4), a vertical plate (3) and an inclined plate (5) are cross-arranged in the inner cavity (7), the horizontal plate (4), the vertical plate (3) and the inclined plate (5) divide the inner cavity (7) into a plurality of sealed installation cavities, the polygonal shell (1) is installed with a power supply module, an MCU module, an artificial intelligence AI module, a three-band infrared sensor module, an ultraviolet sensor module, a camera module and a 4G CAT-1 module, the power supply module, the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module are arranged in the plurality of installation cavities, the MCU module is connected to the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module, and the power supply module is connected to the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module.

2. A flame detector according to claim 1, characterized in that: The power supply module comprises a solar panel and a rechargeable lithium battery pack which are connected to each other. The rechargeable lithium battery pack is installed in the installation cavity and is connected to the MCU module, the artificial intelligence AI module, the three-band infrared sensor module, the ultraviolet sensor module, the camera module and the 4G CAT-1 module. The solar panel is installed on the outer wall surface of the polygonal shell (1).

3. A flame detector according to claim 1, characterized in that: The solar panel comprises a first solar panel (6) and a second solar panel (12); the polygonal shell (1) is a regular pentagonal shell; the first solar panel (6) and the second solar panel (12) are respectively mounted on two opposite inclined surfaces at the upper end of the regular pentagonal shell; the first solar panel (6) and the second solar panel (12) are both provided with a plurality of elastic rods (8); the ends of the elastic rods (8) are provided with barbs (9); and the regular pentagonal shell is provided with through holes (10) for the elastic rods (8) to pass through.

4. A flame detector according to claim 1, characterized in that: The transverse plate (4) is provided with a groove (13), a first depression (14) and a second depression (15) are provided in the groove (13), a first ball (16) and a second ball (17) are provided in the first depression (14) and the second depression (15), respectively, the first ball (16) and the second ball (17) are connected to a first elastic cord (18) and a second elastic cord (19) respectively, the first elastic cord (18) and the second elastic cord (19) have built-in connecting wires, the two connecting wires are respectively connected to the first ball (16) and the second ball (17), the two connecting wires are respectively connected to an early warning module, the early warning module has a buzzer, and the early warning module is activated when the first ball (16) and the second ball (17) collide.

5. A flame detector according to claim 4, characterized in that: The depths of the first depression (14) and the second depression (15) are respectively one quarter to one third of the diameters of the first rolling ball (14) and the second rolling ball (17).

6. A flame detector according to claim 4, characterized in that: The first ball (14) and the second ball (17) are both iron balls. The first ball (14) and the second ball (17) are respectively provided with arc-shaped magnetic blocks (18) and connected thereto. The two arc-shaped magnetic blocks (21) are respectively connected to the first ball (14) and the second ball (17). Lubricants are respectively provided between the arc-shaped magnetic blocks (18) and the first ball (14) and the second ball (17).

Images