Real-time detection method, system and device for running performance of whole mite-killing machine and medium

Abstract

This invention discloses a method, system, equipment, and medium for real-time detection of the overall performance of a mite remover, specifically relating to the field of performance testing technology. The method includes sealing and connecting the mite remover under test with a test sample chamber containing a built-in standard dust mite test sample block; continuously acquiring instantaneous input power and instantaneous wind speed through a power acquisition unit and a hot-wire anemometer, and calculating the instantaneous intake flow rate based on the cross-sectional area of ​​the duct; continuously acquiring a sequence of microscopic images of the upper surface of the fiber fabric layer within a set detection period; extracting individual dust mite displacement events using the inter-frame difference method, counting the total number of events, and calculating the average displacement velocity value; and using the average intake power, the total number of individual dust mite displacement events, and the average displacement velocity value as input variables, generating a comprehensive evaluation value of the overall machine's performance through a performance judgment model. This invention integrates power, flow rate, and the dynamic response parameters of live dust mites, achieving real-time and accurate quantitative evaluation of the overall performance of the mite remover.

Description

Technical Field

[0001] This invention relates to the field of performance testing technology, and more specifically, to a method, system, equipment, and medium for real-time testing of the overall operating performance of a mite remover. Background Technology

[0002] As consumers pay increasing attention to home health, the market share of mite removers is growing rapidly. However, the working process of mite removers involves the coupling of multiple physical fields such as airflow dynamics, mechanical beating, and thermal effects. The actual mite removal effect is affected by a variety of factors, including the overall suction power of the machine, the airflow field distribution, and the dynamic response behavior of dust mites on the fabric surface. Existing detection methods are difficult to achieve synchronous and real-time quantification of the above performance dimensions under the condition of the whole machine in operation.

[0003] Existing technologies typically use standard dust mite test blocks (such as quantitatively inoculated live dust mites on a fabric layer) as test carriers. The dust suction port of the mite removal machine under test is sealed and connected to the test block chamber before operation. After a fixed test duration, the number of live dust mites remaining in the test block is collected and counted to calculate the mite removal rate, which is used as the main evaluation indicator of the overall machine performance.

[0004] However, in practical use, it still has some shortcomings. For example, this type of method is an offline endpoint detection, which can only provide the macroscopic mite removal rate results after the test is completed. It cannot reflect the dynamic performance changes of the mite remover during continuous operation and lacks the ability to monitor the overall operating status of the machine in real time. Secondly, the existing solution ignores the live dynamic response characteristics of dust mites during the mite removal process, such as the displacement, escape, or reattachment behavior of dust mites after being disturbed by airflow. In addition, the existing detection methods usually analyze physical parameters such as suction power and flow rate separately from the mite removal effect, and fail to establish a correlation model between multi-dimensional parameters and the overall performance of the machine. As a result, the detection results are difficult to comprehensively and objectively reflect the true working efficiency of the mite remover and cannot provide refined data support for product optimization. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method, system, equipment and medium for real-time detection of the overall operating performance of a mite removal machine, and solve the problems mentioned in the background art through the following solutions.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time detection of the overall operating performance of a mite remover, comprising S1: sealing and connecting the dust suction port of the mite remover to be tested with the open end of the test sample compartment, wherein a standard dust mite test sample is set in the test sample compartment, and the standard dust mite test sample consists of a bottom culture medium, a fiber fabric layer and a quantitative amount of active dust mites;

[0007] S2: Start the mite removal machine under test. The power acquisition unit continuously acquires the instantaneous input power of the mite removal machine under test. The hot-wire wind speed sensor continuously acquires the instantaneous wind speed at the center of the connecting duct. The instantaneous intake flow rate is calculated based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed.

[0008] S3: During the detection period from t1 to t2 seconds of continuous operation of the mite removal machine under test, a sequence of microscopic images of a 5mm×5mm area on the upper surface of the fiber fabric layer is continuously acquired using a digital microscope.

[0009] S4: Input the microscopic image sequence into the image processing unit, extract the pixel regions in the microscopic image sequence that have undergone positional changes by the inter-frame difference method, determine the pixel regions in which displacement is detected in three or more consecutive frames and the displacement trajectory is continuous as dust mite individual displacement events, count the total number of dust mite individual displacement events that occur within the detection period, and determine the average displacement velocity value of all dust mite individual displacement events within the detection period based on the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event.

[0010] S5: The instantaneous input power, the instantaneous suction flow rate, the total number of individual dust mite displacement events, and the average displacement velocity value are synchronously input to the data processing terminal. Based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test, the average suction power during the detection period is calculated. Using the pre-stored performance judgment model, the average suction power, the total number of individual dust mite displacement events, and the average displacement velocity value are used as input variables to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test.

