Dustproof filter box intelligent life cycle management system

Abstract

The application discloses a dustproof filter box intelligent life cycle management system, and belongs to the technical field of personal protective equipment management, comprising a respiratory resistance intelligent detection device, a self-service distribution and recovery device and a management background system, wherein the self-service distribution and recovery device is in communication connection with the respiratory resistance intelligent detection device, and is used for automatically distributing new dustproof filter boxes of corresponding brands and models to miners in need of replacing dustproof filter boxes according to detection results, and recovering old dustproof filter boxes. Whether the dustproof filter box needs to be replaced is determined through respiratory resistance detection, and the traditional fixed-period replacement mode is replaced, so that resource waste caused by early replacement of the dustproof filter box is avoided, protection failure caused by overuse is eliminated, the risk of pneumoconiosis is effectively reduced, the visual identification function of multiple brands of filter boxes is integrated, the brand and the model of the filter box used by the miner are automatically identified, the self-service distribution and recovery device accurately matches and pops out the new filter box of the corresponding model according to the brand and the model, and the problem of non-matching of the model is avoided.

Description

Technical Field

[0001] This invention relates to the field of personal protective equipment management technology, and more specifically, to an intelligent lifecycle management system for dust filter cartridges. Background Technology

[0002] The mining industry operates in environments with high dust concentrations, and long-term inhalation of dust can easily lead to irreversible occupational diseases such as pneumoconiosis. Dust masks are the most important personal protective equipment for miners, with their core filtering component being the filter cartridge. As dust accumulates during use, the breathing resistance of the filter cartridge gradually increases, and the filtration efficiency decreases. Once the filter exceeds a certain threshold, it will no longer effectively block dust particles, seriously threatening the health of miners. The new national standard GB 18664-2025 clearly stipulates the mandatory replacement criteria for filter cartridges that exceed the breathing resistance standard and requires strengthened life-cycle management of respiratory protective equipment.

[0003] Currently, the following problems are commonly found in the management of dust filter boxes in various mining enterprises: 1. The fixed time period for uniform distribution does not take into account the differences in dust exposure among different positions, resulting in the premature replacement of some filter boxes, which wastes resources, or the use of them beyond their expiration date, which leads to the failure of protection and increases the risk of pneumoconiosis.

[0004] 2. Relying on manual records of distribution and replacement makes it impossible to monitor the actual usage status of each miner's filter box in real time, resulting in management loopholes.

[0005] 3. Long queues during centralized distribution, complicated procedures after miners come up from the mine, and time lost during rest periods all contribute to reduced work enthusiasm.

[0006] 4. The inability to systematically accumulate filter cartridge consumption data makes it difficult to formulate scientific procurement plans and provide data support for safety management decisions.

[0007] While some filter cartridge testing equipment exists in the current technology, most of them are single-channel testing devices, which are inefficient and cannot meet the needs of centralized shift changes in mining teams. Furthermore, they lack automatic identification functions for multi-brand filter cartridges, which can easily lead to model mismatch issues during distribution. At the same time, the testing and distribution processes are separated, failing to form a closed-loop management system, and data cannot be automatically synchronized, making it difficult to achieve full lifecycle traceability of filter cartridges.

[0008] In view of this, the present invention is proposed to solve the above-mentioned technical problems. Summary of the Invention

[0009] The purpose of this invention is to provide an intelligent lifecycle management system for dust filter cartridges, in order to solve the technical problems of low detection efficiency and lack of linkage between dust filter cartridge detection and distribution in the prior art.

[0010] To achieve the above objectives, the present invention provides the following technical solution: A dust filter cartridge intelligent lifecycle management system, comprising: The intelligent breathing resistance detection equipment is used to detect the breathing resistance and brand / model identification of dust filter boxes used by miners, and output the test results. It includes a first industrial control host, a first face recognition camera, 20 test chamber modules, a vision recognition module, and a large screen display. The first face recognition camera is used for miner identity verification. Each test chamber module includes an electric chamber cover, a filter box positioning slot, a differential pressure sensor, an oil-free diaphragm air pump, and a PID flow controller. The PID flow controller is electrically connected to the oil-free diaphragm air pump to drive it to output a constant airflow. The two pressure ports of the differential pressure sensor are connected to the air inlet and air outlet sides of the filter box positioning slot, respectively, to independently complete the breathing resistance detection of a single dust filter box. The vision recognition module includes 20 industrial cameras installed on the top inner wall of each test chamber module, with the lenses vertically aligned with the center of the dust filter box positioning slot, to collect images of the dust filter boxes and identify their brand and model. The large screen display is used to display the team's test results in real time. The self-service dispensing and recycling equipment is connected to the intelligent breathing resistance detection equipment. It is used to automatically dispense new dust filter boxes of the corresponding brand and model to miners who need to replace their dust filter boxes based on the detection results, and to recycle old dust filter boxes. It includes a second industrial control host, a second face recognition camera, a multi-brand classification storage bin, a pop-out mechanism, an old filter box recycling bin, and a touch screen display. The multi-brand classification storage bin is divided into multiple independent sections to store dust filter boxes of different brands and models. Each independent section is equipped with a pop-out mechanism. The pop-out mechanism pops out the matching new dust filter box from the corresponding section based on the detection results. A diffuse reflection photoelectric sensor is installed at the disposal port of the old filter box to detect the insertion status of the old filter box. The management backend system communicates with the intelligent breathing resistance detection device and the self-service dispensing and recycling device, respectively, and is used to store and manage personnel information, device status, detection records and dispensing records, and generate statistical reports.

