Method and system for early detection of micro-leakage of inner bag of electronic grade polysilicon packaging bag

Abstract

This invention relates to the field of electronic-grade polysilicon production technology, and particularly to a method and system for early detection of micro-leakage in the inner bag of electronic-grade polysilicon packaging bags. The method includes: acquiring initial sealing baseline data of the inner bag of the packaged target packaging bag; collecting real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data according to a preset sampling cycle; calculating sealing state characteristic parameters such as the inert gas concentration decay rate, pressure difference change rate, and temperature and humidity compensation correction value based on the above data; inputting the sealing state characteristic parameters into a preset micro-leakage judgment model to obtain the micro-leakage judgment result; generating a leakage warning signal when a micro-leakage is detected; by establishing an initial sealing baseline and continuously monitoring changes in the sealing state, slow micro-leakage that gradually develops during storage, transportation, or stacking can be identified, and interference from environmental fluctuations and bag rebound factors can be reduced through temperature and humidity compensation, thereby improving the stability and reliability of early detection of micro-leakage.

Description

Technical Field

[0001] This invention relates to the field of electronic-grade polysilicon production technology, and in particular to a method and system for early detection of micro-leakage in the inner bag of electronic-grade polysilicon packaging bags. Background Technology

[0002] Electronic-grade polysilicon is an important high-purity raw material in the semiconductor industry. It is extremely sensitive to water vapor, oxygen, dust, and other external pollutants. Therefore, during packaging, storage, and transportation, it is usually necessary to rely on the inner bag of the packaging bag to form a clean and stable isolation environment. The inner bag of the packaging bag is generally the inner layer of the bag that directly contains or directly isolates the electronic-grade polysilicon. It is usually sealed by heat sealing or other methods, and inert gases such as nitrogen and argon can be filled into the bag to maintain a protective atmosphere. The so-called micro-leakage refers to the slow leakage channels at the nanometer to micrometer level formed in the sealing area, film area, or stress bending area of ​​the inner bag. It is different from obvious bag breakage or large leakage, and also different from the normal air permeability of the packaging material itself. This type of micro-leakage is usually not easy to be detected by appearance in the early stage, but may gradually cause the protective atmosphere of the inner bag to fail during storage, long-distance transportation, or stacking adjustment.

[0003] In existing technologies, the sealing control of electronic-grade polysilicon packaging bags mostly adopts a single offline negative pressure leak detection method after heat sealing, or uses a general fixed threshold for leak judgment. That is, after packaging, the sealing is judged by pressure holding test to determine whether the seal is qualified. This method can detect immediate leaks or obvious sealing defects after heat sealing, but it is difficult to cover subsequent stages such as storage and transportation after the product leaves the factory, and it is also difficult to reflect the trend of the slow change of the inner bag sealing status over time. At the same time, existing fixed threshold detection usually does not take into account the changes in inert gas composition, changes in internal and external pressure difference, and the influence of environmental temperature and humidity for comprehensive judgment. It is easily affected by packaging specifications, materials, vacuuming status and external environmental fluctuations, leading to missed detection or misjudgment of early micro-leakage. Therefore, the main problem of existing technologies is that it is difficult to make stable and reliable early judgment of micro-leakage before the inner bag of electronic-grade polysilicon packaging bags has obvious failure.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the general background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] This invention provides a method and system for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags, thereby effectively solving the problems in the background art.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for early detection of micro-leakage in the inner bag of an electronic-grade polycrystalline silicon packaging bag, comprising the following steps: Obtain the initial sealing reference data of the inner bag of the electronic-grade polycrystalline silicon target packaging bag after the encapsulation is completed. The initial sealing reference data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. According to the preset sampling cycle, collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag of the target packaging bag, and current ambient temperature and humidity data. Based on the initial sealing reference data, the real-time gas composition data, the real-time pressure difference data inside and outside the inner bag, and the current ambient temperature and humidity data, the sealing state characteristic parameters of the inner bag are calculated. The sealing state characteristic parameters include the inert gas concentration decay rate, the pressure difference change rate, and the temperature and humidity compensation correction value. The sealing state characteristic parameters are input into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag; When the micro-leakage determination result indicates the presence of a micro-leakage, a corresponding leak warning signal is generated.

[0007] Furthermore, based on the initial sealing reference data and the various data collected in real time, the sealing state characteristic parameters of the inner bag are calculated, specifically including: Based on the initial inert gas component concentration and the real-time gas component data, the inert gas concentration decay rate per unit time is calculated. Based on the initial pressure difference between the inside and outside of the inner bag and the real-time pressure difference data between the inside and outside of the inner bag, the rate of change of pressure difference per unit time is calculated; Based on the initial temperature and humidity data of the packaging environment and the current temperature and humidity data, a temperature and humidity compensation correction value is calculated. The inert gas concentration decay rate and pressure difference change rate are corrected using the temperature and humidity compensation correction value to obtain the corrected sealing state characteristic parameters.

[0008] Furthermore, after calculating the sealing state characteristic parameters of the inner bag, time-series fitting is performed on the sealing state characteristic parameters at multiple sampling times to generate the change trend curve of the sealing state of the inner bag. Based on the trend curve, the sealing status change of the inner bag within a preset time period is predicted by a time series prediction algorithm. When the prediction result reaches the micro-leakage warning threshold, an early warning signal is generated.

[0009] Furthermore, the construction steps of the preset micro-leakage determination model include: Acquire historical sealing sample data of multiple sets of inner bags of electronic-grade polycrystalline silicon packaging bags. The historical sealing sample data includes the sealing state characteristic parameters and corresponding labels of no-leakage samples, micro-leakage samples and macro-leakage samples. The historical sealed sample data is divided into a training set and a validation set; The initial neural network model is trained under supervision using the training set, and the trained model is then validated and its parameters optimized using the validation set to generate the preset micro-leakage determination model.