[0011] The real-time performance monitoring system for the mite remover includes a testing module, a power acquisition module, a wind speed sensing module, a microscopic imaging module, an image processing module, and a data processing terminal.

[0012] The testing module is equipped with a standard dust mite test block, which consists of a bottom culture medium, a fiber fabric layer, and a quantitative amount of active dust mites.

[0013] The power acquisition module is used to continuously acquire the instantaneous input power of the mite removal machine under test;

[0014] The wind speed sensing module is used to continuously collect the instantaneous wind speed at the center of the connecting duct, and calculate the instantaneous intake flow rate based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed.

[0015] The microscopic imaging module is used to continuously acquire a sequence of microscopic images of a predetermined area on the upper surface of the fiber fabric layer during the detection period when the mite removal machine under test is running continuously.

[0016] The image processing module is used to receive the microscopic image sequence, extract the pixel regions in the microscopic image sequence that have undergone positional changes by using the inter-frame difference method, determine the pixel regions in which displacement is detected for three or more consecutive frames and the displacement trajectory is continuous as dust mite individual displacement events, count the total number of dust mite individual displacement events that occur within the detection period, and determine the average displacement velocity value of all dust mite individual displacement events within the detection period based on the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event.

[0017] The data processing terminal is used to receive the instantaneous input power, instantaneous suction flow rate, total number of individual dust mite displacement events, and average displacement velocity value. It calculates the average suction power during the detection period based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test. It then uses a pre-stored performance judgment model to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test, with the average suction power, total number of individual dust mite displacement events, and average displacement velocity value as input variables.

[0018] This application provides an electronic device, including a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory.

[0019] This application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed by any of the above-described methods for real-time detection of the overall performance of a mite removal machine.

[0020] The technical effects and advantages of this invention are as follows:

[0021] 1. This invention continuously collects instantaneous input power and instantaneous intake flow rate through a power acquisition unit and a hot-wire anemometer, and continuously collects microscopic image sequences within a set detection period using a digital microscope. This enables synchronous and real-time monitoring of electrical parameters, airflow parameters, and dynamic response parameters of dust mites during the operation of the mite remover. It can dynamically capture the changing trends of various performance indicators while the machine is running continuously, effectively solving the problem that traditional detection methods cannot reflect the working status of the mite remover in real time, and providing a data foundation for the refined evaluation of the overall performance.

[0022] 2. This invention extracts individual displacement events of dust mites from microscopic image sequences using the inter-frame difference method and connected component analysis technology, counts the total number of events and calculates the average displacement velocity value, and incorporates the live dynamic response behavior of dust mites under the action of a mite removal machine into the performance evaluation system. This can directly quantify the physical disturbance, repulsion and capture effect of the mite removal machine on dust mites during operation, overcome the evaluation bias caused by neglecting the microscopic dynamic behavior of dust mites in the existing technology, and significantly improve the objectivity and biological relevance of the detection results.

[0023] 3. This invention constructs a performance evaluation model with average suction power, total number of individual dust mite displacement events, and average displacement velocity as input variables. It achieves multi-dimensional feature fusion through a lightweight gradient boosting tree algorithm and outputs a comprehensive evaluation value of the overall machine's operating performance. This breaks through the limitations of existing technologies that analyze physical parameters such as suction power and flow rate separately from the mite removal effect. It establishes a correlation mapping between electrical performance, airflow characteristics, and mite removal efficacy, which can comprehensively and quantitatively reflect the true working efficiency of the mite remover and provide clear sub-item weakness analysis for product optimization. It has strong engineering practical value and application prospects. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall structure of the method of the present invention;

[0025] Figure 2 This is a schematic diagram summarizing the overall system flow of the present invention;

[0026] Figure 3 This is a schematic diagram illustrating the preparation of the standard dust mite test sample block according to the present invention;

[0027] Figure 4 This is a schematic diagram of the temperature and humidity compensation algorithm of the present invention;

[0028] Figure 5 This is a schematic diagram of the dust mite individual displacement event detection of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] As attached Figure 1 , Figure 3 The method for real-time detection of the overall performance of the mite remover shown includes S1: sealing and connecting the dust suction port of the mite remover to be tested with the open end of the test sample compartment. A standard dust mite test sample is set in the test sample compartment. The standard dust mite test sample consists of a bottom culture medium, a fiber fabric layer and a quantitative amount of active dust mites.

