Particle counting monitoring device and method combining double-pump system and displacement sensing technology

Abstract

The invention relates to the technical field of particle counting monitoring, and discloses a particle counting monitoring device and method combining a double-pump system and a displacement sensing technology. The device comprises a double-pump module, a displacement sensing module, a particle counting module and a self-adaptive control module. The air suction pump and the air blowing pump of the double-pump module work independently and are started and stopped in a coordinated mode through the synchronous controller, and particles are stably collected and discharged. The displacement sensing module collects displacement data of the sampling head in real time. And the particle counting module calculates the particle distribution density through a regional density conversion algorithm according to the displacement and the sampling data. And the adaptive control module reads sampling head chip identification information, and determines an optimal sampling mode by using an AI algorithm model. The method is accurate in detection and high in adaptability, can effectively monitor particles on the surface of an object, can be widely applied to multiple industries such as electronics and pharmacy, and guarantees the product quality.

Description

Technical Field

[0001] The present invention relates to the technical field of particle counting monitoring, and specifically to a particle counting monitoring device and method combining a dual-pump system and displacement sensing technology. Background Art

[0002] In many industrial production and scientific research fields, the monitoring of particles on the surface of objects is crucial. Taking the electronic manufacturing industry as an example, if there are tiny particles on the surface of a chip, it may cause problems such as circuit short circuits and abnormal signal transmission, seriously affecting the chip performance and product quality; in the pharmaceutical industry, particle contamination on the surface of drug packaging or production equipment may lead to drug contamination and threaten the health of patients.

[0003] Traditional particle counting monitoring methods have many limitations. The early optical microscope detection method, although it can visually observe particles, has extremely low detection efficiency and is difficult to achieve automation and large-area detection. The method based on filter screen filtration and weighing can only obtain the total amount information of particles and cannot obtain the distribution of particles on the surface of the object. For a surface with uneven particle distribution, this detection result cannot accurately reflect the actual situation.

[0004] Some existing particle counting monitoring devices use a single-pump system, and it is easy to have problems with unstable air flow during the process of particle inhalation and discharge. The single pump is responsible for both inhalation and exhaust. During the switching process, the pressure and flow rate of the air flow will fluctuate, resulting in incomplete particle collection or residue inside the device, affecting the accuracy of the detection result. Moreover, the single-pump system is difficult to adapt to the diverse requirements for the particle inhalation and discharge speeds in different detection scenarios.

[0005] The application of displacement sensing technology in particle counting monitoring also has deficiencies. Some devices use displacement sensors with low precision, which cannot accurately measure the tiny displacement of the sampling head, bringing great difficulties to accurately determining the particle sampling position and calculating the particle distribution density per unit area. In addition, the existing displacement sensing modules often cannot work well in coordination with the particle counting module, and the data fusion effect is poor, unable to provide accurate and effective displacement data support for particle counting.

[0006] In terms of sampling mode selection, most existing monitoring devices lack adaptability. Different detection object surfaces have various forms, such as flat surfaces, curved surfaces, irregular surfaces, etc., and different sampling parameters are required to ensure the detection accuracy. However, traditional devices usually adopt a fixed sampling mode and cannot automatically adjust the sampling parameters according to the type of sampling head and the actual situation of the detection surface, resulting in inaccurate detection results and low efficiency.

[0007] With the continuous improvement of product quality requirements in various industries, there is an urgent need for a particle counting monitoring device and method that can overcome the above defects to achieve more efficient, accurate, and intelligent particle detection, meeting the needs of modern industrial production and scientific research. Summary of the Invention

[0008] The purpose of the present invention is to provide a particle counting monitoring device and method combining a dual-pump system with displacement sensing technology to solve the problems raised in the above background technology.

[0009] To achieve the above purpose, the present invention provides the following technical solution: A particle counting monitoring device combining a dual-pump system with displacement sensing technology, including a dual-pump module, a displacement sensing module, a particle counting module, and an adaptive control module; The dual-pump module includes an inhalation pump and a blowing pump. The inhalation pump and the blowing pump perform particle inhalation and discharge operations through independent channels respectively, and their start-stop timings are coordinated by a synchronous controller; The displacement sensing module is used to collect displacement data of the sampling head in real time, including the moving distance, speed, and direction; The particle counting module calculates the particle distribution density per unit area through a regional density conversion algorithm based on the displacement data and particle sampling data; The adaptive control module obtains identification information through a chip embedded in the sampling head and determines the optimal sampling mode of the sampling head based on the identification information through a preset AI algorithm model.

[0010] Preferably, the specific implementation method of the displacement sensing module is as follows: A laser displacement sensor or a capacitive displacement sensor is used to detect the relative position between the sampling head and the object surface in real time; During the sampling process, the displacement data is segmented according to the time series, the cumulative value of each segment of displacement is calculated and data fusion is performed with the particle sampling quantity to generate a particle distribution density map; If the fluctuation amplitude of a certain segment of displacement data exceeds the preset fluctuation threshold, it is marked as an abnormal displacement segment, and the adaptive control module is triggered to recalibrate the sampling parameters.

[0011] Preferably, the synchronous control process of the dual-pump module includes: When the inhalation pump starts, the blowing pump starts after a delay of T1 time to ensure stable inhalation airflow before performing particle discharge; During the operation of the blowing pump, its exhaust pressure value is monitored in real time. If the deviation between the exhaust pressure and the preset pressure value exceeds the threshold, a pump pressure abnormal signal is generated and fed back to the adaptive control module to adjust the power of the blowing pump.

