A Smart Diagnostic System for Security Monitoring Sensors Based on Deep Learning Algorithms

Abstract

This invention relates to the field of monitoring system detection technology, specifically to an intelligent diagnostic system for safety monitoring sensors based on deep learning algorithms. The system includes sensors, a controller, and an alarm. Both the sensor and the alarm are electrically connected to the controller. The sensor includes a mounting base and a housing mounted on the mounting base. Inside the housing is a detection device that detects methane concentration via catalytic combustion. This invention utilizes an inlet pipe with a raised inner wall at the methane detection section of the system. The gas is detected and analyzed by passing through the inlet pipe. The raised inner wall of the inlet pipe reduces the pressure and flow rate of the detected gas, ensuring a consistent gas sampling volume in the methane detection section. This prevents interference from the negative pressure inside the coal mine ventilation ducts, thereby reducing the computational complexity of gas pressure and wind speed correlation in the algorithm, simplifying the diagnostic system's algorithm, and improving its accuracy.

Description

Technical Field

[0001] This invention relates to the field of monitoring system detection technology, specifically to an intelligent diagnostic system for security monitoring sensors based on deep learning algorithms. Background Technology

[0002] Coal mine safety monitoring systems utilize multiple sensors to monitor various hazardous factors within the mine in real time. A controller processes, transmits, analyzes, and stores the sensor data, then feeds the processed data back to the alarm system for timely warnings. This helps workers respond quickly to potential dangers and ensures mine safety. However, sensors can be affected by interference from other factors during operation. Therefore, existing coal mines employ intelligent diagnostic systems for safety monitoring sensors based on deep learning algorithms to ensure the accuracy of sensor monitoring.

[0003] The intelligent diagnostic system for safety monitoring sensors based on deep learning algorithms uses a multi-model ensemble method based on deep neural network learning to combine multiple classification models, such as k-NN, support vector machines, and gradient boosting decision trees (GBDTs), to construct a multi-model ensemble model to predict the health status of the sensors. Cross-validation analysis is added to select important feature information, and the outputs of multiple classification models are integrated using deep neural networks to avoid false alarms from the sensors.

[0004] Some sensors are installed in specific locations within coal mines, making them susceptible to various factors. For example, infrared gas sensors installed in coal mine exhaust pipes are subject to negative pressure in their detection chambers due to the high and fluctuating gas flow rates inside. This affects the inlet pressure and velocity of the detection chamber, altering the gas density and impacting the sensor's accuracy. To address this, pressure and velocity monitoring sensors are installed within the infrared gas sensor, along with algorithms to ensure accuracy. However, this increases the complexity of deep learning algorithms, leading to errors in feature analysis and inaccurate sensor malfunction detection. False alarms not only interfere with workers' assessment of the mine's safety but also create significant psychological stress. Furthermore, an increase in false alarms can lower worker vigilance, delaying the response to potential safety hazards and posing a significant safety risk.

[0005] Therefore, in order to reduce the complexity of the intelligent diagnostic system algorithm and improve the accuracy of the intelligent diagnostic system, a safety monitoring sensor intelligent diagnostic system based on deep learning algorithm is proposed. Summary of the Invention

[0006] The purpose of this invention is to provide a safety monitoring sensor intelligent diagnostic system based on deep learning algorithms. By optimizing the constant intake air in the gas concentration monitoring section, the algorithm of the intelligent diagnostic system is optimized. This not only simplifies the learning algorithm of the intelligent diagnostic system and reduces the calculation steps of the learning algorithm, but also improves the accuracy of the learning algorithm, so that the system can achieve high diagnostic accuracy when applied to coal mine exhaust pipes.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A safety monitoring sensor intelligent diagnostic system based on deep learning algorithms includes a mounting base and a housing. The housing is mounted on the mounting base and contains a detection device that detects gas concentration using infrared light. The housing also contains a detection chamber and an air extraction assembly that draws external air into the detection chamber. The housing has an air inlet connected to the air extraction assembly and the detection chamber. An air inlet pipe connected to the air inlet is also present on the housing. The inner wall of the air inlet pipe has multiple protrusions that reduce air pressure. A cleaning assembly is also present on the housing. This cleaning assembly cleans scale buildup on the inner wall of the air inlet pipe by deforming it, and the cleaning assembly can also use wind power to vibrate the air inlet pipe, reducing the frequency of scale buildup on its inner wall.

