An electronic atomizer leak-proof liquid intelligent monitoring system and method

Abstract

This invention relates to the field of leak detection technology, specifically to an intelligent leak detection method for electronic atomizers. The method includes the following steps: Capacitive, pressure, and temperature / humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. During initial startup calibration, images of the airflow channel are acquired using a miniature camera and infrared light source. Droplet contours are extracted using an edge detection algorithm. Leakage event feedback data is collected via a user app, real leak scenarios are labeled, the training set is updated, and risk levels are converted into specific warning signals. Notifications are sent to the user via device vibration, LED lights, and app push notifications. In cases of high risk, the power is automatically cut off, maintenance suggestions are generated, and a complete leak-proof closed-loop control scheme is obtained. This invention solves the problems of traditional atomizer leak detection methods being singular, unable to provide graded warnings of leak risks, and unable to dynamically adjust monitoring parameters to achieve effective leak-proof closed-loop control.

Description

Technical Field

[0001] This invention belongs to the field of leak detection technology, specifically relating to an intelligent monitoring system and method for leak detection liquid in electronic atomizers. Background Technology

[0002] In various atomizing device applications, leakage prevention has always been a key factor affecting device performance and user experience. Traditional atomizing devices primarily rely on single mechanical detection devices, such as simple float switches, to monitor for leaks. While these devices can detect leaks to some extent, they have several limitations. First, mechanical detection devices have low accuracy, often failing to accurately detect minor leaks. Alarms are only triggered when the leak reaches a certain level, causing a noticeable change in liquid level. This can lead to delays in detecting leaks in their early stages, potentially causing equipment damage or environmental impact. Second, their functionality is limited; they can only detect the presence of a leak, but cannot assess or classify the risk level, making it difficult to take appropriate countermeasures based on different risk levels. Furthermore, traditional detection methods lack the ability to collect and analyze leak data, failing to continuously optimize monitoring parameters based on feedback from actual use, resulting in limited improvement in monitoring effectiveness.

[0003] Existing technologies suffer from several drawbacks: limited detection methods, relying on only a small number of sensor types, making it difficult to comprehensively acquire information on the liquid state inside atomizing devices; inaccurate risk assessment of leakage, making it impossible to provide tiered early warnings; and a lack of data feedback and optimization mechanisms, making it difficult to dynamically adjust monitoring parameters. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides an intelligent monitoring system and method for preventing leakage in electronic atomizers. This system solves the problems of traditional atomizing devices, such as limited monitoring methods, difficulty in comprehensively and accurately acquiring internal liquid status information, inability to provide graded early warnings of leakage risks, lack of data feedback optimization mechanisms, and inability to dynamically adjust monitoring parameters to achieve effective closed-loop control for preventing leakage. To achieve the above objectives, this invention adopts the following technical solution: The intelligent monitoring method for preventing leakage in an electronic atomizer includes the following steps: Capacitive, pressure, and temperature / humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. During initial startup, a calibration procedure is performed to extract sensor baseline parameters and set thresholds to obtain an initial dataset. Images of the airflow channel are acquired using a miniature camera and an infrared light source. An edge detection algorithm is used to extract droplet contours, and morphological processing is combined to eliminate noise. The droplet coverage area ratio and dynamic change rate are used as key indicators to obtain condensate risk characteristics. Sensor data and visual detection results are normalized and processed using a fuzzy logic module. The system assesses risk levels—high, medium, and low—by using a pre-defined rule base, based on factors such as capacitance change rate, pressure fluctuation, temperature and humidity difference, and droplet area, resulting in three levels of leak warnings. Leakage event feedback data is collected via a user app, real-world leak scenarios are labeled, the training set is updated, and historical data is analyzed using clustering algorithms to extract high-frequency leak patterns. Monitoring parameters are then adjusted to obtain an optimized dynamic monitoring module. Risk levels are converted into specific warning signals, which are then communicated to users via equipment vibration, LED lights, and app push notifications. In cases of high risk, the system automatically cuts off power, locks the heating function, and generates maintenance suggestions, resulting in a complete leak-proof closed-loop control solution.