[0012] Preferably, the particle counting implementation method of the particle counting module includes: Step A1: Receive real-time displacement data and particle sampling data from the displacement sensing module, and divide the displacement data into continuous time series segments based on a preset time interval; Step A2: According to the moving distance, speed, and direction of each displacement segment, convert the time series segment into a corresponding surface detection area through the area mapping algorithm, and calculate the area of each detection area; The implementation method of the area density conversion algorithm is as follows: Divide the displacement data into several sampling units, and each unit corresponds to a surface area; Count the number of particle samples in each unit, and calculate the local particle density in combination with the unit area; Perform weighted averaging on the local particle densities of all units to generate an overall particle distribution density value, and eliminate abnormal units with discrete values exceeding the preset range; Step A3: Count the number of particle samples in each detection area, and calculate the local particle density value in combination with the area of the corresponding area; Step A4: Perform spatial interpolation on the local particle density values of adjacent detection areas to generate a particle distribution density map covering the entire detection range; Step A5: Identify abnormal areas where the density difference between adjacent areas in the particle distribution density map exceeds the preset threshold, and perform smoothing correction on the particle density values of the abnormal areas based on the spatial continuity characteristics of historical sampling data; Step A6: Output the corrected particle distribution density map and the particle statistics value per unit area.

[0013] Preferably, the specific operation mode of the adaptive control module is as follows: When replacing the sampling head, read the identification code stored in the chip through radio frequency identification technology, including the sampling head type, applicable surface morphology, and sampling parameter range; Input the identification code into the AI algorithm model, match the corresponding suction pump power curve, displacement sensing sensitivity, and particle count correction coefficient, and automatically switch to the optimal sampling mode.

[0014] Preferably, the sampling head includes three types: flat type, curved surface type, and flexible type; The flat sampling head is made of rigid material and is applicable to flat surfaces. Its sampling parameters include a fixed suction rate and a linear displacement calibration coefficient; The curved surface sampling head is equipped with an adjustable bracket and is applicable to regular curved surfaces. Its sampling parameters include a dynamic suction rate and a curvature compensation coefficient; The flexible sampling head is made of silica gel material and is applicable to irregular surfaces. Its sampling parameters include an adaptive suction power and a displacement fluctuation tolerance range.

[0015] Preferably, the AI algorithm model adopts a neural network algorithm based on dynamic weight allocation, which specifically includes the following steps: Step B1: Encode the sampling head type in the identification code as a three-dimensional orthogonal vector , where correspond to the Boolean activation values of planar, curved, and flexible types respectively; Step B2: Construct a surface morphology feature matrix , where is the surface roughness grade scalar, is the curvature feature vector, , are type-related adaptive coefficients and satisfy ; Step B3: Perform parameter fusion through a dual-channel convolutional kernel, where: Power curve parameter ; Sensitivity coefficient ; Correction coefficient ; where represents three-dimensional tensor convolution, represents the Hadamard product, represents the activation function, represents the rectified linear unit function, represents the hyperbolic tangent function, is the temperature coefficient, respectively represent the weight matrices related to the calculation of the power curve parameter, the sensitivity coefficient, and the correction coefficient; represents the vector and the matrix longitudinal splicing, represents the vector and the matrix Kronecker product, represents the vector and the matrix horizontal splicing; represents the bias term; Step B4: Establish a dynamic loss function , where is the deviation value between the output parameter and the preset parameter library; optimize the weight matrix through backpropagation, so that the weighted sum of squares of the deviation between the output parameter and the preset parameter library is minimized, is the weight coefficient.

[0016] Preferably, it further includes an abnormal feedback module, and its operation mode is: During the particle counting process, if continuous N abnormal displacement segments are detected or the particle density value exceeds the preset reasonable range, a system abnormal signal is generated, where N is a preset positive integer constant; The system abnormal signal triggers the adaptive control module to pause the current sampling process and start a self-check program to detect the double-pump module, displacement sensing module, and particle counting module one by one.

[0017] Preferably, the specific steps of the self-check program include: Conduct a pressure pulse test on the suction pump and the blowing pump. If the pulse response time exceeds the preset threshold, it is determined that there is a pump body failure; Perform zero-drift calibration on the displacement sensing module. If the drift value after calibration still exceeds the tolerance range, it is determined that the sensor fails; Perform a standard particle sample test on the particle counting module. If the counting error exceeds the allowable value, reload the particle density conversion algorithm.

[0018] Preferably, a particle counting monitoring method for a particle counting monitoring device based on the combination of the above double-pump system and displacement sensing technology, the method includes the following steps: Utilize the suction pump and the blowing pump in the double-pump module to perform particle inhalation and discharge operations respectively through independent channels, and the start-stop timing of the two is coordinated by the synchronous controller; With the help of the displacement sensing module, real-time displacement data of the sampling head are collected, including the moving distance, speed, and direction, and the displacement data are transmitted to the particle counting module; The particle counting module calculates the particle distribution density per unit area by using the regional density conversion algorithm according to the received displacement data and particle sampling data; The adaptive control module obtains the identification information through the chip embedded in the sampling head and determines the optimal sampling mode of the sampling head based on the identification information by using the preset AI algorithm model.

[0019] Compared with the prior art, the beneficial effects of the present invention are: The particle counting monitoring device and method combining the dual - pump system and displacement sensing technology proposed by the present invention have significant beneficial effects in many aspects compared with traditional technologies. In terms of detection accuracy, the design of the dual - pump module greatly improves the stability of particle collection and discharge. The suction pump and the blowing pump work through independent channels, and the start - stop timing is coordinated by a synchronous controller, avoiding the problem of unstable air flow in a single - pump system. After the suction pump starts, the blowing pump starts after a delay of T1 time to ensure that the inhaled air flow is stable before discharging the particles, preventing particles from remaining or being re - inhaled in the device and ensuring that each collected particle can be accurately counted. When the blowing pump is operating, the exhaust pressure is monitored in real time. If the pressure deviation exceeds the threshold, the adaptive control module adjusts the power of the blowing pump in a timely manner to further ensure the smoothness and stability of particle discharge, making the particle counting result more reliable.