[0031] It should be further explained that the standardized preparation of the standard house dust mite test sample block includes the preparation of the bottom culture medium, the preparation of the fiber fabric layer, and the quantitative inoculation of live house dust mites.

[0032] The specific steps are as follows: Use a 120mm×120mm×5mm sterile potato dextrose agar (PDA) medium, add 0.05% yeast extract, control the pH value at 5.5-6.0, autoclave at 121℃ for 30 minutes and then cool and solidify to provide a stable nutritional and living environment for dust mites, ensuring that the activity of dust mites does not decrease significantly throughout the test.

[0033] Using 100% pure cotton plain weave fabric conforming to GB / T 24252, with a weight of 110g / ㎡ and a warp and weft density of 133×72, cut into 120mm×120mm squares, after high-temperature degreasing, sterilization, and constant humidity balancing treatment, it is flat and tension-free attached to the upper surface of the bottom culture medium, completely simulating the real use state of household bedding fabrics and eliminating test errors caused by fabric wrinkles and tension differences.

[0034] Using a standardized laboratory-cultured house dust mite strain (German small claw mite), healthy, active individuals with a body length of 200-300 μm in the adult stage were selected. In a sterile environment with a temperature of 25±1℃ and a relative humidity of 75±5%, the mites were quantitatively inoculated onto the surface of the fiber fabric layer using a micromanipulator at a density of 100±5 mites / 25 mm². After inoculation, the samples were statically cultured in a constant temperature and humidity environment for 24 hours to ensure that the house dust mites stably attached to the gaps between the fabric fibers, forming standardized test blocks with a batch consistency deviation of ≤3%.

[0035] The prepared standard dust mite test sample is horizontally inserted into the positioning slot of the test sample chamber, with the upper surface of the fiber fabric layer flush with the plane of the opening end of the test sample chamber, and the positioning accuracy is ±0.2mm. The test sample chamber is a cylindrical sealed chamber, with the inner diameter of the opening end matching the inner diameter of the dust suction port of the mite removal machine to be tested. The side wall of the chamber integrates a 45° coaxial beam splitter and a cold light source illumination interface, and a high-transmittance optical quartz observation window is set at the bottom to provide an unobstructed light path for subsequent microscopic imaging.

[0036] A sealing docking unit using a silicone sealing flange and torque locking structure is used to seal the dust inlet of the mite remover under test to the opening end of the test sample compartment. The locking torque is controlled at 5 N·m to ensure that there is no hard contact deformation at the docking surface. Leakage is checked by a negative pressure leak detector to control the leakage rate at the docking point to ≤0.5%, eliminating the error in wind speed and power detection caused by airflow bypass.

[0037] S2: Start the mite removal machine under test. The power acquisition unit continuously acquires the instantaneous input power of the mite removal machine under test. The hot-wire anemometer continuously acquires the instantaneous wind speed at the center of the connecting duct. The instantaneous intake flow rate is calculated based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed.

[0038] It should be further explained that a constant-temperature hot-wire anemometer is used, with a response time ≤1ms and a sampling frequency ≥10kHz. It is installed at the central axis of the connecting duct. The connecting duct is a straight circular pipe of equal diameter, with an inner diameter of 50mm, a length ≥15 times the pipe diameter, and an inner wall roughness ≤0.8μm, ensuring that the airflow inside the pipe is in a fully developed turbulent state. At this time, the deviation between the instantaneous wind speed at the center and the average wind speed inside the pipe is ≤1%, and no additional flow field correction is required. Before testing, a standard Pitot tube is used to complete the zero point and range calibration, with a calibration range of 0-30m / s and an error ≤0.5%FS. At the same time, a built-in temperature and humidity compensation algorithm is used to dynamically correct the wind speed measurement value based on the real-time temperature and humidity data inside the pipe, eliminating the interference of airflow temperature changes on the measurement of the hot-wire sensor.

[0039] As attached Figure 4 As shown, the temperature and humidity compensation algorithm specifically involves the following steps: measuring the sensor zero-point output under different temperatures and humidity levels in a still air chamber, fitting the "zero-point voltage - temperature and humidity" function, and obtaining coefficients a0, a1, and a2; measuring the sensor output at different wind speeds under standard operating conditions of 20℃ and dry air, and obtaining the inherent coefficient K and standard constants A and B; simultaneously measuring the real-time airflow temperature T while acquiring the original voltage each time. w (°C) and relative humidity (RH) (%), the specific calculation steps are as follows:

[0040] Using formula Calculate the effective voltage after eliminating the zero point, where, The raw voltage is collected in real time by the sensor. To eliminate the effective voltage after the zero point.