[0011] Furthermore, the test chamber module also includes a bidirectional air path switching valve, which is connected to the inlet and outlet sides of the oil-free diaphragm air pump and the filter box positioning groove, respectively, to switch the airflow direction to measure inspiratory resistance and expiratory resistance. The test chamber module uses the constant flow method to detect breathing resistance. The oil-free diaphragm air pump is controlled by a PID flow controller to drive a constant airflow of 85L / min through the dust filter box. The differential pressure sensor measures the pressure difference across the dust filter box. When the measured inspiratory resistance exceeds 350Pa or the expiratory resistance exceeds 250Pa, the dust filter box is deemed unqualified.

[0012] Furthermore, the visual recognition module adopts a lightweight convolutional neural network algorithm based on MobileNetV3 or EfficientNet-Lite, combined with the SE attention mechanism to focus on the brand logo area. The initial recognition library covers no less than 20 dust filter models. When the recognition confidence is ≥90%, it is considered a valid recognition. When the recognition confidence is below 90%, the dust filter image sample is automatically saved to the first industrial control host. It supports updating the recognition model via OTA. The industrial camera automatically triggers shooting after the electric compartment cover is closed. The image acquisition and recognition process is carried out in parallel with the pre-stabilization process of the oil-free diaphragm air pump.

[0013] Furthermore, the total testing time of the intelligent breathing resistance detection device is ≤20 seconds. The specific timing sequence is as follows: 2 seconds for airtightness self-test after the electric compartment cover is closed, 3 seconds for the oil-free diaphragm air pump to start and pre-stabilize through PID control, 10 seconds for the differential pressure sensor to continuously sample at a sampling rate of 50Hz, 2 seconds for the first industrial control host to perform data calculation and judgment, 1 second for the result output to the large screen display and the self-service dispensing and recycling equipment, and 2 seconds for the electric compartment cover to open. The dust filter box image acquisition and recognition and the oil-free diaphragm air pump pre-stabilization are performed in parallel without taking up additional testing time.

[0014] Furthermore, the total capacity of the multi-brand categorized storage warehouse is ≥200 sets. Each independent partition is equipped with a photoelectric inventory counting sensor to monitor the remaining quantity of dust filter boxes in the corresponding partition in real time. When the inventory of a certain partition is lower than the set threshold, the second industrial control host automatically sends a shortage warning message containing the partition number and the corresponding brand and model to the management backend system.

[0015] Furthermore, both the first and second face recognition cameras employ RGB+NIR binocular liveness detection technology, using the ArcFace algorithm for feature extraction. The 1:N retrieval recognition rate is ≥99%, the false recognition rate is ≤0.01%, and the response time is ≤2 seconds. Both the first and second industrial control hosts locally store encrypted face feature databases, without storing original face photos. They support offline mode operation, with detection and distribution data generated in offline mode cached locally and automatically synchronized to the management backend system after network recovery.

[0016] Furthermore, the management backend system adopts a B / S architecture, providing functions such as personnel management, equipment management, inspection record query, replacement record query, statistical report generation, threshold management, inventory warning, and brand and model management. It supports setting different dust filter box resistance replacement thresholds according to the dust exposure level of each position, automatically generating statistical reports on average replacement frequency per person, dust filter box consumption comparison of each position, and monthly / annual consumption trends, and supports exporting reports to Excel format.