[0020] The displacement sensing module uses a laser displacement sensor or a capacitive displacement sensor, which can collect displacement data of the sampling head in real time and with high precision, including the moving distance, speed, and direction. During the sampling process, the displacement data is processed in segments according to the time series and data - fused with the particle sampling quantity. The generated particle distribution density map can intuitively and accurately reflect the distribution of particles on the object surface. Once the fluctuation amplitude of a certain segment of displacement data exceeds the preset threshold, it is marked as an abnormal displacement segment and the adaptive control module is triggered to recalibrate the sampling parameters, effectively avoiding particle counting errors caused by abnormal displacement and improving the detection accuracy.

[0021] The regional density conversion algorithm of the particle counting module is carefully designed, comprehensively considering the displacement data and the particle sampling data. By dividing the displacement data into sampling units, calculating the local particle density of each unit and performing weighted averaging, and excluding abnormal units, the calculated overall particle distribution density value is more accurate. When generating the particle distribution density map, spatial interpolation processing is performed on the local particle density values of adjacent detection regions, and the particle density values of abnormal regions are smoothed and corrected based on the spatial continuity characteristics of historical sampling data, further improving the accuracy and reliability of the map and providing more accurate particle distribution information for users.

[0022] The adaptive control module determines the optimal sampling mode through a preset AI algorithm model by means of the identification information of the chip embedded in the sampling head, demonstrating strong intelligent adaptability. When different types of sampling heads (flat, curved, and flexible) are replaced, it can automatically match the corresponding suction pump power curve, displacement sensing sensitivity, and particle count correction coefficient. The flat sampling head is suitable for flat surfaces, and its fixed suction rate and linear displacement calibration coefficient ensure the efficiency and accuracy of sampling on flat surfaces; the curved sampling head is equipped with an adjustable bracket, and its dynamic suction rate and curvature compensation coefficient enable it to optimize sampling according to the characteristics of regular curved surfaces; the flexible sampling head is made of silicone material, and its adaptive suction power and displacement fluctuation tolerance range enable it to well adapt to irregular surfaces. This adaptive ability greatly improves the adaptability of the device to different detection scenarios, and can ensure high-precision and high-efficiency detection regardless of how complex the detection surface morphology is.

[0023] In addition, the abnormal feedback module and the self-check program provide strong guarantees for the stable operation of the device. During the particle counting process, if continuous N abnormal displacement segments or particle density values exceeding the preset reasonable range are detected, a system abnormal signal is immediately generated, triggering the adaptive control module to pause sampling and start the self-check program. The double-pump module, displacement sensing module, and particle counting module are inspected item by item to promptly discover and solve potential problems, such as pump body failures, sensor failures, algorithm errors, etc., ensuring that the device is always in the best working state and reducing detection errors and production delays caused by equipment failures. Brief Description of the Drawings

[0024] Figure 1 is the working principle diagram of the particle counting monitoring device described in the present invention; Figure 2 is the flowchart of the particle counting module for counting and generating a spectrum; Figure 3 is the schematic diagram of the working characteristics and detection of different types of sampling heads; Figure 4 is the schematic diagram of the AI algorithm model for calculating and optimizing sampling parameters. Detailed Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] Please refer to Figures 1-4, the present invention provides a particle counting monitoring device and method combining a dual-pump system and displacement sensing technology. The particle counting monitoring device mainly consists of a dual-pump module, a displacement sensing module, a particle counting module, and an adaptive control module.

[0027] The dual-pump module is responsible for the inhalation and discharge operations of particles in the entire device. It includes an inhalation pump and a blowing pump. The inhalation pump and the blowing pump work through independent channels, and their start-stop timings are precisely coordinated by a synchronous controller. During actual operation, the inhalation pump starts first to inhale the gas containing particles into the device, and then the blowing pump starts to discharge the inhaled particles from the device, ensuring that the device can continuously and stably perform particle sampling operations.

[0028] The displacement sensing module undertakes the important task of real-time collecting displacement data of the sampling head. The data it collects includes moving distance, speed, and direction. During the sampling process, this module uses specific sensors to continuously monitor the relative position between the sampling head and the object surface. For example, when detecting particles on the surface of an industrial product, the displacement sensing module can continuously obtain various displacement information of the sampling head when it moves on the product surface, and transmit this data to the particle counting module, providing key basic data for subsequent particle counting and distribution density calculation.

[0029] The particle counting module receives the displacement data and particle sampling data from the displacement sensing module, and calculates the particle distribution density per unit area through a regional density conversion algorithm. It first processes the received data, determines the corresponding detection area based on the displacement data, and then combines the number of particle samples in each area. After a series of calculations, it obtains the particle distribution density per unit area, providing intuitive particle distribution information for users.

[0030] The adaptive control module obtains identification information through a chip embedded in the sampling head. When the sampling head is installed on the device, this module can read the information in the chip, and then, with the help of a preset AI algorithm model, determines the optimal sampling mode of the sampling head according to this identification information, so that the device can adapt to different types of sampling heads and detection scenarios, improving the accuracy and efficiency of detection.

[0031] The following further illustrates the present invention in conjunction with Embodiments 1 to 6: Embodiment 1: In this embodiment, the specific implementation method of the synchronous control of the displacement sensing module and the dual-pump module is further elaborated.

[0032] The displacement sensing module uses a laser displacement sensor to detect the relative position between the sampling head and the object surface in real time. The laser displacement sensor emits a laser beam and measures the time it takes for the laser beam to travel from emission to reflection back to the sensor. By utilizing the principle of the constant speed of light, it accurately calculates the distance between the sampling head and the object surface. When detecting particles on the surface of a planar object, the laser displacement sensor can capture the trajectory of the sampling head moving on the plane in real time, obtaining information about its moving distance, speed, and direction.

