Explosion-proof intelligent camera with automatic cleaning system in underground coal mine

Abstract

The invention provides an underground coal mine explosion-proof intelligent camera with an automatic cleaning system, and relates to the technical field of cameras, the underground coal mine explosion-proof intelligent camera comprises a camera body, an automatic cleaning module and an AI processing module, the outer side of the camera body is an explosion-proof shell, and the explosion-proof shell meets the underground coal mine explosion-proof standard; the automatic cleaning module is electrically connected with the AI processing module; through the integrated design of the explosion-proof shell, the automatic cleaning function and the AI processing function, the core pain points that a traditional explosion-proof camera window is prone to dust accumulation and difficult to maintain are fundamentally solved, the automatic cleaning module adopts the combination mode of high-pressure gas purging, high-pressure water flow flushing and silica gel scraping and brushing, the window pollution degree judgment algorithm is combined, and the automatic cleaning function is achieved. And the cleaning action can be automatically started according to the actual pollution condition, frequent manual disassembly is not needed, the safety risk of underground maintenance operation is reduced, and the equipment loss and the labor cost are reduced.

Description

Technical Field

[0001] This invention relates to the field of camera technology, and in particular to an explosion-proof intelligent camera with an automatic cleaning system for underground coal mines. Background Technology

[0002] In existing technologies, underground coal mines, being high-risk working environments, require explosion-proof monitoring equipment for production process supervision and safety risk control. Commonly used explosion-proof cameras are mainly divided into two categories: flameproof and increased safety types. Their core functionality involves strengthening the sealing of the outer shell and designing an explosion-proof structure to isolate internal circuit sparks from external flammable and explosive environments such as gas and coal dust. Combined with manual inspections or simple video capture methods, they achieve basic monitoring of areas such as mining, transportation, and electromechanical chambers. While some traditional explosion-proof cameras are equipped with simple dustproof coatings or manual cleaning ports, they lack active cleaning mechanisms and still rely on regular manual disassembly and cleaning underground.

[0003] On the one hand, the harsh working conditions of high dust, humidity, and vibration in underground coal mines cause coal dust, water mist, and oil stains to easily accumulate in the camera's viewing window. Traditional equipment lacks efficient automatic cleaning functions, requiring manual cleaning 1-2 times per week. This not only increases the safety risks for workers going down into the mine but also causes problems such as equipment disassembly damage and monitoring interruption during cleaning. On the other hand, viewing window contamination leads to blurred monitoring images, significantly reducing the accuracy of AI recognition algorithms that rely on clear image data, such as those for identifying personnel violations and equipment anomalies. Furthermore, traditional monitoring suffers from numerous blind spots, slow response, and fragmented data, failing to meet the policy requirements and actual production needs of risk prediction and efficient supervision in the intelligent transformation of the coal industry. Therefore, this invention proposes an explosion-proof intelligent camera with an automatic cleaning system for underground coal mines to solve the problems existing in the prior art. Summary of the Invention

[0004] To address the aforementioned issues, this invention proposes an explosion-proof intelligent camera with an automatic cleaning system for underground coal mines. This camera, through its integrated design combining an explosion-proof housing, automatic cleaning function, and AI processing capabilities, fundamentally solves the core pain points of traditional explosion-proof cameras, such as easy dust accumulation on the viewing window and difficult maintenance. The automatic cleaning module employs a combination of high-pressure gas blowing, high-pressure water rinsing, and silicone scraping, combined with a viewing window contamination degree judgment algorithm, to automatically initiate cleaning actions based on the actual contamination level. This eliminates the need for frequent manual disassembly, reducing safety risks associated with underground maintenance operations, lowering equipment wear and labor costs, and ensuring uninterrupted clear monitoring 24 hours a day.