[0017] Furthermore, the data flow of this intelligent management system is as follows: After a miner completes facial recognition in front of the intelligent breathing resistance detection device, the first industrial control host allocates an empty test chamber, and the electric cover of the corresponding test chamber opens automatically; after the miner places and secures the dust filter box, the electric cover closes, and the intelligent breathing resistance detection device simultaneously collects breathing resistance data and filter box images, outputs the test results, and pushes them to a large screen display and a self-service distribution and recycling device; after the miner completes facial recognition in front of the self-service distribution and recycling device, the second industrial control host queries the miner's test results, and if a replacement is needed, it controls the pop-up mechanism of the corresponding partition to pop out a new dust filter box of the matching model; after the miner takes out the new dust filter box and puts the old dust filter box into the old filter box recycling bin, the diffuse reflection photoelectric sensor confirms that the old filter box has been put in, the second industrial control host completes the distribution process and records the distribution information; all test data and distribution data are synchronized to the management backend system in real time, forming a complete lifecycle ledger of personnel-time-filter box usage.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By detecting breathing resistance, the need for dust filter cartridge replacement is determined, replacing the traditional fixed-period replacement method. This avoids the waste of resources caused by premature replacement of dust filter cartridges and eliminates the protective failure caused by overuse, effectively reducing the risk of pneumoconiosis. It fully complies with the requirements of the new national standard GB 18664-2025 for determining obsolescence based on actual protective performance.

[0019] 2. The system adopts 20 parallel detection channels, with a single detection time of ≤20 seconds, which can meet the detection needs of a shift at the same time, greatly improving shift change efficiency and reducing miners' waiting time.

[0020] 3. It integrates multi-brand filter cartridge visual recognition function, automatically identifies the brand and model of dust filter cartridges used by miners, and the self-service distribution and recycling equipment accurately matches and pops out the corresponding model of new filter cartridges, avoiding the problem of model mismatch.

[0021] 4. It realizes a closed loop of the entire process of testing, distribution, recycling and data management, and automatically generates a complete life cycle ledger of personnel-time-dust filter box usage, meeting the full life cycle traceability management requirements in the new national standard.

[0022] 5. It has a comprehensive safety design, including anti-pinch function for electric compartment cover and pop-out mechanism, UPS protection against power failure, and encrypted storage of facial data, to ensure the safe operation of equipment and the security of miners' personal information.

[0023] 6. Supports setting differentiated replacement thresholds based on job dust exposure levels, and can generate multi-dimensional statistical reports to provide scientific data support for mine safety management and procurement planning. Attached Figure Description

[0024] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation of the invention. Obviously, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without creative effort. In the drawings: Figure 1 This is a schematic diagram of the structure of the intelligent lifecycle management system for dust filter cartridges provided in the embodiments of this application; Figure 2 A schematic diagram of the breathing resistance intelligent detection device of the dust filter cartridge intelligent life cycle management system provided in this application embodiment; Figure 3 A schematic diagram of the structure of a single test unit of the intelligent lifecycle management system for dust filter cartridges provided in this application embodiment; Figure 4 A forward view of a self-service dispensing and recycling device for a dust filter cartridge intelligent lifecycle management system provided in this application embodiment; Figure 5 A side view of the self-service distribution and recycling equipment of the intelligent lifecycle management system for dust filter cartridges provided in this application embodiment. Detailed Implementation

[0025] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0026] See Figures 1 to 5As shown, an intelligent lifecycle management system for dust filter cartridges includes: an intelligent breathing resistance detection device, a self-service dispensing and recycling device, and a management backend system. The intelligent breathing resistance detection device is used to detect the breathing resistance and identify the brand and model of the dust filter cartridges after use by miners, and outputs the detection results. It includes a first industrial control host, a first face recognition camera, 20 test chamber modules, a vision recognition module, and a large-screen display. The first face recognition camera is used for miner identity verification. Each test chamber module includes an electric chamber cover, a filter cartridge positioning slot, a differential pressure sensor, an oil-free diaphragm air pump, and a PID flow controller. The PID flow controller is electrically connected to the oil-free diaphragm air pump to drive it to output a constant airflow. The two pressure ports of the differential pressure sensor are connected to the inlet and outlet sides of the filter cartridge positioning slot, respectively, to independently complete the breathing resistance detection of a single dust filter cartridge. The vision recognition module includes 20 industrial cameras installed on the top inner wall of each test chamber module, with their lenses vertically aligned with the center of the filter cartridge positioning slot, for... The system collects images of dust filter cartridges and identifies their brand and model. A large-screen display shows the real-time test results from the work team. The self-service distribution and recycling equipment communicates with the intelligent breathing resistance detection equipment. Based on the test results, it automatically distributes new dust filter cartridges of the corresponding brand and model to miners who need to replace their dust filter cartridges and recycles old dust filter cartridges. The system includes a second industrial control host, a second facial recognition camera, a multi-brand classification storage bin, a pop-out mechanism, an old filter cartridge recycling bin, and a touch screen display. The multi-brand classification storage bin is divided into multiple independent sections, each storing different brands and models of dust filter cartridges. Each independent section has a corresponding pop-out mechanism that pops out a matching new dust filter cartridge from the corresponding section based on the test results. A diffuse reflection photoelectric sensor is installed at the disposal port of the old filter cartridge recycling bin to detect the insertion status of the old filter cartridge. The management backend system communicates with both the intelligent breathing resistance detection equipment and the self-service distribution and recycling equipment to store and manage personnel information, equipment status, test records, and distribution records, and to generate statistical reports.