[0010] Further, the step of inputting the sealing state characteristic parameters into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag specifically includes: The sealing state characteristic parameters after temperature and humidity compensation correction are input into the preset micro-leakage judgment model, and the three classification results of no leakage, micro-leakage, and macro-leakage corresponding to the inner bag of the target packaging bag are output. When the classification result is micro-leakage, it is determined that there is micro-leakage in the inner bag of the target packaging bag, and the corresponding leakage level and leakage probability value are output.

[0011] Furthermore, the method also includes: When the microleakage determination result indicates the presence of microleakage, the estimated range of the microleakage aperture and the real-time leakage rate of the inner bag are calculated based on the inert gas concentration decay rate, the pressure difference change rate, the inner bag volume, and the electronic-grade polysilicon volume parameters. Based on the estimated range of the micro-leakage aperture, the real-time leakage rate, and the purity protection threshold of electronic-grade polysilicon, corresponding disposal recommendations are generated.

[0012] Furthermore, the generation of corresponding handling suggestion information specifically includes: The remaining safe storage time is calculated based on the estimated range of the micro-leakage aperture, the real-time leakage rate, and the purity protection threshold of electronic-grade polysilicon. Based on the remaining safe storage time, tiered disposal recommendations are generated.

[0013] Furthermore, the preset sampling period is a dynamically adjustable period, specifically including: Based on the fluctuation range of the current ambient temperature and humidity data, the changing trend of the sealing state characteristic parameters, and the storage and transportation stage of the target packaging bag, the duration and sampling accuracy of the preset sampling period are dynamically adjusted.

[0014] Furthermore, the acquisition of initial sealing reference data for the inner bag of the packaged electronic-grade polysilicon target packaging bag specifically includes: After the electronic-grade polysilicon filling, inert gas encapsulation, and heat sealing of the inner bag of the target packaging bag are completed and a preset standing time is reached, the initial inert gas component concentration inside the inner bag and the initial pressure difference between the inside and outside of the inner bag are collected. At the same time, the initial environmental temperature and humidity data at the sealing point are collected to generate the initial sealing reference data and store it in association with the database corresponding to the unique identifier of the target packaging bag.

[0015] The present invention also includes a micro-leakage early detection system for the inner bag of an electronic-grade polycrystalline silicon packaging bag, the system comprising: The initial reference acquisition module is used to acquire the initial sealing reference data of the inner bag of the packaged electronic-grade polycrystalline silicon target packaging bag. The initial sealing reference data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. The real-time data acquisition module is used to collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data of the target packaging bag according to a preset sampling period. The feature parameter calculation module is used to calculate the sealing state feature parameters of the inner bag based on the initial sealing reference data, the real-time gas composition data collected in real time, the real-time pressure difference data inside and outside the inner bag, and the current ambient temperature and humidity data. The sealing state feature parameters include the inert gas concentration decay rate, the pressure difference change rate, and the temperature and humidity compensation correction value. The micro-leakage determination module is used to input the sealing state characteristic parameters into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag; The early warning signal generation module is used to generate a corresponding leakage early warning signal when the micro-leakage determination result indicates the presence of a micro-leakage.

[0016] The beneficial effects of this invention are as follows: By establishing an initial sealing benchmark after the inner bag of the electronic-grade polycrystalline silicon packaging bag is sealed, and continuously acquiring data such as gas composition, internal and external pressure difference, and ambient temperature and humidity during subsequent testing, the sealing status of the inner bag can be transformed from a traditional single-time pass judgment to an early judgment based on dynamic changes. Compared with existing single offline negative pressure leak detection or fixed threshold leak detection methods after heat sealing, this invention can not only detect sealing abnormalities that already exist when the packaging is completed, but also identify slow micro-leaks that gradually develop during storage, transportation, or stacking. Thus, it generates an early warning signal before the inner bag shows obvious damage, collapse, bulging, or complete failure of the protective atmosphere, providing a time window for early intervention for re-inspection, isolation, or repackaging.

[0017] Meanwhile, by comprehensively considering the decay of inert gas concentration, changes in internal and external pressure difference, and the influence of ambient temperature and humidity during the judgment process, it can avoid the one-sided judgment caused by relying solely on a single pressure difference or a single gas component detection. Through temperature and humidity compensation correction, it can reduce the interference of non-leakage factors such as environmental fluctuations, bag rebound, and transportation disturbances on the detection results, improve the stability and reliability of micro-leakage judgment, and reduce misjudgments and missed detections.

[0018] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of a method for early detection of micro-leakage in the inner bag of electronic-grade polysilicon packaging bags; Figure 2 A flowchart illustrating the sealing state characteristic parameters of the inner bag; Figure 3 This is a schematic diagram of a micro-leakage early detection system for the inner bag of an electronic-grade polycrystalline silicon packaging bag. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Example 1:

[0023] like Figures 1 to 2 As shown, this application provides a method for early detection of micro-leakage in the inner bag of an electronic-grade polycrystalline silicon packaging bag, the method comprising: S10: Obtain the initial sealing reference data of the inner bag of the electronic-grade polysilicon target packaging bag after packaging. The initial sealing reference data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. S20: Collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data of the target packaging bag according to the preset sampling cycle; S30: Based on the initial sealing reference data, real-time gas composition data, real-time pressure difference data inside and outside the inner bag, and current ambient temperature and humidity data, the sealing state characteristic parameters of the inner bag are calculated. The sealing state characteristic parameters include the inert gas concentration decay rate, pressure difference change rate, and temperature and humidity compensation correction value. S40: Input the sealing status characteristic parameters into the preset micro-leakage judgment model to obtain the micro-leakage judgment result of the inner bag of the target packaging bag; S50: When the micro-leakage determination result indicates the existence of a micro-leakage, a corresponding leak warning signal is generated.