[0005] To achieve the objectives of this invention, the invention is implemented through the following technical solution: an explosion-proof intelligent camera with an automatic cleaning system for underground coal mines, comprising a camera body, an automatic cleaning module, and an AI processing module, wherein the outer side of the camera body is an explosion-proof shell, and the explosion-proof shell meets the explosion-proof standards for underground coal mines;

[0006] The automatic cleaning module is electrically connected to the AI ​​processing module. The automatic cleaning module starts the cleaning action according to preset conditions or environmental sensing signals. The camera body continuously collects image data and transmits it to the AI ​​processing module. The AI ​​processing module performs intelligent analysis and risk warning on the image data.

[0007] Further improvements include: the explosion-proof enclosure adopts an explosion-proof or increased safety type structure with an explosion-proof rating of not less than ExdIMb; the sealing components of the explosion-proof enclosure are made of oil-resistant rubber; and the surface of the explosion-proof enclosure is provided with an antistatic coating.

[0008] A further improvement is that the automatic cleaning module determines whether to initiate a cleaning action based on the result of the window contamination level calculation, which uses the following mathematical algorithm:

[0009]

[0010] Where: P is the window contamination degree (unit: %), which characterizes the degree of influence of contaminants on the window surface on imaging; The pixel area (unit: pixel²) of the clear region in the current image is extracted in real time by the camera itself; The total effective imaging pixel area of ​​the camera body (unit: pixel²) is a fixed value for device calibration. For clear region weighting coefficients, the value range is: 0 < ≤1, based on the preset dust concentration underground; the higher the dust concentration, the better. The larger the value; The average contrast of the current image is calculated from the grayscale values. The standard average contrast ratio calibrated for the camera body serves as the baseline contrast value under clean conditions; K2 is the contrast weighting coefficient, with a value range of 0 < ≤1, and K1 satisfies + =1, used to balance the impact of sharp areas and contrast on contamination levels; when P≥ At that time, the AI ​​processing module sends a start signal to the automatic cleaning module. The preset cleaning start threshold ranges from 30% to 60%.

[0011] Further improvements are made in that: the automatic cleaning module includes a high-pressure gas generating unit, a high-pressure water flow unit, a silicone scraper assembly and a drive motor. The nozzles of the high-pressure gas generating unit and the nozzles of the high-pressure water flow unit are concealed on the explosion-proof housing around the camera body window. The silicone scraper assembly is driven by an explosion-proof drive motor and rotates to wipe the surface of the window.

[0012] Further improvements include: the high-pressure water flow unit has the function of heating the water flow to 30-40℃, the heating start-up condition is that the downhole ambient temperature is ≤5℃, the purging pressure of the high-pressure gas is 0.3-0.6MPa, and the flushing pressure of the high-pressure water flow is 0.5-0.8MPa.

[0013] Further improvements are made in the automatic cleaning module, where preset conditions include timed start-up and manual / remote start-up, and environmental sensing signals include window contamination level signal, underground dust concentration signal, and humidity signal.

[0014] Further improvements include: the camera body is equipped with a low-light high-definition sensor, supporting imaging in low-light environments; the camera body has IP68 dust and water resistance, ≥10G vibration resistance, and high and low temperature resistance from -20℃ to 60℃.

[0015] A further improvement is made in that the AI ​​processing module uses a confidence correction algorithm for the identification results of risk scenarios, and the correction formula is as follows:

[0016]

[0017] Where: R is the corrected AI recognition confidence level, with a value range of 0≤R≤1, used to determine the reliability of the recognition result; The initial identification confidence level has a value range of 0 ≤ ≤1 indicates the initial recognition result of the AI ​​algorithm on the original image; This is a correction factor for the impact of pollution levels, with a value range of 0 < ≤0.5 is used to quantify the degree to which window contamination reduces recognition accuracy; P is the current window contamination level (unit: %). The confidence level is stabilized at a contamination threshold (unit: %), with a range of 10%-20%. This means that when the contamination level is below this value, the impact on the identification confidence level is negligible. This is the image sharpness compensation coefficient, with a value range of 0 < ≤0.3 is used to compensate for the impact of image fluctuations on confidence during the cleaning process; D is the current image sharpness score, which is obtained through an edge detection algorithm. The higher the score, the sharper the image. The standard sharpness score ranges from 80 to 100, serving as the baseline sharpness value under clean conditions; when R ≥ At that time, the AI ​​processing module triggers a real-time alert. The threshold for triggering an early warning is 0.7-0.9.