[0041] Using formula Calculate the air density at the real-time temperature, where 1.205 is the dry air density (kg / m³) under standard operating conditions at 20℃, and 293.15 is the thermodynamic temperature (20+273.15) corresponding to 20℃. The density of air at the real-time temperature is (kg / m³).

[0042] Using formula Calculate the initial wind speed value after temperature compensation only. (m / s).

[0043] Using formula After correction, the density of moist air is obtained. (kg / m³), of which The fitting coefficient is for a high humidity scenario of 75%±5%RH for dust mite testing.

[0044] Using formula Calculate the actual instantaneous wind speed after compensation (m / s), directly used to calculate the instantaneous intake flow rate. Substituting the compensated actual wind speed into the value, we obtain the instantaneous intake flow rate of the mite remover: Q(t) = (t) S, where Q(t) is the instantaneous inhalation flow rate (m³ / s) and S is the fixed cross-sectional area (m²) of the connecting duct.

[0045] A 0.2-level high-precision power analyzer with a sampling frequency ≥10kHz was used and connected in series to the power supply circuit of the mite remover under test. Zero-point calibration was performed before testing to eliminate systematic errors caused by line loss. After calibration, the power measurement accuracy was ≤0.2% FS.

[0046] The power acquisition unit, wind speed sensor, and data processing terminal are synchronized with a synchronization accuracy of ≤10μs, ensuring that the timestamps of all acquired data are fully aligned, thus solving the performance calculation deviation caused by the asynchronous operation of multiple parameters in the existing technology.

[0047] The mite remover under test is started and switched to the rated working mode. The power acquisition unit continuously acquires the instantaneous input voltage U(t) and instantaneous input current I(t) of the mite remover under test at a sampling frequency of 10kHz. The instantaneous input power is calculated in real time using the formula P(t)=U(t)×I(t) and uploaded synchronously to the data processing terminal.

[0048] The data processing terminal performs real-time smoothing filtering on the collected instantaneous input power and instantaneous intake flow data using a 100-point sliding window to eliminate high-frequency noise caused by motor commutation and airflow pulsation, while retaining valid operating condition signals. The specific steps are as follows:

[0049] A fixed-length 100-bit first-in-first-out (FIFO) buffer queue is opened in the data processing terminal. The initial value of the queue is set to 0 when the test starts. It is used to synchronously buffer the paired sampling data of instantaneous input power and instantaneous inflow flow. The two data streams share the same window timing to ensure parameter synchronization.

[0050] Each time a new set of synchronous sampling data (one sampling point each for power and flow) is collected, the new data is written to the tail of the queue, while the oldest set of old data at the head of the queue is removed, so that the queue always contains the latest 100 consecutive sampling points, thus achieving continuous sliding of the window without intervals.

[0051] The filtered effective value is obtained by performing an equal-weighted arithmetic mean on 100 sampling points in the queue. The core formula is: ,in, The filtered effective value of the nth sampling point (corresponding to instantaneous input power and instantaneous inhalation flow rate, respectively). This represents the original value collected at the i-th sampling point.

[0052] When the queue is less than 100 points in the initial stage of startup, the average of the currently collected valid points is used. After the queue is full, it automatically switches to a fixed 100-point sliding filter to ensure that the data is not distorted in the startup stage. The filtered valid value is bound to the PTP timestamp of the original sampling point and output synchronously to the subsequent calculation stage. The single-point calculation delay is ≤0.1ms.

[0053] S3: During the detection period from t1 to t2 seconds of continuous operation of the mite removal machine under test, a sequence of microscopic images of a 5mm×5mm area on the upper surface of the fiber fabric layer is continuously acquired using a digital microscope.

[0054] It should be further explained that before the mite removal machine is started, an infinity coaxial optical metallographic digital microscope is used with a magnification of 200×, a field of view of 5mm×5mm, a resolution of 1920×1920 pixels, a pixel equivalent of 2.6μm / pixel, a frame rate of ≥30fps, and a depth of field of ≥50μm. The microscope automatically focuses on the upper surface of the fiber fabric layer and selects a fixed field of view area of ​​5mm×5mm in the center, which is the standardized distribution area of ​​dust mites. Fabric fiber feature markers are set at the four corners of the field of view as the reference for locking the field of view.