[0024] Specifically, the initial sealing baseline data of the inner bag is first obtained. This data includes the initial inert gas concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the sealing environment. The initial inert gas concentration characterizes the initial state of the protective atmosphere inside the inner bag upon completion of sealing. The initial pressure difference between the inside and outside of the inner bag characterizes the pressure maintenance state of the inner bag relative to the external environment upon completion of sealing. The initial temperature and humidity data of the sealing environment serve as a reference for subsequent environmental compensation. To reduce the impact of residual heat from sealing, inflation disturbance, bag rebound, or conveyor compression on the initial data, the initial sealing baseline data can be collected again after a preset stabilization time following sealing. This data is then compared with the batch information of the target packaging bag, sealing time, and sealing equipment information. Alternatively, the packaging bag identification information can be associated and stored. Subsequently, continuous or intermittent sampling is performed on the inner bag of the target packaging bag according to a preset sampling cycle. The collected data includes real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data. The preset sampling cycle can be set according to production cycle, detection sensitivity requirements, and the residence time of the packaging bag. It can employ fixed-cycle sampling or a phased sampling method, such as increasing the sampling frequency in the early stages after sealing and decreasing it after the sealing state stabilizes. When abnormal trends appear in the detection data, the sampling cycle can be automatically shortened to further confirm the presence of micro-leakage. During sampling, real-time gas composition data can include inert gas concentration and oxygen concentration. The data processing unit collects data on inert gas concentration, water vapor content, or tracer gas concentration; real-time differential pressure data reflects the gas retention capacity inside the inner bag; and current ambient temperature and humidity data identifies the impact of environmental changes on the detection results. To improve the reliability of the judgment, the real-time acquired data can be preprocessed before entering subsequent calculations. Preprocessing includes one or more of the following: data smoothing, outlier removal, sensor drift correction, sampling time alignment, and data validity judgment. This avoids misjudgments caused by short-term handling and compression, instantaneous sensor fluctuations, or environmental disturbances. After obtaining the initial sealing reference data and real-time acquired data, the data processing unit calculates the sealing state characteristic parameters of the inner bag. These sealing state characteristic parameters include the inert gas concentration decay rate, differential pressure change rate, and temperature and humidity compensation correction. Positive values ​​are defined as follows: the inert gas concentration decay rate characterizes the degree to which the protective atmosphere inside the inner bag weakens over time, and can be determined based on the changing trend between the initial inert gas component concentration and the real-time inert gas component concentration over a period of time; the pressure difference change rate characterizes the degree to which the pressure difference between the inside and outside of the inner bag decreases or changes over time, and can be determined based on the changing trend between the initial pressure difference and the real-time pressure difference over a period of time; and the temperature and humidity compensation correction value is used to correct the influence of changes in ambient temperature and humidity on gas concentration detection, pressure difference detection, and bag deformation, so that the data used in the final judgment can better reflect the sealing status of the inner bag itself, rather than changes in the external environment. Specifically, when the ambient temperature changes, the gas pressure inside the inner bag may fluctuate normally due to thermal expansion and contraction.When ambient humidity changes, sensor output, bag film flexibility, or the state of the sealing area may also change. Therefore, the current ambient temperature and humidity are compared with the initial ambient temperature and humidity, and corrections are made for changes in gas composition and pressure difference based on a pre-calibrated compensation relationship. The compensation relationship can be established using detection data from leak-free standard packaging bags under different temperature and humidity conditions, or it can be updated based on historical detection data from packaging bags of the same specifications. By introducing temperature and humidity compensation correction values, false alarms caused by environmental changes can be reduced, and the stability of early micro-leakage detection can be improved. After obtaining the sealing state characteristic parameters, the inert gas concentration decay rate, pressure difference change rate, and temperature and humidity compensation correction values ​​are input into a preset micro-leakage detection model. This micro-leakage detection model outputs the micro-leakage detection result of the inner bag of the target packaging bag. The micro-leakage detection model can use a threshold detection model, a weighted scoring model, a statistical discriminant model, or a classification model trained on samples. For threshold detection... The model allows for setting thresholds for inert gas concentration decay, pressure difference change, and comprehensive changes after temperature and humidity compensation. It determines the presence of micro-leakage when multiple detection indicators simultaneously meet abnormal conditions. For the weighted scoring model, standardized parameters of each sealing state are used to generate a comprehensive risk score, which is then used to determine whether the target packaging bag has micro-leakage. For the classification model trained on samples, detection data from leak-free and micro-leakage sample bags can be pre-collected. The trends in inert gas concentration, pressure difference, and temperature and humidity compensation are used as input features to train a model that distinguishes between normal sealing and micro-leakage states. To reduce false alarms caused by occasional fluctuations, the micro-leakage determination model can also combine multiple consecutive sampling results for judgment. Only when the sealing state parameters continuously show a leakage trend within a preset time window, or when the comprehensive risk result continuously reaches the warning condition, will a micro-leakage determination result be output.

[0025] By establishing an initial sealing benchmark after the inner bag of the electronic-grade polycrystalline silicon packaging bag is sealed, and continuously acquiring data such as gas composition, internal and external pressure difference, and ambient temperature and humidity during subsequent testing, the sealing status of the inner bag can be transformed from a traditional single-time pass judgment to an early judgment based on dynamic changes. Compared with existing single offline negative pressure leak detection or fixed threshold leak detection methods after heat sealing, this invention can not only detect sealing abnormalities that already exist when the packaging is completed, but also identify slow micro-leaks that gradually develop during storage, transportation, or stacking. Thus, it generates an early warning signal before the inner bag shows obvious damage, collapse, bulging, or complete failure of the protective atmosphere, providing a time window for early intervention for re-inspection, isolation, or repackaging.

[0026] Meanwhile, the present invention comprehensively considers the decay of inert gas concentration, changes in internal and external pressure difference, and the influence of ambient temperature and humidity during the judgment process. This can avoid the one-sided judgment caused by relying solely on a single pressure difference or a single gas component detection. Through temperature and humidity compensation correction, the interference of non-leakage factors such as environmental fluctuations, bag rebound, and transportation disturbances on the detection results can be reduced, thereby improving the stability and reliability of micro-leakage judgment and reducing misjudgments and missed detections.

[0027] In step S10, the initial sealing reference data of the inner bag of the packaged electronic-grade polysilicon target packaging bag is obtained, specifically including: After the electronic-grade polysilicon filling, inert gas encapsulation, and heat sealing of the inner bag of the target packaging bag are completed and a preset standing time is set, the initial inert gas component concentration inside the inner bag and the initial pressure difference between the inside and outside of the inner bag are collected. At the same time, the initial environmental temperature and humidity data of the sealing point are collected to generate initial sealing reference data and store it in the database corresponding to the unique identifier of the target packaging bag.