[0027] The test chamber module also includes a bidirectional airflow switching valve, which is connected to the inlet and outlet sides of the oil-free diaphragm air pump and the filter cartridge positioning slot, respectively, to switch the airflow direction to measure inspiratory and expiratory resistance. The test chamber module uses a constant flow method to detect breathing resistance. A PID flow controller controls the oil-free diaphragm air pump to drive a constant airflow of 85 L / min through the dust filter cartridge. A differential pressure sensor measures the pressure difference across the dust filter cartridge. When the measured inspiratory resistance exceeds 350 Pa or the expiratory resistance exceeds 250 Pa, the dust filter cartridge is deemed unqualified. Atmosphere is connected to the first port of the bidirectional airflow switching valve through the dust filter screen. The second port of the bidirectional airflow switching valve is connected to the inlet side of the filter cartridge positioning slot, and the outlet side of the filter cartridge positioning slot is connected to the third port of the bidirectional airflow switching valve. The fourth port of the bidirectional air path switching valve is connected to the inlet of the oil-free diaphragm air pump, and the outlet of the oil-free diaphragm air pump is connected to the exhaust port. The high-pressure measuring port of the differential pressure sensor is connected to the inlet side of the filter cartridge positioning slot, and the low-pressure measuring port is connected to the outlet side of the filter cartridge positioning slot. When measuring inspiratory resistance, the bidirectional air path switching valve switches to the first state, allowing atmospheric air to pass through the dust filter, the inlet side of the filter cartridge positioning slot, the filter cartridge, the outlet side of the filter cartridge positioning slot, and the oil-free diaphragm air pump. The pressure difference measured by the differential pressure sensor at this time is the inspiratory resistance. When measuring expiratory resistance, the bidirectional air path switching valve switches to the second state, allowing the airflow output by the oil-free diaphragm air pump to flow in reverse through the outlet side of the filter cartridge positioning slot, the filter cartridge, and the inlet side of the filter cartridge positioning slot before being discharged. The pressure difference measured by the differential pressure sensor at this time is the expiratory resistance. The oil-free diaphragm air pump is controlled by a closed-loop PID flow controller to maintain a constant test flow rate of 85 L / min, ensuring the accuracy and consistency of the test results.

[0028] The visual recognition module adopts a lightweight convolutional neural network algorithm based on MobileNetV3 or EfficientNet-Lite, combined with the SE attention mechanism to focus on the brand logo area. The initial recognition library covers no less than 20 dust filter models. When the recognition confidence is ≥90%, it is considered a valid recognition. When the recognition confidence is below 90%, the dust filter image sample is automatically saved to the first industrial control host. It supports OTA push to update the recognition model. The industrial camera automatically triggers shooting after the electric compartment cover is closed. The image acquisition and recognition process is carried out in parallel with the pre-stabilization process of the oil-free diaphragm air pump.

[0029] The visual recognition module adopts a lightweight convolutional neural network algorithm based on MobileNetV3-Small. An SE (Squeeze-and-Excitation) attention module is inserted into the bottleneck layer of the network. By adaptively calibrating the channel feature weights, it focuses on the brand logo and model label area on the filter box surface, suppresses the interference of background noise, and improves the recognition accuracy.

[0030] The initial recognition library covers no fewer than 20 filter cartridge models from mainstream brands such as 3M, Honeywell, Youan, Haigu, and Baoweikang. For each model, at least 500 training samples were collected under different angles, lighting conditions, and wear levels. After data augmentation (including rotation, scaling, brightness adjustment, and noise addition), the initial model was trained. The model was quantized to INT8 precision, achieving an inference speed of ≤100ms per frame on an Intel Core i5 industrial PC.