[0033] During the sampling process, the displacement data is processed in segments according to the time series. Assume the sampling time is 10 minutes, with each 1 minute as a segment. The displacement data for these 10 minutes is divided into 10 segments. Then, the cumulative value of the displacement for each segment is calculated. For example, within the first time segment, the sampling head moves 5 cm in the X direction and 3 cm in the Y direction. Using the Pythagorean theorem, the cumulative displacement value is approximately 5.83 cm. At the same time, the number of particle samples taken within each time segment is recorded, and the displacement cumulative value is fused with the number of particle samples. Using specialized data processing software, with the displacement cumulative value as the abscissa and the number of particle samples as the ordinate, a particle distribution density map is generated to visually display the particle distribution at different positions.

[0034] If the fluctuation amplitude of a certain segment of displacement data exceeds the preset fluctuation threshold, for example, the preset fluctuation threshold is 1 cm, and the displacement fluctuation of the sampling head reaches 1.5 cm within a certain period of time, it is then marked as an abnormal displacement segment. Once marked as an abnormal displacement segment, it will trigger the adaptive control module to recalibrate the sampling parameters. The adaptive control module will adjust parameters such as the moving speed and sampling frequency of the sampling head according to the previously preset algorithm to ensure the accuracy of subsequent sampling data.

[0035] For the synchronous control of the dual - pump module, when the suction pump starts, the blowing pump starts after a delay of T1 time. The T1 time is set based on actual experiments and experience. For example, after multiple tests, it is found that when T1 is set to 5 seconds, it can ensure that the inhalation airflow is stable before performing the particle discharge operation. During the operation of the blowing pump, the exhaust pressure value is monitored in real time through a pressure sensor. If the deviation between the exhaust pressure and the preset pressure value exceeds the threshold, for example, the preset pressure value is 50 kPa and the threshold is ±5 kPa, when the detected exhaust pressure is 56 kPa, a pump pressure abnormal signal is generated and fed back to the adaptive control module. After receiving the signal, the adaptive control module adjusts the power of the blowing pump according to the preset algorithm to ensure the stable operation of the blowing pump and ensure that the particles can be discharged from the device smoothly.

[0036] In practical application scenarios, such as in an electronic chip manufacturing workshop, when detecting particles on the surface of a chip, the above-mentioned synchronous control method of the displacement sensing module and the dual-pump module can ensure the stability and accuracy of the detection process. The surface of the chip is relatively precise and has high requirements for the accuracy of displacement detection. The laser displacement sensor can meet this requirement; at the same time, the stable operation of the dual-pump module can ensure the effective collection and discharge of particles on the chip surface, avoiding particle residues from affecting the detection results.

[0037] Embodiment 2: This embodiment details the method for realizing particle counting of the particle counting module.

[0038] When the particle counting module performs particle counting, the steps executed include: Step A1: Receive the real-time displacement data and particle sampling data from the displacement sensing module, and divide the displacement data into continuous time series segments based on a preset time interval. Assume that the preset time interval is 10 seconds. Within these 10 seconds, the displacement sensing module continuously collects the displacement data of the sampling head. For example, within these 10 seconds, the sampling head moves 3 cm in the X direction and 2 cm in the Y direction, and at the same time, the particle sampling device collects 10 particles. The particle counting module processes the data of these 10 seconds as one segment.

[0039] Step A2: According to the moving distance, speed, and direction of each displacement segment, convert the time series segment into a corresponding surface detection area through the area mapping algorithm, and calculate the area of each detection area. The specific implementation of the area density conversion algorithm is as follows: Divide the displacement data into several sampling units. Assume that each sampling unit corresponds to a square surface area with a side length of 1 cm. Count the number of particle samplings in each unit. For example, 3 particles are collected in a certain sampling unit, and the area of this unit region is 1 square centimeter, then the local particle density of this unit is 3 particles per square centimeter. Perform a weighted average on the local particle densities of all units to generate an overall particle distribution density value. During the weighted average process, different weights are assigned according to the importance of different units. For example, the units near the detection center area have higher weights. At the same time, eliminate the abnormal units whose discrete values exceed the preset range. Assume that the preset range is ±50% of the average value of the local particle density. If the local particle density of a certain unit is 10 particles per square centimeter, while the overall average value is 3 particles per square centimeter, and the density of this unit exceeds the range, then it is eliminated.

[0040] Step A3: Count the number of particle samplings in each detection area, and calculate the local particle density value in combination with the area of the corresponding area. For example, if the area of a certain detection area is 5 square centimeters and 15 particles are collected, then the local particle density value of this area is 3 particles per square centimeter.

[0041] Step A4: Perform spatial interpolation on the local particle density values in adjacent detection regions to generate a particle distribution density map covering the entire detection range. Through the spatial interpolation algorithm, supplement the particle density information at intermediate positions between adjacent regions to make the map more smoothly and accurately reflect the particle distribution in the entire detection region.

[0042] Step A5: Identify the abnormal regions where the density difference between adjacent regions in the particle distribution density map exceeds the preset threshold, and perform smoothing correction on the particle density values in the abnormal regions based on the spatial continuity characteristics of historical sampling data. Assume the preset threshold is 2 particles per square centimeter. If the particle density difference between two adjacent regions reaches 3 particles per square centimeter, it is determined as an abnormal region. Utilize the particle density change law around this region in the historical sampling data to adjust the particle density values in the abnormal region to make them more in line with the actual situation.

[0043] Execute Step A6: Output the corrected particle distribution density map and the particle statistical value per unit area. These data can be visually displayed to the user through a display screen or transmitted to other devices for further analysis. In practical applications, such as detecting foreign particles on the surface of food packaging in a food processing workshop, through the above workflow of the particle counting module, the particle distribution on the packaging surface can be accurately obtained, helping enterprises promptly discover product quality problems.