[0028] Specifically, when acquiring the initial sealing baseline data of the inner bag of the target packaging bag, after the target packaging bag completes electronic-grade polysilicon filling, inert gas encapsulation, and heat sealing, it first undergoes a preset settling time before initial data acquisition. The purpose of setting the preset settling time is to allow the residual heat of the seal after heat sealing, bag rebound, inflation disturbance, material settling, and internal and external pressure difference to stabilize, avoiding the impact of instantaneous fluctuations immediately after sealing on the initial sealing baseline. The preset settling time can be set according to the packaging bag specifications, inner bag material, heat sealing parameters, inflation pressure, and production cycle. After reaching the preset settling time, the data acquisition unit collects the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial ambient temperature at the sealing point. Humidity data is used to generate initial sealing baseline data. The initial inert gas component concentration is used to characterize the initial protective atmosphere state after the inner bag is sealed, the initial pressure difference between the inside and outside of the inner bag is used to characterize the initial pressure maintenance state of the inner bag, and the initial ambient temperature and humidity data are used as the benchmark for subsequent temperature and humidity compensation and sealing state change analysis. After generating the initial sealing baseline data, it is associated with the unique identifier of the target packaging bag and stored in the database. The unique identifier can be a barcode, QR code, RFID tag or electronic batch number. When performing real-time sampling and micro-leakage judgment, the data processing unit can call the corresponding initial sealing baseline data according to the unique identifier, so that each target packaging bag has an independent judgment reference and facilitates quality traceability.

[0029] As a preferred embodiment of the above, in step S20, the preset sampling period is a dynamically adjustable period, specifically including: Based on the fluctuation range of current environmental temperature and humidity data, the changing trend of sealing status characteristic parameters, and the storage and transportation stage of the target packaging bag, the duration and sampling accuracy of the preset sampling cycle are dynamically adjusted.

[0030] Specifically, when collecting real-time gas composition data, real-time pressure difference data inside and outside the inner bag, and current ambient temperature and humidity data, the data processing unit dynamically adjusts the sampling cycle length and sampling accuracy based on the fluctuation range of the current ambient temperature and humidity data, the changing trend of the sealing status characteristic parameters, and the storage and transportation stage of the target packaging bag. The sampling cycle length refers to the time interval between two adjacent samples, and the sampling accuracy can include sensor sampling resolution, number of repeated samples, duration of a single sample, and data smoothing processing accuracy. When the ambient temperature and humidity fluctuations are small and the sealing status characteristic parameters remain stable, the sampling cycle can be extended and the conventional sampling accuracy can be used. When the ambient temperature and humidity fluctuations are large, or when the sealing status characteristic parameters show continuous changes or approach the warning threshold, the sampling cycle can be shortened and the sampling accuracy can be improved to obtain more continuous and reliable detection data.

[0031] As a preferred embodiment of the above, in step S30, the sealing state characteristic parameters of the inner bag are calculated based on the initial sealing reference data and the various data collected in real time, specifically including: S31: Calculate the inert gas concentration decay rate per unit time based on the initial inert gas component concentration and real-time gas component data; S32: Calculate the rate of change of pressure difference per unit time based on the initial pressure difference between the inside and outside of the inner bag and the real-time pressure difference between the inside and outside of the inner bag; S33: Based on the initial temperature and humidity data of the packaging environment and the current temperature and humidity data, calculate the temperature and humidity compensation correction value, and use the temperature and humidity compensation correction value to correct the inert gas concentration decay rate and pressure difference change rate to obtain the corrected sealing state characteristic parameters.

[0032] Based on the initial temperature and humidity data of the packaging environment and the current temperature and humidity data, a temperature and humidity compensation correction value is calculated. In this embodiment, the temperature and humidity compensation correction value Pc is calculated as follows: ; Wherein, Pc represents the temperature and humidity compensation correction value; T1 represents the current ambient temperature, and T0 represents the initial temperature of the packaging environment; H1 represents the current ambient relative humidity, and H0 represents the initial relative humidity of the packaging environment; a is the temperature influence coefficient, which represents the change in internal air pressure of the inner bag caused by a unit temperature change; b is the humidity influence coefficient, which represents the drift of the gas sensor output caused by a unit relative humidity change; the temperature influence coefficient a and the humidity influence coefficient b are obtained in advance through testing and calibration on standard sample bags of the same specification without leakage under different temperature and humidity conditions, and are stored in the database for retrieval.