[0009] During operation, gas flows into the housing through the intake pipe. The protrusions inside the intake pipe obstruct the airflow, reducing pressure and slowing it down until it gradually reaches equilibrium. This ensures a constant intake volume for the gas sensor, preventing external airflow from affecting the gas monitoring section of the intelligent diagnostic system. This reduces false alarms caused by pressure changes, simplifying the learning algorithm and reducing computational steps, thus improving the diagnostic accuracy of the intelligent diagnostic system. However, coal mine air has high humidity, making the intake pipe prone to scaling and blockage. Since mines typically have many gas sensors, their maintenance requires significant time and labor costs. Therefore, a cleaning component is provided. This component allows for quick cleaning of scale buildup in the intake pipe, reducing sensor maintenance costs. Furthermore, the cleaning component utilizes wind power to further reduce scale buildup, decreasing maintenance frequency and overall system maintenance costs.

[0010] Preferably, the air intake pipe is a retractable corrugated pipe, and the air intake pipe is arranged in a vertical spiral. The cleaning component is connected to the corrugated pipe to keep the corrugated pipe in a stable spiral shape. A dust collection box with an air inlet 2 is connected to the bottom of the air intake pipe.

[0011] By using a retractable bellows in the intake pipe, the creases of the bellows can be used as protrusions to reduce airflow pressure and velocity. The spiral arrangement of the bellows reduces the number of installation positions, allowing for a more compact placement of the gas sensor. Furthermore, the spiral arrangement of the intake pipe creates a spiral flow of air inside, subjecting dust particles in the gas to centrifugal force. This centrifugal force causes the dust particles to rub against the inner wall of the intake pipe, reducing their kinetic energy and causing them to fall off. This prevents dust from entering the housing and adhering to the detection device, thus avoiding a decrease in detection accuracy. The use of a bellows... The design allows dust to impact the creases of the bellows, reducing its kinetic energy and improving cleaning effectiveness. The bellows are retractable; during intake cleaning, repeated stretching causes elastic deformation, allowing dirt to easily detach under reciprocating stress. This eliminates the need for water rinsing, facilitating intake pipe cleaning and maintenance. Furthermore, the connection between the cleaning component and the bellows prevents the bellows from changing its spiral state under airflow impact, thus preventing turbulence and dust re-entrainment and further enhancing cleaning efficiency.

[0012] Preferably, the spiral tilt angle of the intake pipe is 21°-36°.

[0013] When the spiral tilt angle is selected at this angle, particles can easily fall off the air intake pipe, and the volume occupied by the air intake pipe can be reduced, making the related equipment smaller and easier to install.

[0014] Preferably, the cleaning component includes multiple connecting sleeves disposed on the housing. Each connecting sleeve is provided with a locking block and a sliding groove. The locking block engages with the sliding groove of the adjacent connecting sleeve. The sliding groove is arc-shaped. The connecting sleeves located at the top and bottom are connected to the housing. The air intake pipe is connected to the multiple connecting sleeves respectively. The connecting sleeve located at the bottom is connected to a wind receiving part. The wind receiving part can transmit power to the connecting sleeve through wind force to cause the connecting sleeve to vibrate.

[0015] The cleaning assembly employs a multi-sleeve design, ensuring the air intake pipe is always connected to these sleeves. Pulling the bottom sleeve stretches the air intake pipe, causing elastic deformation and dislodging dirt adhering to it. The sleeves are connected by sliding grooves and locking blocks, with the grooves being arc-shaped. This allows the sleeves to rotate during sliding, pulling the air intake pipe from a spiral shape to a vertical one, allowing dirt to fall more effectively to the outside. After cleaning, simply pushing the sleeve upwards compresses the air intake pipe through its vertical and rotational movements, restoring it to its contracted and original spiral state. This improves both the cleaning efficiency and the speed of subsequent installation, reducing maintenance time and costs for the intelligent diagnostic system.