[0005] Furthermore, capacitive, pressure, and temperature / humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. During the initial startup calibration procedure, sensor baseline parameters are extracted and thresholds are set to obtain an initial dataset. This includes the following steps: a capacitive sensor is deployed at the bottom of the liquid storage tank to monitor changes in the liquid dielectric constant in real time to reflect the liquid level; a miniature pressure sensor is installed at the sealing interface of the atomizing core to capture abnormal pressure fluctuations caused by seal failure; a temperature / humidity sensor is integrated into the airflow channel to continuously track the ambient temperature and humidity gradient and predict the risk of condensation formation; and the automated calibration process executed during the initial startup of the device extracts baseline parameters of the sensors under empty tank, standard pressure, and normal temperature dry environments, respectively. The collected values ​​are compared with preset safety ranges, and dynamic thresholds are set to obtain the initial dataset.

[0006] Furthermore, the process of acquiring airflow channel images using a miniature camera and infrared light source, extracting droplet contours using an edge detection algorithm, and eliminating noise through morphological processing, using the droplet coverage area ratio and dynamic change rate as key indicators to obtain condensate risk characteristics, includes the following steps: using a miniature camera paired with a narrow-band infrared light source to directionally acquire images of the inner wall of the airflow channel at a frequency of eight frames per second, ensuring clear capture of droplet adhesion under low light conditions; accurately extracting droplet contour boundaries using the Canny edge detection algorithm, and using a morphological processing method combining opening and closing operations to eliminate small particle noise and uneven lighting interference in the image; performing pixel statistics on the processed binarized image, calculating the ratio of droplet coverage area to the channel cross-sectional area, tracking the area change rate of three consecutive frames, and integrating these into the core feature indicators of condensate accumulation risk.

[0007] Furthermore, the normalization processing of sensor data and visual detection results, using a fuzzy logic module to integrate capacitance change rate, pressure fluctuation, temperature and humidity difference, and droplet area, and assessing the risk level through a preset rule base to obtain high, medium, and low leakage warning results, includes the following steps: Using a minimum-maximum normalization method, the liquid level change rate collected by the capacitive sensor, the fluctuation value monitored by the pressure sensor, the difference recorded by the temperature and humidity sensor, and the droplet coverage area ratio extracted by the visual module are uniformly mapped to the zero-to-one interval; a multi-input single-output module is constructed through the fuzzy logic module, using the parameters collected by the sensor as input variables, and a preset membership function is used to quantify the degree of high, medium, and low risk; reasoning calculations are performed through a rule base formulated by expert experience, and after comprehensive evaluation, a risk level weight value is output. The weight value is then converted into clear high, medium, and low leakage warning results according to a threshold classification standard.

[0008] Furthermore, the process of collecting leakage event feedback data through the user APP, labeling real leakage scenarios, updating the training set, analyzing historical data using a clustering algorithm, extracting high-frequency leakage patterns, and adjusting monitoring parameters to obtain an optimized dynamic monitoring module includes the following steps: Actively collecting leakage event data reported during device use through the feedback entry point built into the user APP; the leakage event data includes the occurrence time, environmental conditions, and original sensor records; manually labeling the feedback data by professionals to clearly distinguish between real leakage and false alarm scenarios, and expanding the labeled data into the existing training set; analyzing historical leakage events using the DBSCAN density clustering algorithm to extract frequently occurring spatiotemporal feature combinations; and adjusting monitoring parameters such as capacitance threshold and pressure fluctuation range based on the pattern analysis results to obtain the optimized dynamic monitoring module.