[0033] Specifically, after the inner bag of the target packaging bag is sealed and initial sealing reference data is collected, the data processing unit uses the initial inert gas component concentration as the reference value of the protective atmosphere of the inner bag, and acquires real-time gas component data in subsequent sampling periods. The real-time gas component data can be the concentration data of inert gases such as nitrogen, argon, and helium inside the inner bag, or it can be a combination of inert gas concentration and oxygen concentration, water vapor content, or tracer gas concentration. Based on the difference between the initial inert gas component concentration and the real-time gas component data, and in conjunction with the corresponding sampling time, the data processing unit calculates the inert gas concentration decay rate per unit time. This inert gas concentration decay rate is used to characterize the rate at which the protective atmosphere inside the inner bag decreases over time. When the inert gas concentration continuously increases or remains within an abnormal range over multiple sampling periods, it indicates a decrease in the inert gas retention capacity of the inner bag. The inert gas concentration decay rate can be determined by the difference between the current sampling point and the initial reference value, or by a sliding time window formed by several recent consecutive sampling points. When using a sliding time window, the data processing unit obtains the decay rate based on the overall trend of gas component data within the window to reduce the impact of single-detection fluctuations, residual gas in the sampling pipeline, instantaneous sensor disturbances, or localized pressure on the bag on the calculation results. Simultaneously, the data processing unit uses the initial pressure difference between the inside and outside of the inner bag as the pressure retention reference value for the target packaging bag and acquires real-time pressure difference data between the inside and outside of the inner bag during subsequent sampling processes, then... Based on the relationship between the initial differential pressure and the real-time differential pressure over time, the rate of change of differential pressure per unit time is calculated. For inner bags sealed under positive pressure, the rate of change of differential pressure reflects the decreasing trend of differential pressure caused by the slow outward escape of gas from inside the inner bag. For inner bags sealed under negative pressure or slightly negative pressure, the rate of change of differential pressure reflects the decreasing trend of the absolute value of differential pressure caused by the slow inward entry of gas from outside the inner bag. Therefore, the data processing unit can determine the direction of differential pressure change according to different sealing conditions, so that the rate of change of differential pressure can characterize the ability of the inner bag to maintain the internal and external pressure difference. To avoid instantaneous differential pressure fluctuations caused by transportation, handling, palletizing, or short-term compression being mistaken for leakage, the rate of change of differential pressure can be determined based on the changing trend of multiple consecutive sampling points. When a certain sampling point shows a change in... When there is a sudden change in the overall trend, the validity of the sampling point can be judged by combining the changes in adjacent sampling points and gas composition. Furthermore, the data processing unit calculates the temperature and humidity compensation correction value based on the initial temperature and humidity data of the packaging environment and the current temperature and humidity data. The temperature and humidity compensation correction value is used to correct the impact of changes in ambient temperature and humidity on the decay rate of inert gas concentration and the rate of change of pressure difference. Since changes in ambient temperature may cause thermal expansion and contraction of gas inside the bag and change the pressure difference reading, and changes in ambient humidity may affect the flexibility of the bag film, the stress state of the sealing area, the sensor response or sampling stability, if no compensation is performed, normal environmental fluctuations may be misjudged as leakage trends, or the real micro-leakage signal may be masked by environmental changes.In this embodiment, the initial temperature and humidity at the time of packaging are used as a reference. The current ambient temperature and humidity are compared with the initial temperature and humidity, and a compensation correction value is calculated based on a pre-established temperature and humidity compensation relationship. The temperature and humidity compensation relationship can be obtained by calibration with a leak-free standard sample bag that is the same or similar to the target packaging bag in terms of specifications, membrane structure, sealing method, and inflation conditions. Alternatively, it can be updated based on stable sample data of packaging bags of the same specifications during historical testing. During actual testing, the data processing unit can first calculate the uncompensated inert gas concentration decay rate and pressure difference change rate, and then correct them using the temperature and humidity compensation correction value. When the gas composition change or pressure difference change mainly comes from non-leakage factors such as temperature rise and fall, and humidity fluctuation, the corresponding abnormality is reduced after compensation. When the environmental change is small but the gas composition and pressure difference still show abnormal changes, the corrected sealing state characteristic parameters can still retain leakage indication.

[0034] In this embodiment, after calculating the sealing state characteristic parameters of the inner bag, time-series fitting is performed on the sealing state characteristic parameters at multiple sampling times to generate the change trend curve of the inner bag sealing state. Based on the trend curve, the sealing status change of the inner bag within a preset time period is predicted by the time series prediction algorithm. When the prediction result reaches the micro-leakage warning threshold, an advanced warning signal is generated.

[0035] Specifically, firstly, the sealing state characteristic parameters of each group are arranged according to the sampling time sequence to form time-series data of the inner bag's sealing state. This time-series data is then time-series fitted to generate a trend curve of the inner bag's sealing state. The sealing state characteristic parameters can include the inert gas concentration decay rate, pressure difference change rate, temperature and humidity compensation correction values, and the compensated comprehensive sealing state parameters. The trend curve can be used to characterize the direction and rate of change of these parameters over time, thus reflecting whether the inner bag's sealing state remains stable, deteriorates slowly, or deteriorates rapidly. Before time-series fitting, outlier removal, smoothing, and time alignment processing can be performed on multiple groups of sealing state characteristic parameters to reduce the single-time... The impact of sampling fluctuations on the trend curve can be assessed through time-series fitting. Linear fitting, piecewise fitting, exponential smoothing, or other fitting methods suitable for continuous sampling data can be employed to ensure the generated trend curve continuously reflects the changing sealing state of the inner bag. After generating the trend curve, the data processing unit uses a time-series prediction algorithm to predict the sealing state changes of the inner bag within a preset future timeframe. This preset timeframe can be set based on the detection scenario, storage cycle, transfer cycle, or quality control requirements. The time-series prediction algorithm can employ autoregressive prediction, moving average prediction, exponential smoothing prediction, Kalman filtering prediction, or other algorithms capable of predicting future changes based on historical trends. In this embodiment, a quadratic exponential smoothing prediction algorithm is used to predict changes in sealing status. The data processing unit first performs quadratic exponential smoothing on the sealing status characteristic parameter values ​​of N consecutive sampling points on the trend curve to obtain a smoothed trend sequence. The process of quadratic exponential smoothing is as follows: first, the original sequence is smoothed once to obtain a smoothed sequence; then, the smoothed sequence is smoothed a second time to obtain a smoothed sequence. Then, based on the last value of the smoothed sequence and its changing trend, the predicted values ​​of the sealing status characteristic parameters for the next M sampling periods are predicted. The predicted values ​​are obtained by adding M times the trend increment to the current level value of the smoothed sequence. When any prediction period... When the predicted value of the sealing status characteristic parameter reaches or exceeds the preset micro-leakage warning threshold during the period, an advance warning signal is generated. Here, N is an integer not less than five, and M is a positive integer preset according to the needs of warehouse management. The micro-leakage warning threshold is preset according to the purity protection requirements of electronic-grade polysilicon and the sealing level of the inner bag of the packaging bag. During the prediction process, the slope of the trend curve, the duration of the abnormality, the fluctuation amplitude, and the consistency of changes among multiple sealing status characteristic parameters can be used as the basis for prediction to determine whether there is a continuous deterioration trend in the sealing status of the inner bag. When the prediction result shows that the sealing status of the inner bag will reach or exceed the preset micro-leakage warning threshold within a preset time period in the future, an advance warning signal is generated.