[0016] Preferably, the air receiving part includes a mounting plate elastically connected to the connecting sleeve located below, the connecting sleeve being connected to the housing via the mounting plate, and two adjacent connecting sleeves being elastically connected. A push rod is connected to the connecting sleeve located below, and a push plate is connected to the push rod. The push plate is inclined upward along the direction of the external airflow.

[0017] By elastically connecting multiple connecting sleeves, and then connecting push rods and push plates to the connecting sleeves, the airflow in the ventilation duct pushes the push plates up and down, causing the connecting sleeves to slide. This stretches the intake pipe, making it easier for dust to fall from the inner wall of the intake pipe into the dust collection box. This prevents dust from accumulating on the inner wall of the intake pipe for extended periods, thus reducing the frequency of scaling on the inner wall of the intake pipe, the maintenance frequency of the intelligent diagnostic system, and the maintenance cost of the system. The elastic connection between the connecting sleeves ensures that the relative displacement between them is the same, thus ensuring that the deformation of each part of the intake pipe is the same. This keeps the intake pipe rotating in a spiral motion, preventing turbulence inside the intake pipe from affecting the constant airflow collection.

[0018] Preferably, the dust collection box is provided with a buffer tube at the air inlet, the buffer tube extends into the air inlet pipe, and the connection between the buffer tube and the air inlet pipe is tangent.

[0019] By installing a buffer tube in the middle of the dust collection box, the airflow can directly enter the air intake pipe during the process of entering the dust collection box, thus ensuring the continuity of air intake. Moreover, the connection between the buffer tube and the air intake pipe is tangent, which makes the airflow entering the air intake pipe more stable. This avoids the problem of gas turbulence inside the dust collection box, which would cause dust to be stirred up again and affect the dust cleaning effect. This improves the dust separation effect and the detection accuracy of the system.

[0020] Preferably, the push plate has a plurality of protrusions spirally arranged, and the spiral arrangement direction of the protrusions is opposite to the spiral direction of the intake pipe.

[0021] The protrusions on the connecting sleeve can prevent the bellows from deforming while being pulled or pushed. When the bellows deforms too much, it will pass over the protrusions and cause the bellows to vibrate. This makes it easier for dirt adhering to the inner wall of the bellows to be removed, thereby improving the cleaning effect of the intake pipe, reducing the maintenance time of the intake pipe, and improving the maintenance efficiency of the intelligent diagnostic system. The protrusions also increase the contact area between the connecting sleeve and the intake pipe, allowing the vibration of the connecting sleeve to be better transmitted to the intake pipe, thus improving the vibration effect of the intake pipe and further improving the cleaning effect of the intake pipe.

[0022] Preferably, the push plate is provided with a plurality of irregularly shaped guide blocks, and the plurality of guide blocks are arbitrarily distributed on the push plate, and the plurality of guide blocks make the center of gravity of the push plate move away from the push rod.

[0023] By setting guide blocks on the push plate, turbulence can be generated at the push plate, causing the push plate to be subjected to different thrusts. This results in inconsistent thrusts from the push plate to the push rod, causing irregular vibrations in the connecting sleeve. This improves the cleaning effect of the intake pipe. Furthermore, the guide blocks can shift the center of gravity of the push plate, causing greater shaking at the push rod, thereby further improving the vibration effect of the intake pipe and enhancing the cleaning effect of dust inside the intake pipe.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0025] 1. By installing an air inlet pipe with a raised inner wall in the gas detection section of the system, the gas is detected and analyzed by passing through the air inlet pipe. The raised inner wall of the air inlet pipe reduces the pressure and flow rate of the gas, ensuring that the gas sampling volume in the gas detection section is always equal. This ensures that the gas sampling volume is not affected by the negative pressure inside the coal mine ventilation duct, thereby reducing the calculation of the correlation between gas pressure and wind speed in the algorithm, simplifying the algorithm of the diagnostic system, and improving the accuracy of the diagnostic system.

[0026] 2. By designing the intake pipe as a spiral shape, the airflow can become spiral while the gas is depressurized and slowed down in the intake pipe. This allows the particulate matter in the gas to be subjected to centrifugal force, and the particulate matter will rub against the intake pipe wall and collide with the protrusions under the action of centrifugal force. This reduces the kinetic energy of the particulate matter and prevents it from entering the sensor's detection area. This avoids the particulate matter affecting the sensor's detection accuracy and improves the accuracy of the diagnostic system.