[0009] Furthermore, the process of converting risk levels into specific warning signals, notifying users via device vibration, LED lights, and APP push notifications, automatically cutting off power and locking the heating function in high-risk situations, generating maintenance suggestions, and obtaining a complete leak-proof closed-loop control scheme includes the following steps: Using a graded coding rule to convert risk levels into differentiated warning signals, the warning signals include: low risk triggers a single short vibration and a flashing green LED; medium risk triggers continuous vibration and a constantly lit yellow LED; high risk executes three strong vibrations combined with a high-frequency flashing red LED; a real-time connection is established between the device's Bluetooth module and the user's APP, synchronously pushing full information including risk type, location of occurrence, and suggested operations; when a high-risk state is detected, the power supply circuit is immediately cut off and the heating function is disabled, while extracting typical fault modes based on historical data to generate a targeted maintenance guide including cleaning and seal replacement steps, thus obtaining a complete leak-proof closed-loop control scheme.

[0010] The second aspect of this invention provides an intelligent monitoring system for preventing leakage in electronic atomizers. This system includes the following modules: a sensor deployment module, used to deploy capacitive, pressure, and temperature / humidity sensors in the liquid storage tank, atomizing core, and airflow channel, respectively; extracting sensor baseline parameters and setting thresholds through a calibration procedure during initial startup to obtain an initial dataset; a visual inspection module, used to acquire images of the airflow channel using a miniature camera and infrared light source; extracting droplet contours using an edge detection algorithm; eliminating noise through morphological processing; and using the droplet coverage area ratio and dynamic change rate as key indicators to obtain condensate risk characteristics; and a risk classification module, used to normalize the sensor data and visual inspection results. The system employs a fuzzy logic module to integrate capacitance change rate, pressure fluctuation, temperature and humidity difference, and droplet area. It assesses risk levels using a pre-defined rule base, resulting in high, medium, and low-level leak warnings. An anomaly detection module collects leak event feedback data via a user app, labels real leak scenarios, updates the training set, analyzes historical data using clustering algorithms, extracts high-frequency leak patterns, and adjusts monitoring parameters to obtain an optimized dynamic monitoring module. A closed-loop control module converts risk levels into specific warning signals, notifying users via equipment vibration, LED lights, and app push notifications. In high-risk situations, it automatically cuts off power, locks the heating function, and generates maintenance suggestions, resulting in a complete leak-proof closed-loop control solution.

[0011] A third aspect of the present invention provides an intelligent monitoring device for preventing leakage of electronic atomizers. The intelligent monitoring device for preventing leakage of electronic atomizers includes a memory and at least one processor. The memory stores instructions. The at least one processor invokes the instructions in the memory to cause the intelligent monitoring device for preventing leakage of electronic atomizers to perform the steps of the intelligent monitoring method for preventing leakage of electronic atomizers as described in any of the preceding claims.

[0012] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions, characterized in that, when executed by a processor, the instructions implement the steps of the intelligent monitoring method for preventing leakage of an electronic atomizer as described in any of the preceding claims.

[0013] In the technical solution provided by this invention, capacitive, pressure, and temperature / humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. During initial startup, a calibration procedure is performed to extract sensor baseline parameters and set thresholds, resulting in an initial dataset. Images of the airflow channel are acquired using a miniature camera and infrared light source. An edge detection algorithm is used to extract droplet contours, and morphological processing is combined to eliminate noise. The droplet coverage area ratio and dynamic change rate are used as key indicators to obtain condensate risk characteristics. Sensor data and visual detection results are normalized, and a fuzzy logic module integrates capacitance change rate, pressure fluctuation, temperature / humidity difference, and droplet area. A preset rule base is used to assess the risk level, resulting in high, medium, and low-level leak warnings. Leakage event feedback data is collected via a user app, real leak scenarios are labeled, the training set is updated, and a clustering algorithm is used to analyze historical data, extracting high-frequency leak patterns and adjusting monitoring parameters to obtain an optimized dynamic monitoring module. The risk level is converted into specific warning signals, which are then communicated to the user via equipment vibration, LED lights, and app push notifications. In cases of high risk, the power is automatically cut off, the heating function is locked, and maintenance suggestions are generated, resulting in a complete leak-proof closed-loop control scheme. This invention solves the problems of traditional atomizing equipment having a single method for leak detection, difficulty in obtaining comprehensive and accurate information on the internal liquid status, inability to classify and warn of leak risks, lack of data feedback and optimization mechanisms, and inability to dynamically adjust monitoring parameters to achieve effective closed-loop control for leak prevention. Attached Figure Description