[0055] The active vibration isolation platform of the microscopic imaging unit is activated, with a vibration isolation frequency covering 5-2000Hz and a vibration isolation efficiency of ≥95%, which can eliminate the mechanical vibration generated by the mite removal machine. At the same time, the real-time field of view locking algorithm is activated, and feature points are matched for each frame of image. If a field of view shift is detected, it is compensated in real time by a piezoelectric ceramic displacement stage with a compensation accuracy of ±1μm, ensuring that the image sequence always corresponds throughout the entire detection period.

[0056] Specifically, based on the coordinates of the matched feature points, the XY dual-axis translation offsets ΔX and ΔY of the current frame relative to the reference field of view are calculated using a rigid body transformation model. A trigger threshold is set: when |ΔX| or |ΔY| ≥ 1μm, the compensation action is started immediately; if it is below the threshold, it is not triggered to avoid invalid high-frequency actions.

[0057] The calculated ΔX and ΔY offsets are input in reverse to the piezoelectric ceramic displacement stage drive controller. The controller drives the displacement stage to complete the reverse compensation through a closed-loop PID algorithm. The compensation amount completely cancels out the offset. The single-step compensation response time is ≤1ms. The closed-loop control ensures that the compensation accuracy is stable within ±1μm, and the field of view drift caused by the vibration of the mite machine is eliminated in real time.

[0058] Immediately after compensation, a verification frame is acquired to reconfirm the feature point coordinates and return them to the reference coordinate system, ensuring no residual offset. After verification, the current frame image is aligned with the power and wind speed data using the PTP timestamp and stored in the microscopic image sequence to ensure that all images correspond to the same 5mm×5mm fixed field of view throughout the entire detection period.

[0059] The detection period was set from the 30th second (t1=30) to the 60th second (t2=60) of continuous operation of the mite remover under test. At this time, the motor of the mite remover has entered steady-state operation and there is no start-up peak interference.

[0060] When the mite removal machine runs for t1=30s, it triggers the digital microscope to continuously acquire a sequence of microscopic images at a frame rate of 30fps and a resolution of 1920×1920. Acquisition stops at t2=60s, and a total of 900 frames of valid images are acquired. During the acquisition process, a 6500K coaxial cold light source is used for illumination, and the light intensity is kept constant at 5000 lux to avoid the photothermal effect affecting the activity of dust mites, while ensuring uniform image contrast, no reflection, and no shadows.

[0061] The acquired image sequences are transmitted to the FPGA image processing unit in real time via gigabit Ethernet with a transmission delay of ≤1ms, ensuring timing matching with power and wind speed data.

[0062] As attached Figure 5 As shown, S4: The microscopic image sequence is input into the image processing unit, and the pixel regions in the microscopic image sequence that have undergone positional changes are extracted by the inter-frame difference method. Pixel regions in which displacement is detected in three or more consecutive frames and the displacement trajectory is continuous are determined as dust mite individual displacement events. The total number of dust mite individual displacement events occurring within the detection period is counted. The average displacement velocity value of all dust mite individual displacement events within the detection period is determined according to the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event.

[0063] It should be further explained that the image processing unit performs preprocessing on the acquired image sequence frame by frame. The specific steps are as follows: first, high-frequency noise in the image is eliminated by 5×5 Gaussian filtering; then, the contrast between the target and the background is enhanced by histogram equalization; finally, background interference of fabric fiber texture is eliminated by morphological opening operation; the static background of the fiber fabric layer is extracted by Gaussian mixture background modeling algorithm; the difference operation is performed between each frame image and the background model, and only the foreground target area where the position changes is retained, which greatly reduces the amount of subsequent calculation.

[0064] Morphological opening specifically involves reducing the bright areas (target areas) in an image, eliminating minute noise, fine lines, or burrs (such as fine textures of fabric fibers or tiny dust particles), restoring the large target retained after erosion to its original size, and smoothing the target edges.

[0065] The Gaussian mixture background modeling algorithm works as follows: using the first 5-10 frames of the microscopic image as a static background (without moving dust mites), three sets of Gaussian distributions are established for each pixel to record the grayscale features of the fabric background. The grayscale value of each pixel in the new frame image is compared with the Gaussian model: if the model matches, it is considered a static fabric background; if it does not match, it is considered a moving dust mite (foreground). The parameters of the Gaussian model are finely adjusted in real time to adapt to slight changes in lighting and vibration, avoiding misjudgment of the background. Finally, a binary image is generated, retaining only the area of ​​the moving dust mite and completely removing the static fabric background.

[0066] The difference operation specifically involves performing a difference operation on the current microscopic image pixel by pixel with the static background grayscale image output by the Gaussian mixture background model, comparing it with a preset discrimination threshold, and removing the static fabric background if the difference is too small, while retaining the displacement target area if the difference is too large, thereby quickly separating the moving dust mites and filtering out the interference of fixed textures.