[0031] When the recognition confidence level is below 90%, the system automatically marks the filter box image as a training sample and saves it to the first industrial control host. The administrator can view the training samples through the management backend system, label the brand and model, and then upload them to the cloud training platform. After the cloud training platform completes the model iteration training, it pushes the updated model to all intelligent breathing resistance detection devices via OTA, without requiring on-site downtime for upgrades, thus achieving continuous expansion of the recognition library. The total testing time of the intelligent breathing resistance detection device is ≤20 seconds. The specific timing sequence is as follows: 2 seconds for airtightness self-test after the electric compartment cover is closed, 3 seconds for pre-stabilization through PID control after the oil-free diaphragm air pump starts, 10 seconds for continuous sampling by the differential pressure sensor at a sampling rate of 50Hz, 2 seconds for data calculation and judgment by the first industrial control host, 1 second for outputting the results to the large screen display and the self-service dispensing and recycling equipment, and 2 seconds for the electric compartment cover to open. The image acquisition and recognition of the dust filter box and the pre-stabilization of the oil-free diaphragm air pump are carried out in parallel without taking up additional testing time.

[0032] The total capacity of the multi-brand categorized storage warehouse is ≥200 sets. Each independent partition is equipped with a photoelectric inventory counting sensor to monitor the remaining quantity of dust filter boxes in the corresponding partition in real time. When the inventory of a certain partition is lower than the set threshold, the second industrial control host automatically sends a stockout warning message containing the partition number and the corresponding brand and model to the management backend system.

[0033] Both the first and second face recognition cameras employ RGB+NIR binocular liveness detection technology, using the ArcFace algorithm for feature extraction. The 1:N retrieval recognition rate is ≥99%, the false recognition rate is ≤0.01%, and the response time is ≤2 seconds. Both the first and second industrial control hosts locally store encrypted face feature databases, without storing original face photos. They support offline mode operation, with detection and distribution data generated in offline mode cached locally and automatically synchronized to the management backend system after network recovery.

[0034] The management backend system adopts a B / S architecture and provides functions such as personnel management, equipment management, inspection record query, replacement record query, statistical report generation, threshold management, inventory warning and brand and model management. It supports setting different dust filter box resistance replacement thresholds according to the dust exposure level of each position, automatically generates statistical reports on average replacement frequency per person, dust filter box consumption comparison of each position, and monthly / annual consumption trends, and supports exporting reports to Excel format.

[0035] The data flow of this intelligent management system is as follows: After a miner completes facial recognition in front of the intelligent breathing resistance detection device, the first industrial control host allocates an empty test chamber, and the electric cover of the corresponding test chamber opens automatically; after the miner places and secures the dust filter box, the electric cover closes, and the intelligent breathing resistance detection device simultaneously collects breathing resistance data and filter box images, outputs the test results, and pushes them to the large screen display and the self-service distribution and recycling equipment; after the miner completes facial recognition in front of the self-service distribution and recycling equipment, the second industrial control host queries the miner's test results, and if a replacement is needed, it controls the pop-up mechanism of the corresponding partition to pop out a new dust filter box of the matching model; after the miner takes out the new dust filter box and puts the old dust filter box into the old filter box recycling bin, the diffuse reflection photoelectric sensor confirms that the old filter box has been put in, the second industrial control host completes the distribution process and records the distribution information; all test data and distribution data are synchronized to the management backend system in real time, forming a complete personnel-time-filter box usage lifecycle ledger.

[0036] Example 1: like Figure 2 As shown, the intelligent breathing resistance testing equipment adopts a vertical cabinet structure with external dimensions of 3000mm wide × 450mm deep × 2000mm high. The cabinet is made of cold-rolled steel plate with a powder coating and a surface anti-corrosion and moisture-proof treatment, with an IP54 protection rating. A face recognition camera, approximately 1800mm high, is installed at the center of the top of the equipment, capable of recognizing faces within a range of 1.5-2.0m. Below the top is a 65-inch 4K industrial high-brightness display screen with a brightness ≥600nit, used to display the team's testing results. Below the display screen are 20 testing compartments, arranged horizontally in two rows. Each compartment is approximately 130mm wide, 100mm high at the opening, and 200mm deep. Each compartment is equipped with a red and green dual-color indicator light and a compartment number label.

[0037] like Figure 3As shown, a single test chamber unit mainly consists of a chamber cover, a filter cartridge positioning slot, a differential pressure sensor, an oil-free diaphragm air pump, and indicator lights. The chamber cover is driven by an electric push rod with a stroke of 50mm and a force of 100N. It is equipped with an airtight ring and a hardware limit switch, and the edges feature a flexible anti-pinch design. The filter cartridge positioning slot adopts an elastic clamping structure, which can accommodate various filter cartridge models. The differential pressure sensor has a range of 0-500Pa, an accuracy of ±2Pa, and includes temperature compensation. It is connected to the main controller via an RS485 interface. The oil-free diaphragm air pump has a rated flow rate of 100L / min, is powered by 24VDC, has a noise level of ≤60dB, and maintains a constant test flow rate of 85L / min through PID closed-loop control.