[0014] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0015] Figure 1 This is a schematic diagram of the first embodiment of an intelligent monitoring method for preventing leakage of electronic atomizers according to an embodiment of the present invention.

[0016] Figure 2 This is a schematic diagram of a second embodiment of an intelligent monitoring method for preventing leakage of electronic atomizers according to an embodiment of the present invention.

[0017] Figure 3 This is a schematic diagram of the third embodiment of an intelligent monitoring method for preventing leakage of electronic atomizers according to an embodiment of the present invention.

[0018] Figure 4This is a schematic diagram of the fourth embodiment of an intelligent monitoring method for preventing leakage of electronic atomizers according to the present invention.

[0019] Figure 5 This is a schematic diagram of the fifth embodiment of an intelligent monitoring method for preventing leakage of electronic atomizers according to the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0022] A smart monitoring method for preventing leakage in electronic atomizers, such as Figure 1 As shown, the process includes the following steps: Capacitive, pressure, and temperature / humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. During initial startup, a calibration procedure is performed to extract sensor baseline parameters and set thresholds to obtain an initial dataset. Images of the airflow channel are acquired using a miniature camera and infrared light source. An edge detection algorithm is used to extract droplet contours, and morphological processing is combined to eliminate noise. The droplet coverage area ratio and dynamic change rate are used as key indicators to obtain condensate risk characteristics. Sensor data and visual detection results are normalized. A fuzzy logic module integrates capacitance change rate, pressure fluctuation, temperature / humidity difference, and droplet area. A preset rule base is used to assess the risk level, resulting in high, medium, and low-level leak warnings. Leakage event feedback data is collected via a user app. Real leak scenarios are labeled, the training set is updated, and a clustering algorithm is used to analyze historical data, extracting high-frequency leak patterns. Monitoring parameters are adjusted to obtain an optimized dynamic monitoring module. The risk level is converted into specific warning signals, which are then communicated to users via equipment vibration, LED lights, and app push notifications. In cases of high risk, the power is automatically cut off, the heating function is locked, and maintenance suggestions are generated, resulting in a complete leak-proof closed-loop control scheme.

[0023] like Figure 2As shown, in this embodiment, a capacitive sensor is deployed at the bottom of the liquid storage tank to monitor the change in the dielectric constant of the liquid in real time to reflect the liquid level. A miniature pressure sensor is installed at the sealing interface of the atomizing core to capture abnormal pressure fluctuations caused by seal failure. A temperature and humidity sensor is integrated in the airflow channel to continuously track the ambient temperature and humidity gradient and predict the risk of condensation formation. Through the automated calibration process executed when the device is first started, the baseline parameters of the sensor under empty tank, standard pressure, and normal temperature dry environment are extracted respectively. After comparing the collected values ​​with the preset safety range, a dynamic threshold is set to obtain the initial dataset.

[0024] Capacitive sensors deployed at the bottom of the liquid storage tank accurately reflect the liquid level in real time, preventing leaks caused by abnormal levels. Miniature pressure sensors installed at the atomizer core's sealing interface promptly detect pressure fluctuations due to seal failure, providing early warning of potential leaks. Temperature and humidity sensors integrated into the airflow channel continuously track environmental changes, effectively predicting the risk of condensation formation. An automated calibration process upon initial startup extracts baseline parameters and sets dynamic thresholds, creating an initial dataset that provides a reliable basis for subsequent accurate monitoring and risk assessment, significantly improving the accuracy and timeliness of leak detection.