[0027] 3. By setting up a cleaning assembly with multiple connecting sleeves, the intake pipe is fixed to multiple connecting sleeves, thereby shaping the spiral shape of the intake pipe. This prevents the spiral shape of the intake pipe from deforming under the influence of airflow, thus avoiding the problem of turbulence inside the intake pipe caused by the deformation of the intake pipe, which leads to changes in air pressure and airflow speed and affects the diagnostic accuracy of the system. The multiple connecting sleeves can extend and rotate during the extension and rotation process, thereby causing the intake pipe to elongate and deform, and make the intake pipe vertical. This allows the dirt on the inner wall of the intake pipe to be quickly removed from the intake pipe and discharged, thereby effectively improving the maintenance efficiency of the system and reducing the maintenance cost of the system.

[0028] 4. An air intake is installed on the connecting sleeve at the bottom of the cleaning component, allowing the system to use air force to self-clean the intake pipe. This reduces the frequency of scaling on the intake pipe, lowers the system's maintenance frequency and costs, and ensures that the deformation of the intake pipe is consistent across all parts when the connecting sleeve causes deformation. This prevents changes in the spiral shape of the intake pipe from causing turbulence inside the pipe and dust to be stirred up. As a result, the cleaning component can perform self-cleaning without stirring up dust, thus affecting the system's dust separation effect. Attached Figure Description

[0029] Figure 1 This is a partial structural cross-sectional view of the present invention;

[0030] Figure 2 For the present invention Figure 1 Cross-sectional view at point AA;

[0031] Figure 3 This is a schematic diagram of the structure of multiple connecting sleeves stretching against each other according to the present invention;

[0032] Figure 4 For the present invention Figure 3 Enlarged view of section B in the middle.

[0033] In the diagram: 1. Mounting base; 2. Housing; 21. Air inlet one; 3. Air inlet pipe; 31. Protrusion; 4. Cleaning assembly; 41. Connecting sleeve; 42. Locking block; 43. Slide groove; 44. Air receiving part; 441. Mounting plate; 442. Push rod; 443. Push plate; 444. Spring; 45. Protrusion; 5. Dust collection box; 51. Air inlet two; 52. Buffer pipe; 6. Guide block. Detailed Implementation

[0034] like Figures 1 to 4 As shown, the details are as follows:

[0035] A safety monitoring sensor intelligent diagnostic system based on deep learning algorithms includes sensors, a controller, and an alarm. Both the sensor and the alarm are electrically connected to the controller. During operation, gas from the coal mine exhaust pipe enters the sensor. The sensor analyzes the data and transmits it to the controller. After analysis and processing by the controller, the sensor data controls the alarm, which consists of a buzzer and an alarm light. Coal mine workers should evacuate quickly upon seeing the alarm light illuminate or hearing the alarm sound. The sensor includes a mounting base 1 and a housing 2. The housing 2 contains a detection chamber, which houses a detection device equipped with a sensor. An air pump is installed inside the detection chamber. The housing 2 has an opening... An air inlet 21 connected to the testing chamber is provided, and an air inlet pipe 3 is connected to the air inlet 21. The air inlet pipe 3 is a retractable corrugated pipe made of polytetrafluoroethylene (PTFE). PTFE has good corrosion resistance, thus preventing dust from adhering to the inside of the air inlet pipe 3 and causing corrosion. The air inlet pipe 3 also has good elasticity, giving it a high energy absorption capacity. The corrugations can form protrusions 31 on the inner wall, and the corrugations can be folded multiple times, making it less susceptible to stress fatigue damage under the impact of airflow, thus extending its service life. After the air pump is started, gas enters the testing chamber along the air inlet pipe 3. After the airflow is dissipated by the protrusions 31, the final gas velocity entering the testing chamber is... Equal gas pressure ensures consistent sampling by the air pump, preventing negative pressure effects. This reduces false alarms from gas sensors caused by pressure changes, simplifying the learning algorithm and reducing computational complexity, thus improving the accuracy of the intelligent diagnostic system. A cleaning assembly 4, consisting of multiple connecting sleeves 41, is connected to the intake pipe 3. Each sleeve 41 has a sliding groove 43 and a locking block 42. The locking blocks 42 of adjacent sleeves 41 engage with the sliding grooves 43. During use, the two sleeves 41 can extend and retract relative to each other, deforming the corrugated intake pipe 3 and causing dirt accumulated on its inner wall over time to fall off. The intake pipe 3 is spirally coiled around multiple connecting sleeves 41, which allows the airflow inside the intake pipe 3 to flow in a spiral. This causes the dust carried in the gas to be subjected to centrifugal force, which causes the dust to rub against the inner wall of the intake pipe 3 and collide with the protrusions 31. This reduces the kinetic energy of the dust and causes it to fall off. During the constant pressure treatment of the airflow, the intake pipe 3 can also prevent dust from entering the housing 2 and adhering to the detection device, thus reducing the detection accuracy of the detection device. The multiple connecting sleeves 41 can keep the intake pipe 3 in a stable spiral shape, preventing the change of the spiral shape of the intake pipe 3 from causing turbulence inside the intake pipe 3 and causing the dust to be stirred up again, thereby further improving the dust removal effect of the intake pipe 3.