[0036] As a preferred embodiment of the above, in step S40, the construction step of the preset micro-leakage determination model includes: S41: Obtain historical sealing sample data for multiple sets of electronic-grade polycrystalline silicon packaging bags inner bags. The historical sealing sample data includes sealing status characteristic parameters and corresponding labels for leak-free samples, micro-leaking samples, and macro-leaking samples. The labels of the above historical sealing sample data are confirmed in the following ways: Leak-free samples are selected from packaging bags inner bags that have been confirmed to be leak-free by helium mass spectrometry leak detector and whose sealing status characteristic parameters remain stable within a preset long-term tracking period; micro-leaking samples are selected from packaging bags inner bags that have been confirmed to have micro-leaking channels by helium mass spectrometry leak detector and whose sealing status characteristic parameters show a slow and continuous deterioration trend; wherein, the equivalent leakage pore size range of micro-leaking samples measured by helium mass spectrometry leak detector is 0.1 micrometer to 10 micrometers; macro-leaking samples are selected from packaging bags inner bags that have been confirmed to have experienced sealing failure by visual inspection or obvious pressure decay. S42: Divide the historical sealed sample data into a training set and a validation set; S43: Supervised training of the initial neural network model is performed using the training set, and the accuracy of the trained model is verified and the parameters are optimized using the validation set to generate a preset micro-leakage judgment model.

[0037] Specifically, firstly, historical sealing sample data of multiple sets of inner bags in electronic-grade polycrystalline silicon packaging bags is acquired. This historical sealing sample data includes sealing state characteristic parameters and labels corresponding to different inner bags. Sealing state characteristic parameters may include inert gas concentration decay rate, pressure difference change rate, temperature and humidity compensation correction values, and corrected sealing state characteristic parameters. Labels include at least no-leakage samples, micro-leakage samples, and macro-leakage samples, used to characterize the actual sealing state of the corresponding samples. No-leakage samples represent a normal inner bag sealing state, micro-leakage samples represent slow leakage without significant failure, and macro-leakage samples represent significant sealing failure. After acquiring the historical sealing sample data, data cleaning, outlier removal, feature normalization, and label verification are performed to improve the effectiveness and consistency of the training data. Subsequently, the processed historical sealing sample data is divided into... The system uses a training set and a validation set, and supervises the training of an initial neural network model. The initial neural network model can be a multi-layer feedforward neural network, or a recurrent neural network or long short-term memory network suitable for time-series data processing. During training, the sealing state feature parameters from the training set are used as model input, and the corresponding labels are used as supervised output, allowing the model to learn the feature differences between no leakage, micro-leakage, and macro-leakage. After training, the accuracy of the trained model is verified using a validation set, and the model parameters are optimized based on the verification results. Parameter optimization may include adjusting the network structure, learning rate, training epochs, class weights, or input feature combinations. When the verification results reach the preset accuracy requirements, the optimized model is determined as the preset micro-leakage detection model. During actual detection, the data processing unit inputs the real-time calculated sealing state feature parameters into this model, and then outputs the micro-leakage detection result of the inner bag of the target packaging bag.

[0038] In this embodiment, in step S40, the sealing state characteristic parameters are input into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag, specifically including: The sealing status characteristic parameters after temperature and humidity compensation correction are input into the preset micro-leakage judgment model, and the output results are classified into three categories: no leakage, micro-leakage, and macro-leakage corresponding to the inner bag of the target packaging bag. When the classification result is micro-leakage, it is determined that there is micro-leakage in the inner bag of the target packaging bag, and the corresponding leakage level and leakage probability value are output at the same time.

[0039] Specifically, the data processing unit inputs the temperature and humidity compensated sealing status characteristic parameters into a preset micro-leakage judgment model to reduce the interference of ambient temperature and humidity changes on gas component detection and pressure difference detection. The temperature and humidity compensated sealing status characteristic parameters may include the corrected inert gas concentration decay rate, the corrected pressure difference change rate, and the temperature and humidity compensation correction value. After receiving the above parameters, the micro-leakage judgment model classifies and identifies the sealing status of the inner bag of the target packaging bag and outputs three classification results: no leakage, micro-leakage, and macro-leakage. Among them, no leakage indicates that the inner bag is in a normal sealing state, micro-leakage indicates that the inner bag has a slow leakage trend but has not yet formed an obvious failure, and macro-leakage indicates that the inner bag has already leaked. If obvious sealing failure occurs, the early micro-leakage state can be distinguished from the normal sealing state and obvious leakage state through three classification results, thereby improving the early identification capability. When the model outputs a classification result of micro-leakage, the data processing unit determines that there is a micro-leakage in the inner bag of the target packaging bag, and simultaneously outputs the corresponding leakage level and leakage probability value. The leakage probability value is used to represent the confidence level of the model in judging that the inner bag belongs to the micro-leakage state, and the leakage level is used to represent the degree of micro-leakage risk. It can be determined based on the leakage probability value, the degree of decay of the corrected inert gas concentration, and the degree of change of the corrected pressure difference. The leakage level can include low level, medium level, and high level, which are used to correspond to different re-inspection, isolation, or repackaging treatment strategies.

[0040] The method further includes the following steps after step S40: When the micro-leakage determination result indicates the presence of micro-leakage, the estimated range of micro-leakage aperture and real-time leakage rate of the inner bag are calculated based on the inert gas concentration decay rate, pressure difference change rate, combined with the inner bag volume and the volume parameters of the filled electronic-grade polysilicon. Based on the estimated range of micro-leakage aperture, real-time leakage rate, and purity protection threshold of electronic-grade polysilicon, corresponding disposal recommendations are generated.

[0041] In this embodiment, the calculation of the micro-leakage aperture prediction range adopts an estimation method based on the gas leakage rate equation. For the escape process of inert gas through the micro-leakage channel, the relationship between the real-time leakage rate Q and the micro-leakage aperture d satisfies: ; Where Q represents the real-time leakage rate in Pascals cubic meters per second; d represents the micro-leakage aperture in micrometers; P1 represents the internal air pressure of the inner bag, P0 represents the external ambient air pressure, and (P1-P0) is the real-time pressure difference between the inside and outside of the inner bag in Pascals; μ represents the dynamic viscosity of the inert gas filled into the inner bag in Pascals per second, which can be obtained from known physical property parameters according to the gas type; C is the flow coefficient related to the geometry of the leakage channel, and n is the flow state index. The range of C and n is determined according to the morphological characteristics of the leakage channel. Usually, C is between 0.01 and 100, and n is between 2 and 4.