[0025] like Figure 3 As shown, in this embodiment, a miniature camera paired with a narrow-band infrared light source is used to directionally acquire images of the inner wall of the airflow channel at a frequency of eight frames per second, ensuring clear capture of the droplet adhesion state under low light conditions. The droplet contour boundary is accurately extracted using the Canny edge detection algorithm, and a morphological processing method combining opening and closing operations is used to eliminate small particle noise and uneven lighting interference in the image. The processed binarized image is then subjected to pixel statistics to calculate the proportion of the droplet coverage area to the channel cross-sectional area, and the area change rate of three consecutive frames is tracked and integrated into the core feature indicator of condensate accumulation risk.

[0026] A miniature camera paired with a narrow-band infrared light source captures images at a specific frequency, clearly revealing the adhesion of droplets to the inner wall of the airflow channel even in low light conditions, laying the foundation for accurate monitoring. The Canny edge detection algorithm accurately delineates the droplet outline, while morphological processing methods effectively eliminate noise and light interference, improving image quality. By performing pixel statistics on the processed images, calculating the droplet coverage area ratio and the rate of area change in consecutive frames, and integrating these into core feature indicators, the risk of condensate accumulation can be accurately reflected, providing reliable data for subsequent risk assessment and early warning, and enhancing the sensitivity and accuracy of leak detection.

[0027] like Figure 4As shown, in this embodiment, the min-max normalization method is used to map the liquid level change rate collected by the capacitive sensor, the fluctuation value monitored by the pressure sensor, the difference recorded by the temperature and humidity sensor, and the droplet coverage area ratio extracted by the vision module to a uniform range of zero to one. A multi-input single-output module is constructed through a fuzzy logic module, using the parameters collected by the sensor as input variables, and using a preset membership function to quantify the degree of high, medium, and low risk. Reasoning calculations are performed through a rule base formulated by expert experience, and after comprehensive evaluation, a risk level weight value is output. The weight value is then converted into a clear high, medium, and low three-level leakage warning result according to the threshold classification standard.

[0028] A min-max normalization method is employed to uniformly map data from different types of sensors and vision modules, eliminating differences in data dimensions, making the data comparable, and facilitating subsequent comprehensive processing. The fuzzy logic module constructs a multi-input single-output architecture, using preset membership functions to quantify the risk level of each parameter, enabling it to handle uncertain information more accurately in practice. Leveraging a rule base developed with expert experience, the system comprehensively evaluates and outputs risk level weights, transforming them into clear three-level early warning results. This accurately and intuitively presents leakage risk, providing a reliable basis for timely countermeasures and effectively improving the intelligence level of leakage prevention monitoring.

[0029] like Figure 5 As shown, in this embodiment, leakage event data reported during device use is actively collected through the feedback entry point built into the user's APP. The leakage event data includes the occurrence time, environmental conditions, and original sensor records. Professional personnel manually annotate the feedback data to clearly distinguish between real leakage and false alarm scenarios, and the annotated data is expanded into the existing training set. The DBSCAN density clustering algorithm is used to analyze historical leakage events and extract high-frequency spatiotemporal feature combination patterns. Based on the pattern analysis results, the monitoring parameters such as capacitance threshold and pressure fluctuation range are adjusted in a targeted manner to obtain an optimized dynamic monitoring module.

[0030] Leakage event data is proactively collected through user app feedback, covering multiple key information aspects and providing rich material for subsequent analysis. Professionally labeled data effectively distinguishes between real and false alarm scenarios, improving data quality and expanding the training set to enhance model generalization ability. The DBSCAN algorithm is used to analyze historical data, accurately extracting high-frequency spatiotemporal feature patterns for deeper understanding of leakage patterns. Based on this, monitoring parameters are adjusted to make the dynamic monitoring module more closely match actual usage, enabling more accurate and timely detection of leakage risks.