[0036] Dust settles by reducing its kinetic energy; therefore, it's necessary to ensure the dust isn't stirred up by the airflow from the pump. Typically, the airflow rate of the pump is 1-5 L / min. According to Stokes' Law, the resistance F experienced by particulate matter moving in a fluid medium... d The relationship between its velocity v and particle radius r is:

[0037] F d =3πμrv

[0038] Where μ represents the viscosity of the fluid, which is approximately 1.81 × 10⁻⁵ Pa·s at atmospheric pressure.

[0039] From the equilibrium equation of gravity and terminal velocity, the terminal velocity v of the particle can be obtained. t for:

[0040] v t =(2rgρ) / (9μ)

[0041] Where g is the acceleration due to gravity, and ρ is the density of the particulate matter.

[0042] Therefore, the ability of particulate matter to be carried by the airflow can be determined based on the terminal velocity. When the terminal velocity is less than the airflow velocity, the particulate matter can be carried by the airflow; otherwise, it cannot. Thus, when the airflow rate is 1-5 L / min, it cannot carry particulate matter with a diameter greater than 4.4 to 8.3 μm.

[0043] Specific calculations:

[0044] When the airflow rate is 1L / min, for a particle with a diameter of 4.4μm, its terminal velocity is about 0.0178mm / s, which is less than the air velocity corresponding to 1L / min (about 0.0183m / s), so it cannot be carried by the airflow.

[0045] Similarly, when the airflow rate is 5 L / min, for a particle with a diameter of 8.3 μm, its terminal velocity is about 0.262 mm / s, which is less than the air velocity corresponding to 5 m / s (about 0.306 m / s), so it cannot be carried by the airflow.

[0046] Infrared gas sensors can be affected by dust or particles larger than 10 μm in diameter, while dust or particles larger than 20 μm in diameter may significantly affect the sensor's performance. This method, by eliminating the kinetic energy of the dust, can remove particles larger than 4.4 μm, thereby avoiding the impact of large-diameter dust on the monitoring accuracy of the gas sensor.

[0047] To ensure that particles falling onto the inner wall of intake pipe 3 fall off more easily, according to Newton's second law, the net force on dust in equilibrium is zero, therefore:

[0048] sinθ=μcosθ

[0049] Where sinθ represents the inclination angle of the inclined plane, and μ represents the coefficient of friction between the inclined plane and the dust.

[0050] Dividing both sides of the above equation by cosθ, we get:

[0051] tanθ=μ

[0052] Therefore, given that the particle diameter is 4.4 μm and the coefficient of friction μ on the inclined plane is known, the inclination angle θ at which the particle can remain on the inclined plane is arctan-μ.