[0018] Further improvements are made in that the intelligent recognition algorithm integrated into the AI ​​processing module is used to identify three core scenarios: personnel not wearing safety helmets, unauthorized entry, and abnormal equipment operation.

[0019] Further improvements include: the AI ​​processing module communicates wirelessly with the downhole monitoring platform, uploading the collected image data, recognition results, and equipment status information in real time, providing remote viewing and parameter configuration functions.

[0020] The beneficial effects of this invention are as follows:

[0021] 1. This invention fundamentally solves the core pain points of traditional explosion-proof cameras, such as easy dust accumulation and difficult maintenance, by integrating an explosion-proof shell, automatic cleaning function, and AI processing function into an integrated design. The automatic cleaning module adopts a combination of high-pressure gas blowing, high-pressure water flushing, and silicone scraping. Combined with the window contamination degree judgment algorithm, it can automatically start the cleaning action according to the actual contamination situation, eliminating the need for frequent manual disassembly. This not only reduces the safety risks of underground maintenance operations, but also reduces equipment wear and labor costs, and ensures 24-hour uninterrupted clear monitoring.

[0022] 2. The explosion-proof structure and environmental adaptability design of this invention are fully adapted to the complex working conditions in coal mines. The explosion-proof shell meets the coal mine-specific explosion-proof standards such as ExdIMb, and also has IP68 dustproof and waterproof rating, ≥10G vibration resistance, and high and low temperature resistance characteristics of -20℃ to 60℃. It can operate stably in high-risk areas such as mining faces and gas accumulation points, thus expanding the application scenarios of explosion-proof cameras and filling the technical gap of intelligent monitoring equipment in harsh environments.

[0023] 3. This invention achieves closed-loop management of monitoring data acquisition and intelligent analysis through the collaborative work of the AI ​​processing module and the automatic cleaning module. The low-light high-definition sensor ensures imaging quality in low-light environments. The AI ​​recognition algorithm combined with the confidence correction algorithm effectively improves the recognition accuracy of risk scenarios such as personnel not wearing safety helmets and abnormal equipment operation. The real-time early warning function provides reliable data support for underground safety supervision and production scheduling, helping the coal industry to achieve intelligent transformation and improve safety management. Attached Figure Description

[0024] Figure 1 This is the front view of the present invention. Detailed Implementation

[0025] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0026] Example 1

[0027] according to Figure 1 As shown, this embodiment proposes an explosion-proof intelligent camera with an automatic cleaning system for underground coal mines, including a camera body, an automatic cleaning module, and an AI processing module. The outer side of the camera body is an explosion-proof shell, and the explosion-proof shell meets the explosion-proof standards for underground coal mines.

[0028] The automatic cleaning module is electrically connected to the AI ​​processing module. The automatic cleaning module initiates cleaning actions based on preset conditions or environmental sensing signals. The camera continuously collects image data and transmits it to the AI ​​processing module, which performs intelligent analysis and risk warnings on the image data. This achieves synergistic linkage between explosion-proof safety, intelligent cleaning, and risk warning, eliminating the compatibility limitations of single-function devices; it ensures uninterrupted monitoring during cleaning, providing continuous data support for 24-hour safety management underground.

[0029] The explosion-proof enclosure adopts an explosion-proof or increased safety type structure with an explosion-proof rating of not less than ExdIMb. The sealing components of the explosion-proof enclosure are made of oil-resistant rubber, and the surface of the explosion-proof enclosure is coated with an antistatic coating. The oil-resistant rubber sealing components improve the sealing durability of the enclosure and prevent the explosion-proof performance from deteriorating due to moisture and oil contamination; the antistatic coating effectively prevents dust accumulation on the enclosure from generating static sparks, reducing the risk of gas and coal dust explosions.