[0038] The main controller of the inspection equipment is an Intel Core i5 industrial PC with 8GB RAM, 256GB SSD, and running Ubuntu 20.04. Twenty industrial vision recognition cameras are connected to the PC via a USB 3.0 hub. Each camera is 5 megapixels, has built-in LED lighting, and IP40 protection. The equipment is equipped with an 8-port industrial Ethernet switch and a 1KVA UPS uninterruptible power supply.

[0039] Example 2: See Figure 4 and Figure 5 As shown, the self-service dispensing and recycling equipment also adopts a vertical cabinet structure, with external dimensions of 800mm wide × 600mm deep × 1900mm high, and an IP54 protection rating. A second facial recognition camera, approximately 1600mm high, is installed at the top front of the cabinet. Below the camera is a 10.1-inch IPS touchscreen display for showing operation guidance and retrieval results. Below the display are 20 pop-up retrieval slots, each capable of storing two dust filter cartridges. Inside the cabinet are multi-brand filter cartridge categorized storage compartments, divided into five independent sections, each storing different brands and models of dust filter cartridges. Each independent section is equipped with an independent conveyor belt, a stepper motor pop-out mechanism, and a photoelectric inventory counting sensor. At the bottom of the cabinet is a removable used filter cartridge recycling bin with a capacity of ≥200 sets. A diffuse reflection photoelectric sensor is installed at the recycling port to detect whether a used filter cartridge has been inserted.

[0040] The main controller of the self-service distribution and recycling equipment is an Intel Core i5 industrial computer with 8GB RAM, 128GB SSD, and a 500VA UPS uninterruptible power supply.

[0041] Example 3: (1) Working process of intelligent respiratory resistance detection equipment The complete workflow of the intelligent respiratory resistance detection device is as follows: After miners come to the surface and arrive at the intelligent breathing resistance detection equipment, the system automatically activates facial recognition and completes identity verification within 2 seconds. After successful authentication, the first industrial control host allocates an empty test chamber, the red and green dual-color indicator lights of the corresponding chamber flash, the electric chamber cover opens automatically, prompting the miner to put in the dust filter box; A diffuse reflection photoelectric sensor is installed in the filter box positioning slot. When it detects that the filter box is correctly placed and fixed, the electric compartment cover closes automatically and performs a 2-second airtightness self-check. The oil-free diaphragm air pump starts, and the PID flow controller stabilizes the flow rate at 85L / min within 3 seconds. At the same time, the industrial camera in the warehouse automatically captures images of the dust filter box for brand and model identification. The differential pressure sensor continuously collects differential pressure data for 10 seconds at a sampling rate of 50Hz. The first industrial control host filters out noise and takes the average value. The system compares the measured resistance value with the set threshold for the miner's corresponding position. If the value is within the threshold, the corresponding warehouse light will turn on green; otherwise, the light will turn on red. The test results are pushed to the large screen display and the self-service distribution and recycling equipment in real time. The electric bin cover opens automatically, and the miners retrieve their dust masks.

[0042] (2) Equipment distribution process The complete workflow for self-service distribution and recycling equipment is as follows: When miners arrive at the self-service dispensing and recycling equipment, the system automatically initiates facial recognition. The system checks the miner's test results. If no replacement is needed, the screen displays "Your dust filter does not need to be replaced. Please continue using it." If replacement is required, the system will control the ejection mechanism of the corresponding compartment to eject the new dust filter box according to the identified dust filter box brand and model. The corresponding compartment indicator light will light up, and the screen will prompt "Please take out the new dust filter box (XX brand) and put the old dust filter box into the recycling port". The miners took out the new dust filter box and put the old dust filter box into the recycling port; Once the recycling sensor confirms that the old dust filter box has been put in, the compartment automatically closes, and the system records the distribution information and uploads it to the backend.

[0043] Example 4: The management backend system is developed using a MySQL database and the Spring Boot framework, providing a web-based access interface. Its core functionalities include: Personnel Management: Supports miner information entry, facial photo collection, department and work group configuration, and real-time synchronization of the facial database to the device. Equipment Management: Real-time monitoring of the operating status, online / offline status, and alarm records of two devices; supports remote parameter configuration. Record Query: Historical testing records and issuance / recovery records can be filtered and queried by time, personnel, test results, and other criteria; Statistical reports: Automatically generates reports such as average filter cartridge replacement frequency per person, filter cartridge consumption comparison for each position, and monthly / annual consumption trends, and supports exporting to Excel; Threshold management: Supports setting differentiated filter cartridge resistance replacement thresholds for different job positions to adapt to working environments with different dust concentrations; Inventory Management: Real-time monitoring of the inventory of dust filter cartridges of various brands; when the inventory falls below the set threshold, an automatic replenishment alert is sent to the administrator via SMS and system message. Brand and Model Management: Maintains the dust filter box brand and model library and adaptation mapping table, and supports adding new models, uploading training samples, and pushing recognition model updates.