[0053] The coefficient of friction depends on the material of the inclined surface and the particles. Telescopic bellows are typically made of polymer materials such as rubber and plastic, which generally have a low coefficient of friction. For typical polymer materials, the coefficient of friction is approximately between 0.1 and 0.4. Therefore, when using a smooth plastic inclined surface to hold 4.4μm dust, the inclination angle is approximately between 14 and 24 degrees. However, in practice, the coefficient of friction is affected by factors such as humidity. Therefore, a safety factor of 1.5 is used, and the spiral inclination angle of the intake pipe 3 is calculated to be 21°-36°. Using this angle, particles easily fall off the intake pipe 3, and the volume occupied by the intake pipe 3 is reduced, making the related equipment smaller and easier to install. Multiple connecting sleeves 41 are interconnected by springs 444, and the lowest connecting sleeve 41 is equipped with a push rod 442. The push rod 442 is connected to a spring that tilts towards the airflow direction inside the ventilation duct. During operation, the airflow can blow the push plate 443 and push the connecting sleeve 41 upward through the push rod 442. This allows the intake pipe 3 to experience a certain amount of expansion and contraction vibration during the testing process, making it easier for dust on the intake pipe 3 to fall into the dust collection box 5 connected to the intake pipe 3. This reduces the frequency of dust accumulation on the intake pipe 3 and the frequency of maintenance. The spring 444 ensures that the relative displacement of adjacent connecting sleeves 41 is equal when multiple connecting sleeves 41 move. This causes the intake pipe 3 to elongate at multiple positions with the same amount of extension and ensures that the rotation angle of each part of the intake pipe 3 is equal, preventing different deformations at different positions of the intake pipe 3 and thus avoiding damage to the intake pipe 3. Irregular deformation causes turbulence inside the intake pipe 3, altering the gas pressure and flow rate, thus affecting the diagnostic accuracy of the system. Multiple protrusions 45 are provided on the connecting sleeve 41. These protrusions 45 fit snugly against the intake pipe 3, and since the intake pipe 3 is corrugated, the protrusions 45 fit snugly against the outer folds of the intake pipe 3, thereby increasing the contact area between the intake pipe 3 and the connecting sleeve 41. This allows the vibration generated by the connecting sleeve 41 to be better transmitted to the intake pipe 3, improving the vibration effect of the intake pipe 3 and thus enhancing its cleaning effect. The push plate 443 is equipped with irregular guide blocks 6, which can be square, irregularly shaped, triangular, etc., allowing the push plate 443 to... The surface is prone to turbulence, causing inconsistent thrust on the pusher plate 443. This results in different thrusts, causing irregular vibrations in the connecting sleeve 41 and irregular deformations in the intake pipe 3, thereby improving the cleaning effect of the intake pipe 3. The guide block 6 can shift the center of gravity of the pusher plate 443, causing greater swaying at the push rod 442, further enhancing the vibration effect of the intake pipe 3 and improving the cleaning effect of dust inside the intake pipe 3. The side of the guide block 6 with a larger center of gravity can be set towards one edge of the pusher plate 443, so that the center of gravity of the pusher plate 443 is concentrated at that edge. The guide block 6 can also be made of different materials, such as a high-density metal at the edge of the pusher plate 443.The guide block 6, positioned opposite the edge, is made of low-density engineering plastic. When dust falls from the inlet pipe 3 into the dust collection box 5, a buffer pipe 52 is installed on the dust collection box 5. This buffer pipe 52 is tangential to the inlet pipe 3. When gas enters the inlet pipe 3 from the inlet port 51, the airflow into the dust collection box 5 is stable, preventing turbulence within the dust collection box 5. This avoids the collected dust or dust falling from the inlet pipe 3 into the dust collection box 5 being re-raised, thus improving the dust removal effect of the inlet pipe 3 and enhancing the diagnostic accuracy of the system.