[0044] Example 3: In the entire particle counting monitoring device, the adaptive control module and the sampling head cooperate with each other, enabling the device to automatically adjust to the optimal sampling mode according to different detection objects and scenarios, thereby ensuring the accuracy and efficiency of the detection results.

[0045] When it is necessary to replace the sampling head, the adaptive control module plays a key role. It reads the identification code stored in the chip embedded in the sampling head through radio frequency identification technology. Radio frequency identification technology uses radio frequency signals to achieve non-contact reading of the identification code, which is convenient to operate, has a fast reading speed and high accuracy. The identification code contains rich information, covering the sampling head type, applicable surface morphology and sampling parameter range, etc. For example, when installing a new sampling head, the adaptive control module quickly reads the identification code and learns that the sampling head is a curved surface type. This information provides an important basis for subsequent adjustment of the sampling mode.

[0046] In actual application scenarios, different types of sampling heads are suitable for different detection surfaces. The flat sampling head is made of rigid materials, and its structure is firm and stable, which enables it to maintain good stability and sealing performance when conforming to flat surfaces. When detecting particles on the surface of a printed circuit board, the flat sampling head can closely conform to the surface of the circuit board, ensuring effective particle collection and reducing external interference. Its sampling parameters include a fixed suction rate and a linear displacement calibration coefficient. The fixed suction rate is set based on the research on the particle characteristics of the printed circuit board surface and actual detection experience. For example, it is set to 4 liters per minute. Such a suction rate can not only ensure that the particles on the circuit board surface are fully sucked into the detection device but also prevent damage to the circuit board or affect the detection accuracy due to excessive suction. The linear displacement calibration coefficient is used to compensate for the displacement error that may occur when the sampling head moves on the plane. It is obtained through the analysis and calibration of a large amount of experimental data to ensure the accuracy of the displacement data, thereby improving the accuracy of the particle distribution density calculation.

[0047] The curved surface sampling head is equipped with an adjustable bracket inside, and this design enables it to flexibly adapt to regular curved surfaces. Taking the detection of the curved surface of an automotive engine cylinder block as an example, the adjustable bracket can be adjusted according to the curved shape of the cylinder block. Before detection, the operator can manually or automatically control the angle and position of the bracket to match the surface of the cylinder block, making the sampling head better conform to the curved surface and ensuring the comprehensiveness and accuracy of sampling. Its sampling parameters include a dynamic suction rate and a curvature compensation coefficient. The dynamic suction rate is adjusted in real time according to the different positions and curvature changes of the curved surface. In areas with a larger curvature, since the particle distribution may be more complex, in order to ensure effective collection of sufficient particles, the suction rate will be appropriately increased; in areas with a smaller curvature, the suction rate will be decreased to avoid over-sampling or damage to the surface. The curvature compensation coefficient is used to compensate for the detection error caused by the curved surface shape. By accurately measuring and analyzing the geometric shape of the cylinder block curved surface and combining a large amount of experimental data, an appropriate curvature compensation coefficient is determined, so that when calculating the particle distribution density, the actual particle situation on the curved surface can be more accurately reflected.

[0048] The flexible sampling head is made of silica gel, which has good flexibility and elasticity, enabling it to adapt to various irregular surfaces. When detecting particles on the surface of handicrafts with complex shapes, the advantages of the flexible sampling head are fully demonstrated. It can closely conform to various concave and convex shapes on the surface of the handicraft, leaving no area where particles may exist unexamined. Its sampling parameters include adaptive suction power and displacement fluctuation tolerance range. The adaptive suction power is automatically adjusted according to the actual situation of the surface. In areas with more surface protrusions, the suction power is appropriately increased to ensure that particles on the protrusions can be collected; in concave areas, the suction power is reduced to prevent damage to the surface or inhalation of excessive impurities that may interfere with the test results. The displacement fluctuation tolerance range allows for displacement fluctuations of the sampling head within a certain range. Due to the irregularity of the handicraft surface, it is difficult for the sampling head to maintain absolute stability during movement. Setting a reasonable displacement fluctuation tolerance range can avoid misjudging displacement abnormalities due to minor displacement fluctuations, ensuring the stability and reliability of the detection process.

[0049] After obtaining the identification code of the sampling head, the adaptive control module inputs it into the AI algorithm model. Based on the information in the identification code, the AI algorithm model quickly matches the corresponding suction pump power curve, displacement sensing sensitivity, and particle counting correction coefficient, and automatically switches to the optimal sampling mode. For example, for a curved surface sampling head, the AI algorithm model will adjust the suction pump power curve according to the relevant information in its identification code. In areas with a larger curvature, the power of the suction pump is increased to improve the suction rate; in areas with a smaller curvature, the suction pump power is reduced to ensure the stability of sampling. At the same time, the displacement sensing sensitivity is increased so that the displacement sensing module can more accurately detect the minute displacement changes of the sampling head on the curved surface, providing more reliable displacement data for accurately calculating the particle distribution density. In addition, the particle counting correction coefficient is adjusted to compensate for the impact of the curved surface shape on particle counting, ensuring that the final test results can truly reflect the particle distribution on the curved surface. In this way, the adaptive control module works in coordination with different types of sampling heads, enabling the particle counting monitoring device to efficiently and accurately complete particle detection tasks on various surfaces, meeting the detection requirements of different industries.

[0050] Example 4: This example is used to describe the neural network algorithm based on dynamic weight allocation in the AI algorithm model. This algorithm plays a core role in determining the optimal sampling mode throughout the particle counting monitoring device.

[0051] In step B1, the sampling head type in the identification code is encoded as a three-dimensional orthogonal vector . Among them, , , Boolean activation values corresponding to planar, curved, and flexible sampling heads respectively. For example, if a planar sampling head is currently in use, then , , ; if it is a curved sampling head, then , , ; for a flexible sampling head, , , . This coding method can present the characteristics of different types of sampling heads in a concise and easily computable vector form, providing basic data support for subsequent steps.