[0044] Example 5: This system features comprehensive safety and reliability design to ensure stable operation in complex mining environments: Mechanical safety: All moving parts (electric push rod, stepper motor pop-out mechanism) are equipped with limit switches to prevent damage from excessive travel; the edges of the electric compartment cover and pop-out mechanism adopt a flexible anti-pinch design, and will automatically reverse if the resistance increases beyond the set threshold during the closing process; at the beginning of each shift, the system automatically runs a full compartment self-check program. Compartments that fail the self-check are automatically disabled and a "Maintenance in Progress" message is displayed on the large screen, while an alarm message is sent to the management backend system.

[0045] Electrical safety: The entire equipment is reliably grounded and equipped with a residual current circuit breaker; all control circuits use 24V low-voltage power supply, and 220V AC power is only used for powering the industrial control host and monitor. Strong and weak current are wired separately and isolated by baffles; equipped with UPS uninterruptible power supply, the testing equipment is equipped with a 1KVA UPS, which can support the completion of all ongoing testing processes and safe shutdown after a power outage, and the dispensing equipment is equipped with a 500VA UPS, which can support the completion of the current dispensing process.

[0046] Data security: Facial feature data is encrypted and stored using the AES-256 algorithm. No original facial photos are stored on the device or in the backend. Local data is automatically backed up to the backend server every hour to prevent single-point data loss. All operations are logged in their entirety, including personnel login, detection operations, distribution operations, parameter modifications, etc. The logs are kept for no less than 3 years and support audit traceability.

[0047] Troubleshooting: When a single component such as the differential pressure sensor or the oil-free diaphragm air pump malfunctions, only the corresponding compartment is disabled, while the remaining compartments operate normally without affecting overall use; when the network is interrupted, the system automatically switches to offline mode, and all detection and dispensing data are cached locally. Once the network is restored, the data is automatically resumed from the interruption point to the backend; when the filter cartridge identification fails, the system prompts the miner to contact the administrator to manually confirm the brand and model, and automatically saves the image sample for subsequent model training.

[0048] 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 preferred examples and are not intended to limit 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. A dust filter cartridge intelligent lifecycle management system, characterized in that, include: The intelligent breathing resistance detection device is used to detect the breathing resistance and brand / model identification of dust filter boxes used by miners, and output the detection results. It includes a first industrial control host, a first facial recognition camera, 20 test chamber modules, a vision recognition module, and a large-screen display. The first facial recognition camera is used for miner identity verification. Each test chamber module includes an electric chamber cover, a filter box positioning slot, a differential pressure sensor, an oil-free diaphragm air pump, and a PID flow controller. The PID flow controller is electrically connected to the oil-free diaphragm air pump to drive it to output a constant airflow. The two pressure ports of the differential pressure sensor are connected to the inlet and outlet sides of the filter box positioning slot, respectively, to independently complete the breathing resistance detection of a single dust filter box. The vision recognition module includes 20 industrial cameras installed on the top inner wall of each test chamber module, with lenses vertically aligned with the center of the filter box positioning slot, to acquire images of the dust filter boxes and identify their brand and model. The large-screen display is used to display the team's detection results in real time. The self-service dispensing and recycling equipment is communicatively connected to the intelligent breathing resistance detection equipment. It is used to automatically dispense new dust filter boxes of the corresponding brand and model to miners who need to replace their dust filter boxes based on the detection results, and to recycle old dust filter boxes. The equipment includes a second industrial control host, a second facial recognition camera, a multi-brand classification storage compartment, a pop-out mechanism, an old filter box recycling bin, and a touch screen display. The multi-brand classification storage compartment is divided into multiple independent partitions, which store dust filter boxes of different brands and models. Each independent partition is equipped with a corresponding pop-out mechanism. The pop-out mechanism pops out a matching new dust filter box from the corresponding partition based on the detection results. A diffuse reflection photoelectric sensor is installed at the dispensing port of the old filter box to detect the insertion status of the old filter box. The management backend system communicates with the intelligent breathing resistance detection device and the self-service dispensing and recycling device, respectively, and is used to store and manage personnel information, device status, detection records and dispensing records, and generate statistical reports.