[0031] In this embodiment, a hierarchical coding rule is used to convert risk levels into differentiated warning signals. The warning signals include: low risk triggers a single short vibration and a flashing green LED; medium risk triggers continuous vibration and a constantly lit yellow LED; and high risk triggers three strong vibrations combined with a high-frequency flashing red LED. A real-time connection is established between the device and the user's APP via the device's Bluetooth module, and full information including risk type, location of occurrence, and suggested operations is pushed synchronously. When a high-risk state is detected, the power supply circuit is immediately cut off and the heating function is disabled. At the same time, based on historical data, typical fault modes are extracted to generate a targeted maintenance guide including cleaning and sealing replacement steps, resulting in a complete leak-proof closed-loop control scheme.

[0032] The tiered coding system translates risk levels into differentiated warning signals, with different vibration and light cues corresponding to different risks, allowing users to quickly and intuitively perceive the degree of risk. Real-time connection via Bluetooth module and app pushes comprehensive information, enabling users to fully understand the risk situation and corresponding actions. In high-risk situations, cutting off power and disabling the heating function effectively prevents leaks from causing more serious consequences. Simultaneously, targeted maintenance guidelines are generated based on historical data, providing users with clear solutions. This overall system forms a complete closed-loop control solution.

[0033] This invention also provides an intelligent monitoring system for preventing leakage in electronic atomizers, comprising the following modules: a sensor deployment module, used to deploy capacitive, pressure, and temperature / humidity sensors in the liquid storage tank, atomizing core, and airflow channel, respectively; through a calibration procedure during initial startup, extracting sensor baseline parameters and setting thresholds to obtain an initial dataset; a visual inspection module, used to acquire images of the airflow channel using a miniature camera and infrared light source, extracting droplet contours using an edge detection algorithm, combining morphological processing to eliminate noise, and using the droplet coverage area ratio and dynamic change rate as key indicators to obtain condensate risk characteristics; and a risk classification module, used to normalize and process the sensor data and visual inspection results. A fuzzy logic module integrates capacitance change rate, pressure fluctuation, temperature and humidity difference, and droplet area. It assesses risk levels using a pre-defined rule base, resulting in high, medium, and low-level leak warnings. An anomaly detection module collects leak event feedback data via a user app, labels real leak scenarios, updates the training set, analyzes historical data using clustering algorithms, extracts high-frequency leak patterns, and adjusts monitoring parameters to obtain an optimized dynamic monitoring module. A closed-loop control module converts risk levels into specific warning signals, notifying users via equipment vibration, LED lights, and app push notifications. In high-risk situations, it automatically cuts off power, locks the heating function, and generates maintenance suggestions, resulting in a complete leak-proof closed-loop control solution.

[0034] This invention also provides an intelligent monitoring device for preventing liquid leakage in electronic atomizers. This device may further include one or more power supplies, one or more wired or wireless network interfaces, one or more input / output interfaces, and / or one or more operating systems, such as Windows Server, MacOSX, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that the structure of the intelligent monitoring device for preventing liquid leakage in electronic atomizers does not constitute a limitation on the computer device provided by this invention, and may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements.

[0035] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform each step of the intelligent monitoring method for preventing leakage of electronic atomizers provided in the above embodiments.