[0042] The data processing unit calculates the micro-leakage aperture d in reverse using the real-time leakage rate Q, the real-time pressure difference between the inside and outside of the inner bag (P1-P0), and the known dynamic viscosity μ of the inert gas. The lower limit estimate of the solution is used as the minimum aperture and the upper limit estimate is used as the maximum aperture to form the micro-leakage aperture prediction range.

[0043] Specifically, the data processing unit retrieves the inert gas concentration decay rate and pressure difference change rate, and obtains the inner bag volume and the volume parameters of the filled electronic-grade polysilicon, thereby determining the remaining gas space volume inside the inner bag. The electronic-grade polysilicon volume parameters can be determined based on the filling mass, bulk density, or batch material parameters. Subsequently, based on the protective atmosphere loss trend reflected by the inert gas concentration decay rate and the internal and external gas exchange trend reflected by the pressure difference change rate, combined with the gas space volume inside the inner bag, the data processing unit calculates the real-time leakage rate of the inner bag. The real-time leakage rate can represent the amount of inert gas escaping or the amount of external gas entering per unit time, and can be based on... The trend of change over multiple consecutive sampling periods is determined to reduce the impact of fluctuations in a single sampling. Furthermore, the data processing unit, based on the real-time leakage rate, the pressure difference between the inside and outside of the inner bag, the gas type, the gas space volume, and the packaging bag structural parameters, calls a pre-established aperture estimation relationship to obtain the estimated range of the micro-leakage aperture. In this embodiment, the calculation of the estimated range of the micro-leakage aperture adopts an estimation method based on the gas leakage rate equation. For the escape process of inert gas through the micro-leakage channel, the real-time leakage rate and the micro-leakage aperture satisfy the following relationship: the real-time leakage rate is proportional to the nth power of the micro-leakage aperture, proportional to the pressure difference between the inside and outside of the inner bag, and proportional to the dynamic force of the inert gas. Viscosity is inversely proportional, where n is an exponent related to the shape of the leakage channel, ranging from two to four. The data processing unit calculates the micro-leakage pore size in reverse based on the currently calculated real-time leakage rate, the current real-time pressure difference between the inside and outside of the inner bag, and the known dynamic viscosity of the inert gas. The lower limit of the preset confidence interval of the solution is used as the minimum pore size, and the upper limit is used as the maximum pore size, forming a micro-leakage pore size prediction range. In other embodiments, estimation can also be performed by looking up a standard curve obtained from the calibration of sample packaging bags with known micropore sizes, or by inputting the real-time leakage rate and pressure difference data into a pre-trained pore size prediction model. This is used to characterize the leakage rate and pressure difference. The correlation between the variation and the leakage pore size is established because the actual leakage channel may be a sealing gap, a micropore in the membrane material, or an irregular crack. Therefore, the degree of micro-leakage is characterized by the estimated pore size range, rather than being limited to a single pore size value. Subsequently, the data processing unit generates disposal recommendation information based on the estimated micro-leakage pore size range, the real-time leakage rate, and the purity protection threshold of electronic-grade polysilicon. The purity protection threshold may include the minimum requirements for maintaining the protective atmosphere, the risk requirements for allowing water vapor or oxygen to enter, and the pollution control requirements corresponding to the product grade. The disposal recommendation information may include continued monitoring, shortening the retesting cycle, re-inspection, isolation and temporary storage, repackaging, prohibition of leaving the factory, or evaluation of scrapping.

[0044] In this embodiment, corresponding disposal suggestion information is generated, specifically including: The remaining safe storage time is calculated based on the estimated range of micro-leakage aperture, real-time leakage rate, and purity protection threshold of electronic-grade polysilicon. Based on the remaining safe storage time, generate tiered disposal recommendations.

[0045] Specifically, when a micro-leak is detected in the inner bag of the target packaging bag, the data processing unit combines the estimated micro-leakage aperture range, real-time leakage rate, remaining gas volume inside the inner bag, current inert gas concentration, and purity protection threshold to calculate the time the protective atmosphere of the inner bag can maintain the purity protection requirements for electronic-grade polysilicon under the current leakage state. The purity protection threshold may include the minimum inert gas holding concentration, allowable oxygen intake, allowable water vapor intake, or pollution risk limits corresponding to the product grade. The calculation can use the current leakage rate or a conservative calculation based on the higher risk side within the estimated aperture range. When continuous sampling... When the leakage rate changes, the remaining safe storage time can be dynamically updated based on the sampling results. Furthermore, the data processing unit generates tiered disposal recommendations based on the remaining safe storage time. These recommendations can be categorized into different risk levels according to the remaining safe storage time, and corresponding to different handling methods: when the remaining safe storage time is long, recommendations to continue monitoring or shorten the retesting cycle are generated; when the remaining safe storage time is in the middle range, recommendations to re-inspect, isolate and temporarily store, or prioritize handling are generated; when the remaining safe storage time is below the safety lower limit, recommendations to immediately repackage, prohibit shipment, stop transportation, or conduct a disposal assessment are generated. Example 2:

[0046] This invention also includes a micro-leakage early detection system for the inner bag of an electronic-grade polycrystalline silicon packaging bag, the system comprising: The initial baseline acquisition module is used to acquire the initial sealing baseline data of the inner bag of the packaged electronic-grade polysilicon target packaging bag. The initial sealing baseline data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. The real-time data acquisition module is used to collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data of the target packaging bag according to a preset sampling period. The feature parameter calculation module is used to calculate the sealing state feature parameters of the inner bag based on the initial sealing reference data, real-time gas composition data, real-time pressure difference data inside and outside the inner bag, and current ambient temperature and humidity data. The sealing state feature parameters include the inert gas concentration decay rate, pressure difference change rate, and temperature and humidity compensation correction value. The micro-leakage determination module is used to input the sealing state characteristic parameters into the preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag; The early warning signal generation module is used to generate a corresponding leakage early warning signal when the micro-leakage determination result indicates that a micro-leakage exists.