[0054] When cleaning the dust and dirt from the intake pipe 3, the mounting plate 441 is fixed to the housing 2 by a slot connection. In use, simply turn counterclockwise and remove the dust collection box 5 connected to the intake pipe 3. Then, manually pull the push rod 442. At this time, the connecting sleeve 41 extends and rotates, causing the intake pipe 3 to extend and rotate to a vertical position. After the intake pipe 3 is deformed, the dust and dirt can no longer adhere and fall off, and finally be discharged directly along the intake pipe 3. During the stretching process, the protrusion 45 on the connecting sleeve 41 will lock the protrusion 31 on the outer wall of the intake pipe 3, thereby causing the intake pipe 3 to vibrate, thereby improving the cleaning effect of the cleaning component 4. To avoid the protrusion 45 squeezing and damaging the intake pipe 3, the protrusion 45 is set as a hemispherical shape. When reinstalling the intake pipe 3, simply push the multiple connecting sleeves 41 to retract, and the intake pipe 3 will return to its original spiral shape. Finally, tighten the connecting bolts of the push rod 442 and the mounting plate 441. This method of maintenance and post-maintenance installation is quick and can effectively reduce the system maintenance time.

[0055] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A safety monitoring sensor intelligent diagnostic system based on deep learning algorithm, comprising a sensor, a controller, and an alarm, wherein the sensor and the alarm are electrically connected to the controller, the sensor comprising a mounting base (1) and a housing (2) disposed on the mounting base (1), the housing (2) having a detection device for detecting gas concentration by infrared light inside, the housing (2) having an air extraction component for drawing outside air into the housing (2), and the housing (2) having an air inlet (21) communicating with the air extraction component, characterized in that, The housing (2) is provided with an air inlet pipe (3) connected to the air inlet (21). The inner wall of the air inlet pipe (3) is provided with a plurality of protrusions (31) to reduce air pressure. The housing (2) is provided with a cleaning component (4). The cleaning component (4) cleans the scale on the inner wall of the air inlet pipe (3) by deforming the air inlet pipe (3). The cleaning component (4) can also rely on wind power to make the air inlet pipe (3) vibrate to reduce the frequency of scale buildup on the inner wall of the air inlet pipe (3). The air inlet pipe (3) is made of a retractable corrugated pipe and is arranged in a vertical spiral. The cleaning component (4) is connected to the air inlet pipe (3) to keep the air inlet pipe (3) in a stable spiral shape. A dust collection box (5) with an air inlet port (51) is connected below the air inlet pipe (3). The cleaning component (4) includes multiple connecting sleeves (41) disposed on the housing (2). Each connecting sleeve (41) is provided with a locking block (42) and a sliding groove (43). The locking block (42) cooperates with the sliding groove (43) on the adjacent connecting sleeve (41). The sliding groove (43) is arc-shaped. The connecting sleeves (41) located at the top and bottom are connected to the housing (2). The air inlet pipe (3) is connected to multiple connecting sleeves (41) at the same time. The connecting sleeve (41) located at the bottom is connected to a wind receiving part (44). The wind receiving part (44) can transmit power to the connecting sleeve (41) through wind force to make the connecting sleeve (41) vibrate. The air receiving part (44) includes a mounting plate (441) elastically connected to the lower connecting sleeve (41). The lower connecting sleeve (41) is connected to the housing (2) through the mounting plate (441), and two adjacent connecting sleeves (41) are connected to each other by springs (444). A push rod (442) is connected to the lower connecting sleeve (41), and a push plate (443) is connected to the push rod (442). The push plate (443) is inclined upward along the direction of the external airflow.

2. The intelligent diagnostic system for security monitoring sensors based on deep learning algorithms according to claim 1, characterized in that: The push plate (443) has a plurality of protrusions (45) spirally arranged on it, and the spiral arrangement direction of the protrusions (45) is opposite to that of the spiral direction of the air intake pipe (3).

3. The intelligent diagnostic system for security monitoring sensors based on deep learning algorithms according to claim 1, characterized in that: The spiral tilt angle of the intake pipe (3) is 21°-36°.

4. The intelligent diagnostic system for security monitoring sensors based on deep learning algorithms according to claim 1, characterized in that: The dust collection box (5) is provided with a buffer tube (52) at the second air inlet (51). The buffer tube (52) extends into the air inlet pipe (3) and the connection between the buffer tube (52) and the air inlet pipe (3) is tangent.

5. The intelligent diagnostic system for security monitoring sensors based on deep learning algorithms according to claim 1, characterized in that: The push plate (443) is provided with a plurality of irregularly shaped guide blocks (6), and the plurality of guide blocks (6) make the center of gravity of the push plate (443) close to the edge of the push plate (443).

Images