[0036] 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 method for intelligent monitoring of leakage prevention in electronic atomizers, characterized in that, The intelligent monitoring method for preventing leakage in electronic atomizers includes the following steps: Capacitive, pressure, and temperature and humidity sensors are deployed in the liquid storage tank, atomizing core, and airflow channel, respectively. The sensor reference parameters are extracted and thresholds are set through the calibration procedure during the first start-up to obtain the initial dataset. Images of the airflow channel are acquired by a miniature camera and an infrared light source. The droplet contour is extracted by an edge detection algorithm and noise is eliminated by morphological processing. The droplet coverage area ratio and dynamic change rate are used as key indicators to obtain the risk characteristics of the condensate. The sensor data and visual detection results are normalized and processed. The fuzzy logic module is used to integrate the capacitance change rate, pressure fluctuation, temperature and humidity difference and droplet area. The risk level is evaluated through a preset rule base to obtain three levels of leakage warning results: high, medium and low. By collecting leakage event feedback data through user APP, labeling real leakage scenarios, updating the training set, using clustering algorithm to analyze historical data, extracting high-frequency leakage patterns, and adjusting monitoring parameters, an optimized dynamic monitoring module is obtained. The risk level is converted into specific warning signals, which are then sent to users via equipment vibration, LED lights, and APP push notifications. In case of high risk, the power is automatically cut off, the heating function is locked, and maintenance suggestions are generated, resulting in a complete closed-loop control solution to prevent liquid leakage.

2. The intelligent monitoring method for preventing leakage in an electronic atomizer according to claim 1, characterized in that, The system employs capacitive, pressure, and temperature / humidity sensors, which are respectively deployed in the liquid storage tank, atomizing core, and airflow channel. Through the initial startup calibration procedure, sensor baseline parameters are extracted and thresholds are set to obtain an initial dataset, including the following steps: A capacitive sensor is deployed at the bottom of the liquid storage tank to monitor changes in the liquid dielectric constant in real time to reflect the liquid level status. A miniature pressure sensor is installed at the sealing interface of the atomizing core to capture abnormal pressure fluctuations caused by seal failure. Temperature and humidity sensors are integrated into the airflow channel to continuously track the ambient temperature and humidity gradient, predict the risk of condensation formation, and extract the reference parameters of the sensors under empty chamber, standard pressure and normal temperature dry environment through the automated calibration process executed when the equipment is first started. After comparing the collected values ​​with the preset safety range, a dynamic threshold is set to obtain the initial dataset.

3. The intelligent monitoring method for preventing leakage in an electronic atomizer according to claim 1, characterized in that, The process involves acquiring airflow channel images using a miniature camera and infrared light source, extracting droplet contours using an edge detection algorithm, eliminating noise through morphological processing, and using the droplet coverage area ratio and dynamic change rate as key indicators to obtain condensate risk characteristics. This includes the following steps: A miniature camera paired with a narrow-band infrared light source is used to collect images of the inner wall of the airflow channel at a frequency of eight frames per second, ensuring clear capture of the droplet adhesion state under low light conditions; The droplet contour boundary is accurately extracted using the Canny edge detection algorithm, and a morphological processing method combining opening and closing operations is used to eliminate small particle noise and uneven lighting interference in the image. The processed binarized image is statistically analyzed to calculate the ratio of the droplet coverage area to the channel cross-sectional area. The area change rate of three consecutive frames is tracked and integrated into the core feature indicator of condensate accumulation risk.

4. The intelligent monitoring method for preventing leakage in an electronic atomizer according to claim 1, characterized in that, The process of normalizing sensor data and visual detection results, integrating capacitance change rate, pressure fluctuation, temperature and humidity difference, and droplet area using a fuzzy logic module, and assessing the risk level through a preset rule base to obtain high, medium, and low leakage warning results includes the following steps: The minimum-maximum normalization method is used to map the liquid level change rate collected by the capacitive sensor, the fluctuation value monitored by the pressure sensor, the difference recorded by the temperature and humidity sensor, and the droplet coverage area ratio extracted by the vision module to the zero-to-one range. A multi-input single-output module is constructed by using a fuzzy logic module. The parameters collected by the sensor are used as input variables, and a preset membership function is used to quantify the degree of high, medium, and low risk. The system uses a rule base developed by experts to perform reasoning and calculations, and outputs risk level weight values ​​after comprehensive evaluation. These weight values ​​are then converted into clear high, medium, and low-level leakage warning results according to threshold classification standards.