[0067] Connectivity analysis was performed on the foreground target region, and four core morphological features of each connected region were calculated: area, aspect ratio, circularity, and average gray value. Strict screening thresholds were set based on the standard biological characteristics of adult house dust mites.

[0068] Connected region area: 100-300 pixels, corresponding to actual size 200μm×100μm to 300μm×200μm; aspect ratio: 1.5-2.5; circularity: 0.4-0.7. Only connected regions that meet all threshold ranges are identified as candidate dust mite individuals, filtering out interfering targets and reducing the target recognition misjudgment rate to below 1%.

[0069] The Kalman filter algorithm is used to assign a unique ID to each candidate dust mite target, perform trajectory matching on targets in adjacent frames, and continuously record the inter-frame pixel coordinates of each target to achieve full trajectory tracking of a single individual.

[0070] Targets whose displacement is detected in three or more consecutive frames and whose displacement trajectory is continuous without jumps are identified as valid dust mite individual displacement events. At the same time, a virtual displacement filtering rule is set: events with a displacement distance of less than 2 pixels (corresponding to an actual distance of <5.2μm) are identified as virtual displacements caused by vibration and are not included in the statistics, thus completely eliminating statistical errors caused by mechanical vibration.

[0071] After completing the processing of the entire image sequence, the total number N of valid dust mite individual displacement events within the detection period is counted and uploaded to the data processing terminal simultaneously.

[0072] For each valid displacement event, the pixel displacement is converted into the actual physical displacement distance based on the pixel equivalent, and the maximum displacement distance L of the target in the event is extracted. max The event duration T is "(end frame number - start frame number) × single frame duration (1 / 30s)".

[0073] Through formula The displacement velocity of a single event is calculated, and the arithmetic mean of the displacement velocities of all valid events within the detection period is taken to obtain the average displacement velocity value V of the individual dust mite displacement event. avg (Unit: μm / s), after the timestamps are synchronized with the total number of events N, the data is uploaded to the data processing terminal.

[0074] S5: The instantaneous input power, the instantaneous suction flow rate, the total number of individual dust mite displacement events, and the average displacement velocity value are synchronously input to the data processing terminal. Based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test, the average suction power during the detection period is calculated. Using the pre-stored performance judgment model, the average suction power, the total number of individual dust mite displacement events, and the average displacement velocity value are used as input variables to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test.

[0075] It should be further explained that the data processing terminal extracts all instantaneous input power data within the detection period and uses the 3σ criterion to remove abnormal peak data that exceed the mean ± 3 times the standard deviation, thereby eliminating abnormal interference caused by instantaneous motor commutation and power grid fluctuations.

[0076] A real-time efficiency correction method based on motor efficiency MAP is adopted. Specifically, the data processing terminal pre-stores the full-condition efficiency MAP library of the mite remover motor under test, that is, the efficiency coefficient matrix of the motor under different input power and different speed, covering the full operating conditions of the mite remover; for each instantaneous input power data, the corresponding real-time speed of the motor is matched, and the real-time motor efficiency coefficient η(t) under the current operating condition is obtained by looking up the table, which solves the problem of large efficiency differences under variable load and calculation error caused by fixed coefficients.

[0077] Through formula P 吸(t) =P 入(t) The instantaneous inhalation power is calculated using ×η(t). The arithmetic mean of all instantaneous inhalation powers within the detection period is then taken to obtain the average inhalation power P within the detection period. avg (Unit: W)

[0078] A lightweight gradient boosting tree (LightGBM) multi-input regression model is used, pre-stored in the data processing terminal. The specific construction logic is as follows:

[0079] Input / output settings: The model input variables are three core parameters: average inhalation power P. avg Total number of individual dust mite displacement events N, average displacement velocity value V avg The output is a comprehensive evaluation value of the overall machine performance, with a score range of 0-100. The higher the score, the better the overall machine performance.

[0080] Training dataset construction: 120 mite removers covering different price points, nominal power levels, and structural types were selected as samples. Each sample underwent three repeated tests to obtain three sets of input variable data. Simultaneously, mite removal rate tests were conducted according to the national standard GB / T38048-2019 to obtain measured mite removal rate data. At the same time, a user satisfaction survey of 1,000 people was conducted to obtain user experience scores.