[0030] The automatic cleaning module is triggered based on the window contamination level calculation result, which uses the following mathematical algorithm:

[0031]

[0032] Where: P is the window contamination degree (unit: %), which characterizes the degree of influence of contaminants on the window surface on imaging; The pixel area (unit: pixel²) of the clear region in the current image is extracted in real time by the camera itself; The total effective imaging pixel area of ​​the camera body (unit: pixel²) is a fixed value for device calibration. For clear region weighting coefficients, the value range is: 0 < ≤1, based on the preset dust concentration underground; the higher the dust concentration, the better. The larger the value; The average contrast of the current image is calculated from the grayscale values. The standard average contrast value calibrated for the camera body is the baseline contrast value under clean conditions; This is the contrast weighting coefficient, with a value range of 0 < ≤1, and satisfy + =1, used to balance the impact of sharp areas and contrast on contamination levels; when P≥ At that time, the AI ​​processing module sends a start signal to the automatic cleaning module. A preset cleaning initiation threshold is set, ranging from 30% to 60%. By quantifying the degree of contamination, the cleaning action can be precisely started and stopped, avoiding resource waste caused by blind cleaning. The weighting coefficient is dynamically adjusted in combination with the underground dust concentration to adapt to the contamination judgment needs under different working conditions and improve the targeting of cleaning.

[0033] The automatic cleaning module includes a high-pressure gas generating unit, a high-pressure water flow unit, a silicone scraper assembly, and a drive motor. The nozzles of the high-pressure gas generating unit and the nozzles of the high-pressure water flow unit are concealed within the explosion-proof housing surrounding the camera's viewing window. The silicone scraper assembly, driven by the explosion-proof motor, rotates and wipes the viewing window surface. The concealed nozzle and nozzle design prevents the cleaning components from obstructing the imaging field of view, ensuring the integrity of the monitoring image. The silicone scraper, rotating and wiping within the viewing window, combined with the gas and water flow, achieves multi-dimensional cleaning, thoroughly removing stubborn contaminants.

[0034] The high-pressure water flow unit has the function of heating the water flow to 30-40℃. The heating start-up condition is that the underground ambient temperature is ≤5℃. The purging pressure of the high-pressure gas is 0.3-0.6MPa, and the flushing pressure of the high-pressure water flow is 0.5-0.8MPa. The 30-40℃ heated water flow can quickly soften frozen oil and coal dust under low-temperature conditions, improving the cleaning effect. Precise control of gas and water flow pressure ensures both cleaning power and avoids damage to the viewing window glass due to excessive pressure.

[0035] The automatic cleaning module includes preset conditions for timed start-up and manual / remote start-up. Environmental sensing signals include window contamination level signal, downhole dust concentration signal, and humidity signal. The combination of timed start-up and manual / remote start-up meets the dual needs of planned maintenance and emergency cleaning. Multi-dimensional environmental sensing signals trigger cleaning, enabling proactive response to complex downhole contamination scenarios and reducing the risk of monitoring failure.

[0036] The camera body is equipped with a low-light high-definition sensor, supporting imaging in low-light environments. The camera body features IP68 dust and water resistance, ≥10G vibration resistance, and high and low temperature resistance from -20℃ to 60℃. The low-light high-definition sensor ensures clear imaging in low-light underground environments (such as nighttime operations and tunnel corners), eliminating monitoring blind spots; IP68 protection, wide temperature range, and vibration resistance ensure long-term stable operation of the equipment in complex working conditions such as mining faces and transport tunnels.