2. The intelligent lifecycle management system for dust filter cartridges according to claim 1, characterized in that, The test chamber module also includes a bidirectional air path switching valve, which is connected to the inlet side and outlet side of the oil-free diaphragm air pump and the filter cartridge positioning groove, respectively, and is used to switch the airflow direction to measure the inhalation resistance and exhalation resistance respectively. The test chamber module uses a constant flow method to detect breathing resistance. The PID flow controller controls the oil-free diaphragm air pump to drive a constant airflow of 85L / min through the dust filter box. The differential pressure sensor measures the pressure difference across the dust filter box. When the measured inspiratory resistance exceeds 350Pa or the expiratory resistance exceeds 250Pa, the dust filter box is deemed unqualified.

3. The intelligent lifecycle management system for dust filter cartridges according to claim 2, characterized in that, The visual recognition module employs a lightweight convolutional neural network algorithm based on MobileNetV3 or EfficientNet-Lite, combined with an SE attention mechanism to focus on the brand logo area. The initial recognition library covers no fewer than 20 dust filter models. When the recognition confidence is ≥90%, it is considered a valid recognition. When the recognition confidence is below 90%, the dust filter image sample is automatically saved to the first industrial control host. It supports OTA updates to the recognition model. The industrial camera automatically triggers shooting after the electric compartment cover is closed. The image acquisition and recognition process is performed in parallel with the pre-stabilization process of the oil-free diaphragm air pump.

4. The intelligent lifecycle management system for dust filter cartridges according to claim 3, characterized in that, The total detection time of the intelligent breathing resistance detection device is ≤20 seconds. The specific timing sequence is as follows: after the electric compartment cover is closed, the airtightness self-test is started for 2 seconds; the oil-free diaphragm air pump is started and pre-stabilized by PID control for 3 seconds; the differential pressure sensor continuously samples at a sampling rate of 50Hz for 10 seconds; the first industrial control host performs data calculation and judgment for 2 seconds; the result is output to the large screen display and the self-service dispensing and recycling equipment for 1 second; and the electric compartment cover is opened for 2 seconds. The image acquisition and recognition of the dust filter box and the pre-stabilization of the oil-free diaphragm air pump are performed in parallel without occupying additional detection time.

5. The intelligent lifecycle management system for dust filter cartridges according to claim 4, characterized in that, The total capacity of the multi-brand categorized storage warehouse is ≥200 sets. Each independent partition is equipped with a photoelectric inventory counting sensor to monitor the remaining quantity of dust filter boxes in the corresponding partition in real time. When the inventory of a certain partition is lower than the set threshold, the second industrial control host automatically sends a stockout warning message containing the partition number and the corresponding brand and model to the management backend system.

6. The intelligent lifecycle management system for dust filter cartridges according to claim 5, characterized in that, Both the first and second face recognition cameras employ RGB+NIR binocular liveness detection technology, using the ArcFace algorithm for feature extraction. The 1:N retrieval recognition rate is ≥99%, the false recognition rate is ≤0.01%, and the response time is ≤2 seconds. Both the first and second industrial control hosts locally store encrypted face feature databases, but do not store original face photos. They support offline mode operation, and the detection and distribution data generated in offline mode are cached locally and automatically synchronized to the management backend system after the network is restored.

7. The intelligent lifecycle management system for dust filter cartridges according to claim 5, characterized in that, The management backend system adopts a B / S architecture and provides functions such as personnel management, equipment management, inspection record query, replacement record query, statistical report generation, threshold management, inventory warning and brand and model management. It supports setting different dust filter box resistance replacement thresholds according to the dust exposure level of each position, automatically generates statistical reports on average replacement frequency per person, dust filter box consumption comparison of each position, and monthly / annual consumption trends, and supports exporting reports to Excel format.

8. The intelligent lifecycle management system for dust filter cartridges according to any one of claims 1-7, characterized in that, The data flow of the intelligent management system is as follows: after the miner completes facial recognition in front of the breathing resistance intelligent detection device, the first industrial control host allocates an empty test chamber, and the electric chamber cover of the corresponding test chamber opens automatically. After the miner places and secures the dust filter box, the electric compartment cover closes, and the intelligent breathing resistance detection device simultaneously collects breathing resistance data and filter box images, outputs the detection results, and pushes them to the large screen display and the self-service dispensing and recycling equipment. After the miner completes facial recognition in front of the self-service dispensing and recycling equipment, the second industrial control host queries the miner's detection results. If a replacement is required, it controls the pop-out mechanism of the corresponding partition to pop out a new dust filter box of the matching model. After the miner takes out the new dust filter box and puts the old dust filter box into the old filter box recycling bin, the diffuse reflection photoelectric sensor confirms that the old filter box has been put in. The second industrial control host completes the distribution process and records the distribution information. All detection data and distribution data are synchronized to the management backend system in real time to form a complete personnel-time-filter box usage life cycle ledger.

Images