[0052] Step B2: Construct the surface morphology feature matrix . Here, represents the surface roughness grade scalar, which is obtained through a professional surface roughness measuring instrument. The larger the value, the rougher the object surface. For example, after measurement, the surface roughness grade scalar of a certain object is 5. is the curvature feature vector, which describes the degree of curvature of the object surface and can be obtained by analyzing and calculating the geometric shape of the object surface. , serve as type-related adaptive coefficients, , where is the modulus of the vector , calculated through the formula . Assuming , then , , is determined according to the result of and practical experience, so as to ensure that the surface morphology feature matrix can be accurately constructed and can accurately reflect the comprehensive characteristics of the object surface.

[0053] Enter step B3, and perform parameter fusion through a dual-channel convolution kernel. For the power curve parameter , is a three-dimensional tensor convolution operation, which can perform convolution operations on data in three-dimensional space to extract data features. is the weight matrix related to the calculation of the power curve parameter, and its value is optimized through a large number of experiments and data training, determining the influence degree of the input data on the power curve parameter. represents the vertical splicing of the vector and the matrix , integrating the sampling head type information and the surface morphology feature information together. is the bias term, used to adjust the calculation result to avoid model deviation. As an activation function, a common one is the Sigmoid function, which can map the calculation result to the interval, enabling the model to have non-linear expression ability and making the calculation result of the power curve parameters more in line with the actual requirements.

[0054] Sensitivity coefficient , is the Hadamard product, that is, element-wise multiplication; is the weight matrix related to the calculation of the sensitivity coefficient; represents the vector and the matrix of the Kronecker product; is the rectified linear unit function. When the calculation result is greater than 0, the original value is output. When it is less than 0, 0 is output. This can filter out effective information, enhance the model's ability to capture key information, and thus accurately calculate the sensitivity coefficient that meets the actual detection scenario.

[0055] Correction coefficient , is the hyperbolic tangent function, which can normalize the calculation result to the interval, making the correction coefficient more reasonable. is the temperature coefficient, which can adjust the convergence speed and stability during model training. Usually, its optimal value is determined through multiple experiments during the model training process. is the weight matrix related to the calculation of the correction coefficient; represents the vector and the matrix of the horizontal concatenation, comprehensively considering the sampling head type and surface morphology characteristics to calculate the correction coefficient.

[0056] In step B4, a dynamic loss function is established. Here, , , are the deviations of the output values of the power curve parameters, sensitivity coefficient, and correction coefficient from the corresponding standard values in the preset parameter library, respectively. The preset parameter library is accumulated through a large number of experiments and practical applications and contains the ideal parameter values under different sampling head types and surface morphologies. , , are the weight coefficients, which are set according to the importance of different parameters in the detection process. For example, in a high-precision detection scenario, the weight of the deviation of the power curve parameters may be increased . Through the backpropagation algorithm, the weight matrices , , are continuously adjusted so that the loss function Minimization means minimizing the weighted sum of squares of the deviation between the output parameters and the preset parameter library, thereby optimizing the model and improving its accuracy and adaptability. In the field of electronic device manufacturing, when performing particle detection on the surface of a chip, this AI algorithm model can accurately adjust the sampling parameters according to the complex conditions on the chip surface, improving the accuracy and efficiency of particle detection.

[0057] Example 5: During the particle counting process, the abnormal feedback module continuously monitors the data of the displacement sensing module and the particle counting module. If it detects N consecutive abnormal displacement segments or the particle density value exceeds the preset reasonable range, a system abnormal signal will be generated. Here, N is a positive integer constant preset according to the actual application scenario and experience. For example, when performing particle detection on the surface of a precision optical lens, since the surface quality requirements of the lens are extremely high, the particle distribution density should be strictly controlled within a certain range. Suppose . If the displacement sensing module detects that the displacement fluctuation amplitude of the sampling head exceeds the preset fluctuation threshold for 3 consecutive times. For example, the preset fluctuation threshold is 0.5 mm, and the actually detected displacement fluctuation amplitudes for 3 consecutive times are 0.6 mm, 0.7 mm, and 0.65 mm respectively, or the particle density value calculated by the particle counting module exceeds the preset reasonable range for 3 consecutive times. Suppose the preset reasonable range of particle density is 0 - 5 particles per square centimeter, and the actually calculated particle density values for 3 consecutive times are 6 particles per square centimeter, 7 particles per square centimeter, and 6.5 particles per square centimeter respectively, then the abnormal feedback module will quickly generate a system abnormal signal.

[0058] This system abnormal signal will immediately trigger the adaptive control module to pause the current sampling process. After receiving the signal, the adaptive control module will immediately stop the operation of the suction pump and the blowing pump, and at the same time pause the data acquisition of the particle counting module and the displacement sensing module to prevent the further accumulation of incorrect data. Subsequently, a self-check program is started to conduct item-by-item detection on the double-pump module, the displacement sensing module, and the particle counting module.

[0059] When detecting the double-pump module, a pressure pulse test is carried out by sending pressure pulse signals with specific frequencies and intensities to the suction pump and the blowing pump. For example, a pressure pulse signal with a frequency of 20 Hz and an intensity of 60 kPa is sent. If the pulse response time of the suction pump or the blowing pump exceeds the preset threshold, suppose the preset threshold is 0.4 s, and the actually detected pulse response time of a certain pump is 0.6 s, then it is determined that there is a fault in that pump body. The pump body fault may be caused by reasons such as wear of the impeller inside the pump, blockage of the pipeline, or loose sealing. Once a pump body fault is found, it is necessary to repair or replace the relevant components in time to ensure that the double-pump module can work normally and ensure the stable operation of the particle inhalation and discharge operations.