[0037] The AI ​​processing module uses a confidence correction algorithm to correct the risk scenario identification results. The correction formula is as follows:

[0038]

[0039] Where: R is the corrected AI recognition confidence level, with a value range of 0≤R≤1, used to determine the reliability of the recognition result; The initial identification confidence level has a value range of 0 ≤ ≤1 indicates the initial recognition result of the AI ​​algorithm on the original image; This is a correction factor for the impact of pollution levels, with a value range of 0 < ≤0.5 is used to quantify the degree to which window contamination reduces recognition accuracy; P is the current window contamination level (unit: %). The confidence level is stabilized at a contamination threshold (unit: %), with a range of 10%-20%. This means that when the contamination level is below this value, the impact on the identification confidence level is negligible. This is the image sharpness compensation coefficient, with a value range of 0 < ≤0.3 is used to compensate for the impact of image fluctuations on confidence during the cleaning process; D is the current image sharpness score, which is obtained through an edge detection algorithm. The higher the score, the sharper the image. The standard sharpness score ranges from 80 to 100, serving as the baseline sharpness value under clean conditions; when R ≥ At that time, the AI ​​processing module triggers a real-time alert. The warning activation threshold ranges from 0.7 to 0.9. Through contamination compensation and clarity correction, the reliability of AI recognition under contaminated conditions is effectively improved, reducing false alarms and missed alarms caused by blurry images. The warning threshold can be adjusted according to different risk scenarios to adapt to the stringent safety control requirements of different areas underground.

[0040] The AI ​​processing module integrates an intelligent recognition algorithm to identify three core scenarios: personnel not wearing safety helmets, unauthorized entry, and abnormal equipment operation. Focusing on high-frequency risk scenarios underground, it achieves accurate identification and rapid response to safety hazards, improving regulatory efficiency; covering three core safety dimensions—personnel, area, and equipment—it constructs a comprehensive underground safety early warning system.

[0041] The AI ​​processing module communicates wirelessly with the underground monitoring platform, uploading collected image data, recognition results, and equipment status information in real time, providing remote viewing and parameter configuration capabilities. Wireless communication reduces the safety risks and costs of underground cabling construction and is suitable for installation in complex roadways. Remote parameter configuration supports dynamic adjustment of equipment operating parameters based on working conditions, enabling maintenance without manual intervention underground, thus improving management efficiency.

[0042] Example 2

[0043] according to Figure 1As shown in the figure, this embodiment proposes an explosion-proof intelligent camera with an automatic cleaning system for use in coal mine working faces:

[0044] The explosion-proof intelligent camera with an automatic cleaning system is installed 5m above the coal mining machine's operating area in the underground coal mine. The explosion-proof housing adopts a flameproof structure with an explosion-proof rating of ExdIMb, meeting the requirements for gas and coal dust explosion risk prevention and control at the mining face. The image acquisition module uses a 1080P low-light high-definition sensor, supporting low-light imaging down to 0.005 lux, adapting to the changing lighting conditions at the working face during day and night operations.

[0045] The automatic cleaning module's parameters are set as follows: preset cleaning interval is 4 hours, and the window contamination threshold is set. =40%, Clear Area Weighting Coefficient =0.6, contrast weighting coefficient =0.4, standard average contrast ratio =85. When the coal mining machine generates a large amount of coal dust, the clear area pixel area extracted in real time by the image acquisition module. =1.2 million pixels (total area of ​​effective imaging pixels) =2 million pixels), current image average contrast =42, calculated by the algorithm of claim 3: P=(1-(120 / 200)×0.6-(42 / 85)×0.4)×100%≈(1-0.36-0.197)×100%≈44.3%, since P≥40%, the AI ​​processing module starts automatic cleaning.

[0046] The cleaning process is as follows: high-pressure gas (0.4 MPa) blowing for 10 seconds to remove surface dust; 35℃ high-pressure water (0.6 MPa) rinsing for 8 seconds to soften stubborn coal dust; and silicone scraper wiping for 5 seconds to remove residual deposits. During the cleaning process, the image acquisition module continuously captures images, and the AI ​​processing module corrects the recognition confidence level using the algorithm of claim 6: initial recognition confidence level... =0.65, pollution level P=44.3%, confidence level stable pollution threshold =15%, Pollution level impact correction factor =0.3, Image sharpness compensation coefficient =0.2, current image sharpness score D=65, standard sharpness score =90, after correction R=0.65×(1-0.3×|44.3-15|%)+0.2×(65 / 90)≈0.65×(1-0.0879)+0.2×0.722≈0.592+0.144≈0.736≥0.7, triggering an abnormal warning of loose coal mining machine chain, with a response time of 0.3s.