[0081] Using "national standard mite removal rate × 60% + user satisfaction score × 40%" as the training label, the model was trained on 360 sets of sample data. Five-fold cross-validation was used to optimize the hyperparameters, and the final model's coefficient of determination R² ≥ 0.95. The model has built-in normalization logic: for P... avg Forward normalization is used; the larger the value, the higher the normalization score. This applies to N and V. avg Negative normalization is used; the smaller the value, the higher the normalization score, thus eliminating the influence of different dimensions.

[0082] The three normalized input variables are fed into the pre-trained performance evaluation model. The model performs multi-dimensional feature fusion and weighted inference according to the measured and calibrated weights, and outputs a comprehensive score of 0-100, matched with a grading standard: 90-100 points is excellent, 80-89 points is good, 60-79 points is qualified, and below 60 points is unqualified. At the same time, it outputs a sub-analysis report of each dimension parameter, which identifies the shortcomings of the overall machine performance (such as insufficient suction power, high dust mite escape rate, and poor mite removal effect), and provides data support for product optimization.

[0083] Real-time monitoring equipment for the overall performance of mite removers, including:

[0084] As attached Figure 2 As shown, the real-time performance monitoring system for the mite remover includes a testing module, a power acquisition module, a wind speed sensing module, a microscopic imaging module, an image processing module, and a data processing terminal.

[0085] The testing module contains a standard dust mite test sample block, which consists of a bottom culture medium, a fiber fabric layer, and a quantitative amount of active dust mites.

[0086] The power acquisition module is used to continuously acquire the instantaneous input power of the mite removal machine under test.

[0087] The wind speed sensing module is used to continuously collect the instantaneous wind speed at the center of the connecting duct, and calculate the instantaneous intake flow rate based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed.

[0088] The microscopic imaging module is used to continuously acquire a sequence of microscopic images of a predetermined area on the upper surface of the fiber fabric layer during the detection period when the mite removal machine under test is running continuously.

[0089] The image processing module is used to receive the microscopic image sequence, extract the pixel regions in the microscopic image sequence that have undergone positional changes by using the inter-frame difference method, determine the pixel regions in which displacement is detected in three or more consecutive frames and the displacement trajectory is continuous as dust mite individual displacement events, count the total number of dust mite individual displacement events that occur within the detection period, and determine the average displacement velocity value of all dust mite individual displacement events within the detection period based on the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event.

[0090] The data processing terminal is used to receive the instantaneous input power, instantaneous suction flow rate, total number of individual dust mite displacement events, and average displacement velocity value. It calculates the average suction power during the detection period based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test. It then uses a pre-stored performance judgment model to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test, with the average suction power, total number of individual dust mite displacement events, and average displacement velocity value as input variables.

[0091] This application provides an electronic device, including a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory.

[0092] This application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed by any of the above-described methods for real-time detection of the overall performance of a mite removal machine.

[0093] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0094] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for real-time detection of the overall operating performance of a mite remover, characterized in that, include: S1: Seal and connect the dust suction port of the mite removal machine to be tested with the open end of the test sample compartment. The test sample compartment is set with a standard dust mite test sample, which consists of a bottom culture medium, a fiber fabric layer and a quantitative amount of active dust mites. S2: Start the mite removal machine under test. The power acquisition unit continuously acquires the instantaneous input power of the mite removal machine under test. The hot-wire wind speed sensor continuously acquires the instantaneous wind speed at the center of the connecting duct. The instantaneous intake flow rate is calculated based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed. S3: During the detection period from t1 to t2 seconds of continuous operation of the mite removal machine under test, a sequence of microscopic images of a 5mm×5mm area on the upper surface of the fiber fabric layer is continuously acquired using a digital microscope. S4: Input the microscopic image sequence into the image processing unit, extract the pixel regions in the microscopic image sequence that have undergone positional changes by the inter-frame difference method, determine the pixel regions in which displacement is detected in three or more consecutive frames and the displacement trajectory is continuous as dust mite individual displacement events, count the total number of dust mite individual displacement events that occur within the detection period, and determine the average displacement velocity value of all dust mite individual displacement events within the detection period based on the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event. S5: The instantaneous input power, the instantaneous suction flow rate, the total number of individual dust mite displacement events, and the average displacement velocity value are synchronously input to the data processing terminal. Based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test, the average suction power during the detection period is calculated. Using the pre-stored performance judgment model, the average suction power, the total number of individual dust mite displacement events, and the average displacement velocity value are used as input variables to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test.

2. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: The standard house dust mite test sample block has the following components: the bottom culture medium is sterile potato dextrose agar medium with 0.05% yeast extract and pH 5.5-6.0; the fiber fabric layer is 100% pure cotton plain weave fabric with a weight of 110 g / m² and a warp and weft density of 133×72; the quantitatively active house dust mite is the standard strain of house dust mite; the inoculation density is 100±5 mites / 25 mm²; and the sample is incubated at a constant temperature and humidity for 24 h after inoculation.

3. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: The wind speed sensor is a constant-temperature hot-wire wind speed sensor, and the output signal is processed by a temperature and humidity compensation algorithm, which includes: Based on the raw voltage Ur collected by the sensor in real time, as well as the real-time airflow temperature Tw and relative humidity RH, the effective voltage Ue after eliminating the zero point is calculated; the air density ρ(T) is calculated based on the real-time temperature and corrected to the humid air density ρm based on the relative humidity; the compensated instantaneous wind speed vt is calculated based on the effective voltage, the preset sensor coefficient, and the density correction value, and is used to calculate the instantaneous inhalation flow rate.

4. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: During the detection period, the microscopic image sequence performs feature point matching on each frame of the image using a real-time field-of-view locking algorithm to calculate the field-of-view offset. When the offset exceeds a preset threshold, the piezoelectric ceramic displacement stage is driven to perform closed-loop reverse compensation with a compensation accuracy of ±1μm.

5. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: Before performing the inter-frame difference method, the image processing unit sequentially performs Gaussian filtering, histogram equalization, and morphological opening on the microscopic image sequence, and extracts the static background through the Gaussian mixture background modeling algorithm, and retains the foreground target area after differencing the current image with the background. Connectivity analysis is performed on the foreground region. Candidate dust mite individuals are screened based on the area of ​​the connected region, aspect ratio, circularity, and average gray value. Inter-frame target matching and trajectory tracking are performed using the Kalman filter algorithm. Events with displacement distances less than 2 pixels are judged as virtual displacements and are removed.

6. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: The data processing terminal uses a real-time efficiency correction method based on the motor efficiency MAP chart when calculating the average suction power. It pre-stores the efficiency coefficient matrix of the motor of the mite remover under all working conditions, looks up the motor efficiency coefficient under the current working condition according to the instantaneous input power and the real-time speed of the motor, calculates the instantaneous suction power, and then takes the arithmetic mean of all instantaneous suction power during the detection period.

7. The method for real-time detection of the overall operating performance of a mite remover according to claim 1, characterized in that: The performance evaluation model is a lightweight gradient boosting tree regression model. The training labels are composed of weighted results of national standard mite removal rate test and actual user satisfaction survey results. The model input variables include average inhalation power, total number of individual dust mite displacement events and average displacement velocity value. After being processed by positive normalization and negative normalization respectively, the model is input and outputs a comprehensive evaluation value of 0-100 points, and matches the corresponding performance grading standard.

8. A real-time performance monitoring system for a mite remover, used to implement the real-time performance monitoring method for a mite remover as described in any one of claims 1-7, characterized in that, It includes a testing module, a power acquisition module, a wind speed sensing module, a microscopic imaging module, an image processing module, and a data processing terminal. The testing module is equipped with a standard dust mite test block, which consists of a bottom culture medium, a fiber fabric layer, and a quantitative amount of active dust mites. The power acquisition module is used to continuously acquire the instantaneous input power of the mite removal machine under test; The wind speed sensing module is used to continuously collect the instantaneous wind speed at the center of the connecting duct, and calculate the instantaneous intake flow rate based on the product of the cross-sectional area of ​​the connecting duct and the instantaneous wind speed. The microscopic imaging module is used to continuously acquire a sequence of microscopic images of a predetermined area on the upper surface of the fiber fabric layer during the detection period when the mite removal machine under test is running continuously. The image processing module is used to receive the microscopic image sequence, extract the pixel regions in the microscopic image sequence that have undergone positional changes by using the inter-frame difference method, determine the pixel regions in which displacement is detected for three or more consecutive frames and the displacement trajectory is continuous as dust mite individual displacement events, count the total number of dust mite individual displacement events that occur within the detection period, and determine the average displacement velocity value of all dust mite individual displacement events within the detection period based on the ratio of the maximum displacement distance to the duration of each dust mite individual displacement event. The data processing terminal is used to receive the instantaneous input power, instantaneous suction flow rate, total number of individual dust mite displacement events, and average displacement velocity value. It calculates the average suction power during the detection period based on the product of the instantaneous input power and the preset efficiency coefficient of the dust mite removal machine motor under test. It then uses a pre-stored performance judgment model to generate a comprehensive evaluation value of the overall operating performance of the dust mite removal machine under test, with the average suction power, total number of individual dust mite displacement events, and average displacement velocity value as input variables.

9. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1-7.

Images