5. The intelligent monitoring method for preventing leakage in an electronic atomizer according to claim 1, characterized in that, The process of collecting leakage event feedback data through user APP, labeling real leakage scenarios, updating the training set, analyzing historical data using clustering algorithms, extracting high-frequency leakage patterns, and adjusting monitoring parameters to obtain an optimized dynamic monitoring module includes the following steps: The system proactively collects data on leakage events reported during device use through the feedback portal built into the user's app. The leakage event data includes the time of occurrence, environmental conditions, and original sensor records. By having professionals manually annotate the feedback data, we can clearly distinguish between real leakage scenarios and false alarm scenarios, and then expand the annotated data into the existing training set. The DBSCAN density clustering algorithm was used to analyze historical leakage events and extract high-frequency spatiotemporal feature combination patterns. Based on the pattern analysis results, the monitoring parameters such as capacitance threshold and pressure fluctuation range were adjusted to obtain an optimized dynamic monitoring module.

6. The intelligent monitoring method for preventing leakage in an electronic atomizer according to claim 1, characterized in that, The process involves converting risk levels into specific early warning signals, notifying users via equipment vibration, LED lights, and APP push notifications, automatically cutting off power and locking the heating function in high-risk situations, generating maintenance suggestions, and obtaining a complete leak-proof closed-loop control scheme, including the following steps: The risk level is converted into a differentiated warning signal using a hierarchical coding rule. The warning signal includes: low risk triggers a single short vibration and green LED flashing; medium risk triggers continuous vibration and yellow LED staying on; and high risk triggers three strong vibrations combined with high-frequency red LED flashing. The device establishes a real-time connection with the user's APP via Bluetooth module, and synchronously pushes full information including risk type, location of occurrence, and suggested actions; When a high-risk condition is detected, the power supply circuit is immediately cut off and the heating function is disabled. At the same time, typical fault modes are extracted based on historical data to generate a targeted maintenance guide that includes steps such as cleaning and replacing seals, resulting in a complete leak-proof closed-loop control scheme.

7. A smart monitoring system for preventing leakage in electronic atomizers, characterized in that, The electronic atomizer leak-proof intelligent monitoring system includes the following modules: The sensor deployment module is used to deploy capacitive, pressure, and temperature and humidity sensors in the liquid storage tank, atomizing core, and airflow channel, respectively. Through the calibration procedure during the first startup, the sensor reference parameters are extracted and thresholds are set to obtain the initial dataset. The visual inspection module is used to acquire images of the airflow channel through a miniature camera and an infrared light source, extract the droplet contour using an edge detection algorithm, and eliminate noise by combining morphological processing. The droplet coverage area ratio and dynamic change rate are used as key indicators to obtain the risk characteristics of the condensate. The risk classification module is used to normalize sensor data and visual detection results. It uses a fuzzy logic module to integrate capacitance change rate, pressure fluctuation, temperature and humidity difference and droplet area. It evaluates the risk level through a preset rule base and obtains three levels of leakage warning results: high, medium and low. The anomaly detection module is used to collect leakage event feedback data through the user's APP, label real leakage scenarios, update the training set, analyze historical data using clustering algorithms, extract high-frequency leakage patterns, adjust monitoring parameters, and obtain an optimized dynamic monitoring module. The closed-loop control module is used to convert risk levels into specific warning signals, which are then sent to users via equipment vibration, LED lights, and APP push notifications. In case of high risk, the power is automatically cut off, the heating function is locked, and maintenance suggestions are generated to obtain a complete leak-proof closed-loop control solution.

8. A smart monitoring device for preventing leakage in electronic atomizers, characterized in that, The electronic atomizer leak-proof intelligent monitoring device includes a memory and at least one processor. The memory stores instructions, and the at least one processor calls the instructions in the memory to cause the electronic atomizer leak-proof intelligent monitoring device to perform each step of the electronic atomizer leak-proof intelligent monitoring method as described in any one of claims 1-6.

9. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement each step of the intelligent monitoring method for preventing leakage of electronic atomizers as described in any one of claims 1-6.

Images