[0060] Perform zero drift calibration on the displacement sensor module. Ideally, when the sampling head is stationary, the output value of the displacement sensor module should be 0. However, in actual use, factors such as sensor aging and environmental interference may cause deviations in the output value, i.e. zero drift. Use professional calibration equipment to calibrate the displacement sensor module. If the drift value after calibration still exceeds the tolerance range, for example, the tolerance range is set to ±0.05 mm, and the drift value after calibration is 0.1 mm, the sensor is judged to be failed. Sensor failure will seriously affect the accuracy of the displacement data, and thus affect the calculation results of particle counts and distribution density, so it is necessary to replace the new sensor in time.

[0061] Perform a standard particle sample test on the particle counting module. Prepare a standard sample with known particle quantity and distribution, place it in the detection device, and let the particle counting module detect it. If the counting error exceeds the allowable value, for example, the allowable counting error is ±3%, and the actual counting error reaches 5%, reload the particle density conversion algorithm. Reloading the algorithm can try to fix counting error problems caused by algorithm operation errors, data anomalies, or unreasonable parameter settings, to ensure that the particle counting module can accurately calculate the particle distribution density. In the pharmaceutical industry, when performing particle detection on the surface of drug packaging, the effective operation of the abnormal feedback module can promptly discover problems in the detection process, ensure the quality and safety of drugs, and prevent unqualified products from entering the market.

[0062] Embodiment 6: When performing a pressure pulse test on the dual pump module, a special pressure pulse generator is used to send a pressure pulse signal to the suction pump and the blow pump. The frequency and intensity of the pressure pulse signal are carefully set according to the pump model, specifications and design requirements. For example, for a certain model of suction pump and blow pump, after technical analysis and experimental verification, the pressure pulse frequency is set to 15Hz and the intensity is set to 55kPa. After sending the pressure pulse signal, the pulse response time of the pump is accurately recorded by a high-precision pressure sensor and time measurement device. If the mechanical parts inside the pump body, such as the impeller, bearing, etc., are worn, or there is blockage or leakage in the pipeline, the pump's response ability will decrease and the pulse response time will be extended. When the pulse response time exceeds the preset threshold, for example, the preset threshold is 0.5 seconds, and the actual pulse response time of a pump is detected to be 0.7 seconds, it is determined to be a pump body failure. Once the pump body failure is determined, the maintenance personnel will perform maintenance according to the specific situation. If the impeller is worn, replace it with a new one; if the pipeline is blocked, clean the pipeline; if the seal is damaged and causes leakage, replace the seal to ensure that the pump body returns to normal working condition and ensure the efficiency and stability of particle transportation.

[0063] For the zero-drift calibration of the displacement sensing module, a professional calibration platform is used. The displacement sensing module is installed on the calibration platform, and the output value of the displacement sensing module is measured when the calibration platform is in a stationary state. Under normal circumstances, the output value at this time should be 0, but due to the influence of various factors, there may be deviations. If the calibrated drift value still exceeds the tolerance range, for example, the tolerance range is set to ±0.08 mm, and the measured drift value after calibration is 0.12 mm, it is determined that the sensor has failed. The failure of the sensor will cause errors in the collected displacement data, resulting in deviations in the calculated particle distribution density and affecting the accuracy of the detection results. Therefore, once it is determined that the sensor has failed, a new sensor of the same model and specification needs to be replaced in time to ensure the reliability of the displacement data.

[0064] When performing a standard particle sample test on the particle counting module, a standard particle sample is selected. The number and distribution of the standard particle sample have been strictly measured and verified, with high accuracy and reliability. The standard particle sample is placed in the detection device, and the particle counting module is started for detection. By comparing the detection results of the particle counting module with the actual situation of the standard particle sample, the counting error is calculated. If the counting error exceeds the allowable value, assuming the allowable counting error is ±4% and the actual counting error reaches 6%, the particle density conversion algorithm is reloaded. The process of reloading the algorithm includes re-initializing the parameters of the algorithm, reading the latest calibration data, etc. After reloading the algorithm, the standard particle sample is detected again to observe whether the counting error is within the allowable range. If it still does not meet the requirements, it may be necessary to further check whether there are problems in the code logic of the algorithm and the data transmission process, or to detect and repair the hardware of the particle counting module to ensure that the particle counting module can accurately calculate the particle distribution density. In the field of aerospace component manufacturing, when performing particle detection on the surface of components, a strict self-inspection procedure can effectively ensure the accuracy and reliability of the detection device, ensure that the quality of the components meets high-standard requirements, and guarantee the safe operation of aerospace equipment.

[0065] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0066] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A particle counting monitoring device combining a dual-pump system with displacement sensing technology, characterized in that It includes a dual-pump module, a displacement sensing module, a particle counting module, and an adaptive control module; The dual-pump module includes an inhalation pump and a blowing pump. The inhalation pump and the blowing pump perform particle inhalation and discharge operations through independent channels respectively, and the start-stop timing of both is coordinated by a synchronous controller; The displacement sensing module is used to collect the displacement data of the sampling head in real time, including the moving distance, speed, and direction; The particle counting module calculates the particle distribution density per unit area through a regional density conversion algorithm based on the displacement data and particle sampling data; The adaptive control module obtains identification information through a chip embedded in the sampling head, and determines the optimal sampling mode of the sampling head based on the identification information through a preset AI algorithm model.

2. The particle counting monitoring device according to claim 1, characterized in that, The specific implementation method of the displacement sensing module is as follows: A laser displacement sensor or a capacitive displacement sensor is used to detect the relative position between the sampling head and the object surface in real time; During the sampling process, the displacement data is segmented by time series, the cumulative value of each segment of displacement is calculated and data fusion is performed with the particle sampling quantity to generate a particle distribution density map; If the fluctuation amplitude of a certain segment of displacement data exceeds the preset fluctuation threshold, it is marked as an abnormal displacement segment, and the adaptive control module is triggered to recalibrate the sampling parameters.