[0047] Example 3

[0048] according to Figure 1 As shown, this embodiment proposes an explosion-proof intelligent camera with an automatic cleaning system for use in underground coal mines, specifically in transport roadways.

[0049] The calculated window contamination level P = (1 - (130 / 200) × 0.5 - (50 / 85) × 0.5) × 100% ≈ (1 - 0.325 - 0.294) × 100% ≈ 38.1%. Since P ≥ 35% (preset threshold), the AI ​​processing module triggers an automatic cleaning process. The cleaning sequence is as follows: the high-pressure gas generator outputs 0.35MPa dry compressed air, which is blown through three concealed nozzles around the window for 12 seconds to remove the floating dust raised by the belt conveyor; then the high-pressure water flow unit is activated, heating the water flow to 32℃, and rinsing through two nozzles at a pressure of 0.55MPa for 10 seconds to soften and wash away the coal dust and water mist mixture attached to the window; finally, the explosion-proof drive motor drives the silicone scraper assembly to rotate and wipe at a speed of 60r / min for 6 seconds to remove residual deposits.

[0050] During the cleaning process, the image acquisition module continuously captures images of the belt conveyor in operation, and the AI ​​processing module identifies scenarios such as "personnel entering the transport channel without wearing safety helmets," with an initial recognition confidence level. =0.68. Calculated using the confidence correction algorithm of claim 6: contamination level P = 38.1%, confidence level stable contamination threshold. =18%, Pollution level impact correction factor =0.25, image sharpness compensation coefficient =0.25, current image sharpness score D=70 (obtained via Sobel edge detection algorithm), standard sharpness score =90, after correction R=0.68×(1-0.25×|38.1-18|%)+0.25×(70 / 90)≈0.68×(1-0.05025)+0.25×0.778≈0.646+0.194≈0.84≥0.8 (early warning activation threshold), the AI ​​processing module immediately triggered the audible and visual warning and uploaded the warning information to the underground monitoring platform, with a response time of 0.4s, successfully avoiding the risk of collision between personnel and the operating belt conveyor.

[0051] Example 4

[0052] according to Figure 1 As shown, this embodiment proposes an explosion-proof intelligent camera with an automatic cleaning system for use in underground coal mines, specifically in electromechanical chambers.

[0053] In this embodiment, the camera is installed on the top of the underground electromechanical chamber, 3 meters away from the transformer equipment. It is mainly used to monitor the operating status of the switchgear and transformer (such as temperature, abnormal noise, loose wiring, etc.). The explosion-proof enclosure adopts an explosion-proof structure with an explosion-proof rating of ExdIMb. The sealing components of the enclosure are made of fluororubber, which is oil-resistant and aging-resistant, and suitable for the high-temperature environment generated by the equipment operation inside the chamber. The equipment as a whole has the characteristics of high and low temperature resistance from -20℃ to 60℃, IP68 dustproof and waterproof rating, and ≥10G vibration resistance, which can withstand the airflow impact generated by the chamber ventilation system and the vibration of equipment start-up and shutdown.

[0054] The automatic cleaning module adopts a dual-trigger mode of "timed start + contamination sensing", with a preset cleaning interval of 8 hours and a contamination threshold in the display window. =30%. Because the dust concentration in the electromechanical chamber is lower than that in the mining face and transport roadways, the clear area weighting coefficient is... =0.4, contrast weighting coefficient =0.6, standard average contrast ratio =90 (calibrated under clean conditions). When dust accumulates in the ventilation ducts inside the chamber and falls into the camera's viewing window, the image acquisition module extracts the pixel area of ​​the clear area in real time. =1.5 million pixels (total area of ​​effective imaging pixels) =2 million pixels), current image average contrast =60, calculated by the algorithm of claim 3: P=(1-(150 / 200)×0.4-(60 / 90)×0.6)×100%≈(1-0.3-0.4)×100%=30%, reaching the cleaning start threshold, the AI ​​processing module starts the cleaning program.