3. The particle counting monitoring device according to claim 1, characterized in that, The synchronous control process of the dual-pump module includes: When the inhalation pump starts, the blowing pump starts after a delay of T1 time to ensure that the inhalation air flow is stable before performing particle discharge; During the operation of the blowing pump, the exhaust pressure value is monitored in real time. If the deviation between the exhaust pressure and the preset pressure value exceeds the threshold, a pump pressure abnormal signal is generated and fed back to the adaptive control module to adjust the power of the blowing pump.

4. The particle counting monitoring device according to claim 1, characterized in that, The particle counting implementation method of the particle counting module includes: Step A1: Receive the real-time displacement data and particle sampling data from the displacement sensing module, and divide the displacement data into continuous time series segments based on a preset time interval; Step A2: According to the moving distance, speed, and direction of each segment of displacement, convert the time series segment into a corresponding surface detection area through a regional mapping algorithm, and calculate the area of each detection area; The implementation method of the regional density conversion algorithm is as follows: The displacement data is divided into several sampling units, and each unit corresponds to a surface area; Count the number of particle samples in each unit, and calculate the local particle density in combination with the unit area; Perform weighted averaging on the local particle densities of all units to generate an overall particle distribution density value, and eliminate abnormal units whose discrete values exceed the preset range; Step A3: Count the number of particle samples in each detection area, and calculate the local particle density value in combination with the area of the corresponding area; Step A4: Perform spatial interpolation processing on the local particle density values of adjacent detection areas to generate a particle distribution density map covering the entire detection range; Step A5: Identify abnormal areas where the density difference between adjacent areas in the particle distribution density map exceeds the preset threshold, and perform smoothing correction on the particle density values of the abnormal areas based on the spatial continuity characteristics of historical sampling data; Step A6: Output the corrected particle distribution density map and the particle statistics value per unit area.

5. The particle counting monitoring device according to claim 1, characterized in that, The specific operation mode of the adaptive control module is as follows: When the sampling head is replaced, the identification code stored in the chip is read through radio frequency identification technology, including the sampling head type, applicable surface morphology and sampling parameter range; The identification code is input into the AI algorithm model to match the corresponding air pump power curve, displacement sensor sensitivity and particle count correction coefficient, and automatically switch to the optimal sampling mode.

6. The particle counting monitoring device according to claim 5, characterized in that, The sampling head includes three types: flat type, curved type and flexible type; The planar sampling head is made of rigid material and is suitable for flat surfaces. Its sampling parameters include a fixed suction rate and a linear displacement calibration coefficient. The curved surface sampling head has a built-in adjustable bracket and is suitable for regular curved surfaces. Its sampling parameters include dynamic air suction rate and curvature compensation coefficient. The flexible sampling head is made of silica gel and is suitable for irregular surfaces. Its sampling parameters include adaptive suction power and displacement fluctuation tolerance range.

7. The particle counting monitoring device according to claim 6, characterized in that, The AI algorithm model adopts a neural network algorithm based on dynamic weight allocation, which specifically includes the following steps: Step B1: Encode the sampling head type in the identification code as a three-dimensional orthogonal vector , where correspond to the Boolean activation values of planar, curved, and flexible types respectively; Step B2: Construct the surface morphology feature matrix , where is the surface roughness grade scalar, is the curvature feature vector, , are type-related adaptive coefficients and satisfy ; Step B3: Parameter fusion is performed through a dual-channel convolution kernel, where: Power curve parameters ; Sensitivity coefficient ; Correction factor ; Among them represents 3D tensor convolution, represents the Hadamard product, represents the activation function, represents the rectified linear unit function, represents the hyperbolic tangent function, is the temperature coefficient, respectively represent the weight matrices related to the calculation of power curve parameters, related to the calculation of sensitivity coefficients, and related to the calculation of correction coefficients; represents the vector and the matrix for vertical concatenation, represents the vector and the matrix for Kronecker product, represents the vector and the matrix for horizontal concatenation; represents the bias term; Step B4: Establish a dynamic loss function , where is the deviation value between the output parameter and the preset parameter library; optimize the weight matrix through backpropagation, so that the weighted sum of squares of the deviation between the output parameter and the preset parameter library is minimized, is the weight coefficient.

8. The particle counting monitoring device according to claim 1, wherein It also includes an exception feedback module, which operates as follows: During the particle counting process, if N consecutive abnormal displacement segments are detected or the particle density value exceeds the preset reasonable range, a system abnormality signal is generated, where N is a preset positive integer constant; The system abnormal signal triggers the adaptive control module to suspend the current sampling process, and starts the self-check program to check the dual pump module, the displacement sensor module and the particle counting module item by item.

9. The particle counting monitoring device according to claim 8, characterized in that The specific steps of the self-test procedure include: Perform pressure pulse test on the suction pump and the blowing pump. If the pulse response time exceeds the preset threshold, it is judged as a pump failure. Perform zero drift calibration on the displacement sensor module. If the drift value after calibration still exceeds the tolerance range, the sensor is considered to be failed. Perform a standard particle sample test on the particle counting module. If the counting error exceeds the allowable value, reload the particle density conversion algorithm.

10. A particle counting monitoring method based on the device according to any one of claims 1-9, characterized in that, The following steps are involved: The suction pump and the blowing pump in the dual-pump module are used to perform particle suction and discharge operations through independent channels, and the start and stop timing of the two are coordinated by the synchronous controller; The displacement sensor module collects the displacement data of the sampling head in real time, including the moving distance, speed and direction, and transmits the displacement data to the particle counting module; The particle counting module calculates the particle distribution density within a unit area using the regional density conversion algorithm based on the received displacement data and particle sampling data; An adaptive control module is used to obtain identification information through a chip embedded in the sampling head, and a preset AI algorithm model is used to determine the optimal sampling mode of the sampling head based on the identification information.

Images