[0055] The cleaning process has been optimized as follows: high-pressure gas (0.3 MPa) is used to blow away loose dust on the surface for 8 seconds; since there are no stubborn deposits in the chamber, the high-pressure water rinsing step is omitted, and the silicone scraper assembly is directly started to rotate and wipe for 5 seconds, reducing water consumption. After cleaning, the image acquisition module detects abnormal temperature at the transformer terminals (assisted by infrared imaging), and the AI ​​processing module initially identifies the confidence level. =0.72, according to the confidence adjustment algorithm: P=30%, =15%, =0.2, =0.2, D=75, =90, calculated as R=0.72×(1-0.2×|30-15|%)+0.2×(75 / 90)=0.72×(1-0.03)+0.2×0.833≈0.698+0.167≈0.865≥0.85 (equipment abnormality warning threshold), immediately send "transformer terminal high temperature warning" to the monitoring platform, and activate the chamber cooling system in conjunction with it, with a response time of 0.35s, to avoid short circuit faults caused by overheating of the equipment.

[0056] Validation data:

[0057] The performance comparison data of traditional explosion-proof cameras and the present invention in the same underground environment (covering mining faces, transport roadways, and electromechanical chambers) was conducted over a period of 12 months, with a sample size of 50 units / group. The environmental conditions were: dust concentration of 100~500mg / m³, humidity of 60%~95%, temperature of -10℃~45℃, vibration frequency of 5~50Hz, and gas concentration of 0~0.5% (volume fraction).

[0058]

[0059] This explosion-proof intelligent camera with an automatic cleaning system for underground coal mines fundamentally solves the core pain points of traditional explosion-proof cameras, such as easy dust accumulation on the viewing window and difficult maintenance, through its integrated design of explosion-proof housing, automatic cleaning function, and AI processing function. The automatic cleaning module adopts a combination of high-pressure gas blowing, high-pressure water flushing, and silicone scraping, combined with a viewing window contamination degree judgment algorithm, which can automatically start the cleaning action according to the actual contamination level, eliminating the need for frequent manual disassembly. This not only reduces the safety risks of underground maintenance operations but also reduces equipment wear and labor costs, ensuring 24-hour uninterrupted clear monitoring. Furthermore, the explosion-proof structure and environmental adaptability design of this invention are fully adapted to the complex working conditions of underground coal mines. The explosion-proof housing meets coal mine-specific explosion-proof standards such as ExdIMb, and also has IP68 dustproof and waterproof rating, ≥10G vibration resistance, and high and low temperature resistance characteristics from -20℃ to 60℃. It can operate stably in high-risk areas such as mining faces and gas accumulation points, broadening the application scenarios of explosion-proof cameras and filling the technological gap of intelligent monitoring equipment in harsh environments. Meanwhile, through the collaborative work of the AI ​​processing module and the automatic cleaning module, this invention achieves closed-loop management of monitoring data acquisition and intelligent analysis. The low-light high-definition sensor ensures imaging quality in low-light environments, and the AI ​​recognition algorithm combined with the confidence correction algorithm effectively improves the recognition accuracy of risk scenarios such as personnel not wearing safety helmets and abnormal equipment operation. The real-time early warning function provides reliable data support for underground safety supervision and production scheduling, helping the coal industry to achieve intelligent transformation and improve safety management.

[0060] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An explosion-proof intelligent camera with an automatic cleaning system for underground coal mines, comprising a camera body, an automatic cleaning module, and an AI processing module, characterized in that: The outer side of the camera body is an explosion-proof shell, and the explosion-proof shell meets the explosion-proof standards for underground coal mines; The automatic cleaning module is electrically connected to the AI ​​processing module. The automatic cleaning module starts the cleaning action according to preset conditions or environmental sensing signals. The camera body continuously collects image data and transmits it to the AI ​​processing module. The AI ​​processing module performs intelligent analysis and risk warning on the image data.

2. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The explosion-proof enclosure adopts an explosion-proof or increased safety type structure with an explosion-proof rating of not less than ExdIMb. The sealing components of the explosion-proof enclosure are made of oil-resistant rubber, and the surface of the explosion-proof enclosure is provided with an antistatic coating.

3. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The automatic cleaning module is triggered based on the window contamination level calculation result, which uses the following mathematical algorithm: ； Where: P is the window contamination degree (unit: %), which characterizes the degree of influence of contaminants on the window surface on imaging; The pixel area (unit: pixel²) of the clear region in the current image is extracted in real time by the camera itself; The total effective imaging pixel area of ​​the camera body (unit: pixel²) is a fixed value for device calibration. For clear region weighting coefficients, the value range is: 0 < ≤1, based on the preset dust concentration underground; the higher the dust concentration, the better. The larger the value; The average contrast of the current image is calculated from the grayscale values. The standard average contrast value calibrated for the camera body is the baseline contrast value under clean conditions; This is the contrast weighting coefficient, with a value range of 0 < ≤1, and satisfy + =1, used to balance the impact of sharp areas and contrast on contamination levels; when P≥ At that time, the AI ​​processing module sends a start signal to the automatic cleaning module. The preset cleaning start threshold ranges from 30% to 60%.

4. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The automatic cleaning module includes a high-pressure gas generating unit, a high-pressure water flow unit, a silicone scraper assembly, and a drive motor. The nozzles of the high-pressure gas generating unit and the spray nozzles of the high-pressure water flow unit are concealed on the explosion-proof housing around the camera body window. The silicone scraper assembly is driven by an explosion-proof drive motor and rotates to wipe the surface of the window.

5. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 4, characterized in that: The high-pressure water flow unit has the function of heating the water flow to 30-40℃. The heating start-up condition is that the downhole ambient temperature is ≤5℃. The purging pressure of the high-pressure gas is 0.3-0.6MPa, and the flushing pressure of the high-pressure water flow is 0.5-0.8MPa.

6. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The automatic cleaning module includes preset conditions such as timed start and manual / remote start, and environmental sensing signals include window contamination level signal, underground dust concentration signal, and humidity signal.

7. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The camera body is equipped with a low-light high-definition sensor, which supports imaging in low-light environments. The camera body has IP68 dust and water resistance, ≥10G vibration resistance, and high and low temperature resistance from -20℃ to 60℃.

8. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 3, characterized in that: The AI ​​processing module uses a confidence correction algorithm to correct the risk scenario identification results. The correction formula is as follows: ； Where: R is the corrected AI recognition confidence level, with a value range of 0≤R≤1, used to determine the reliability of the recognition result; The initial identification confidence level has a value range of 0 ≤ ≤1 indicates the initial recognition result of the AI ​​algorithm on the original image; This is a correction factor for the impact of pollution levels, with a value range of 0 < ≤0.5 is used to quantify the degree to which window contamination reduces recognition accuracy; P is the current window contamination level (unit: %). The confidence level is stable at a contamination threshold (unit: %), with a range of 10%-20%. This means that when the contamination level is below this value, the impact on the identification confidence level is negligible. This is the image sharpness compensation coefficient, with a value range of 0 < ≤0.3 is used to compensate for the impact of image fluctuations on confidence during the cleaning process; D is the current image sharpness score, which is obtained through an edge detection algorithm. The higher the score, the sharper the image. The standard sharpness score ranges from 80 to 100, serving as the baseline sharpness value under clean conditions; when R ≥ At that time, the AI ​​processing module triggers a real-time alert. The threshold for triggering an early warning is 0.7-0.

9.

9. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The intelligent recognition algorithm integrated into the AI ​​processing module is used to identify three core scenarios: personnel not wearing safety helmets, unauthorized entry, and abnormal equipment operation.

10. The explosion-proof intelligent camera with an automatic cleaning system for underground coal mines according to claim 1, characterized in that: The AI ​​processing module communicates wirelessly with the downhole monitoring platform, uploading the collected image data, recognition results, and equipment status information in real time, providing remote viewing and parameter configuration functions.

Images