[0047] The adjustment system described above in this invention can effectively realize the method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags. The technical effects it can achieve are as described in the above embodiments, and will not be repeated here.

[0048] Similarly, the above-mentioned optimization schemes for the system can also achieve the optimization effects corresponding to the methods in Embodiment 1, which will not be repeated here.

[0049] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and accompanying drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for early detection of micro-leakage in the inner bag of an electronic-grade polycrystalline silicon packaging bag, characterized in that, The method includes: Obtain the initial sealing reference data of the inner bag of the electronic-grade polycrystalline silicon target packaging bag after the encapsulation is completed. The initial sealing reference data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. According to the preset sampling cycle, collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag of the target packaging bag, and current ambient temperature and humidity data. Based on the initial sealing reference data, the real-time gas composition data, the real-time pressure difference data inside and outside the inner bag, and the current ambient temperature and humidity data, the sealing state characteristic parameters of the inner bag are calculated. The sealing state characteristic parameters include the inert gas concentration decay rate, the pressure difference change rate, and the temperature and humidity compensation correction value. The sealing state characteristic parameters are input into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag; When the micro-leakage determination result indicates the presence of a micro-leakage, a corresponding leak warning signal is generated.

2. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, Based on the initial sealing reference data and the real-time collected data, the sealing state characteristic parameters of the inner bag are calculated, specifically including: Based on the initial inert gas component concentration and the real-time gas component data, the inert gas concentration decay rate per unit time is calculated. Based on the initial pressure difference between the inside and outside of the inner bag and the real-time pressure difference data between the inside and outside of the inner bag, the rate of change of pressure difference per unit time is calculated; Based on the initial temperature and humidity data of the packaging environment and the current temperature and humidity data, a temperature and humidity compensation correction value is calculated. The inert gas concentration decay rate and pressure difference change rate are corrected using the temperature and humidity compensation correction value to obtain the corrected sealing state characteristic parameters.

3. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, After calculating the sealing state characteristic parameters of the inner bag, time-series fitting is performed on the sealing state characteristic parameters at multiple different sampling times to generate the change trend curve of the inner bag sealing state. Based on the trend curve, the sealing status change of the inner bag within a preset time period is predicted by a time series prediction algorithm. When the prediction result reaches the micro-leakage warning threshold, an early warning signal is generated.

4. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, The steps for constructing the preset micro-leakage determination model include: Acquire historical sealing sample data of multiple sets of inner bags of electronic-grade polycrystalline silicon packaging bags. The historical sealing sample data includes the sealing state characteristic parameters and corresponding labels of no-leakage samples, micro-leakage samples and macro-leakage samples. The historical sealed sample data is divided into a training set and a validation set; The initial neural network model is trained under supervision using the training set, and the trained model is then validated and its parameters optimized using the validation set to generate the preset micro-leakage determination model.

5. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 4, characterized in that, The step of inputting the sealing state characteristic parameters into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag specifically includes: The sealing state characteristic parameters after temperature and humidity compensation correction are input into the preset micro-leakage judgment model, and the three classification results of no leakage, micro-leakage, and macro-leakage corresponding to the inner bag of the target packaging bag are output. When the classification result is micro-leakage, it is determined that there is micro-leakage in the inner bag of the target packaging bag, and the corresponding leakage level and leakage probability value are output.

6. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, The method further includes: When the microleakage determination result indicates the presence of microleakage, the estimated range of the microleakage aperture and the real-time leakage rate of the inner bag are calculated based on the inert gas concentration decay rate, the pressure difference change rate, the inner bag volume, and the electronic-grade polysilicon volume parameters. Based on the estimated range of the micro-leakage aperture, the real-time leakage rate, and the purity protection threshold of electronic-grade polysilicon, corresponding disposal recommendations are generated.

7. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 6, characterized in that, The generation of corresponding handling suggestion information specifically includes: The remaining safe storage time is calculated based on the estimated range of the micro-leakage aperture, the real-time leakage rate, and the purity protection threshold of electronic-grade polysilicon. Based on the remaining safe storage time, tiered disposal recommendations are generated.

8. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, The preset sampling period is a dynamically adjustable period, specifically including: Based on the fluctuation range of the current ambient temperature and humidity data, the changing trend of the sealing state characteristic parameters, and the storage and transportation stage of the target packaging bag, the duration and sampling accuracy of the preset sampling period are dynamically adjusted.

9. The method for early detection of micro-leakage in the inner bag of electronic-grade polycrystalline silicon packaging bags according to claim 1, characterized in that, The acquisition of initial sealing reference data for the inner bag of the packaged electronic-grade polysilicon target packaging bag specifically includes: After the electronic-grade polysilicon filling, inert gas encapsulation, and heat sealing of the inner bag of the target packaging bag are completed and a preset standing time is reached, the initial inert gas component concentration inside the inner bag and the initial pressure difference between the inside and outside of the inner bag are collected. At the same time, the initial environmental temperature and humidity data at the sealing point are collected to generate the initial sealing reference data and store it in association with the database corresponding to the unique identifier of the target packaging bag.

10. A micro-leakage early detection system for the inner bag of an electronic-grade polycrystalline silicon packaging bag, characterized in that, The system includes: The initial reference acquisition module is used to acquire the initial sealing reference data of the inner bag of the packaged electronic-grade polycrystalline silicon target packaging bag. The initial sealing reference data includes the initial inert gas component concentration inside the inner bag, the initial pressure difference between the inside and outside of the inner bag, and the initial temperature and humidity data of the packaging environment. The real-time data acquisition module is used to collect real-time gas composition data, real-time pressure difference data between the inside and outside of the inner bag, and current ambient temperature and humidity data of the target packaging bag according to a preset sampling period. The feature parameter calculation module is used to calculate the sealing state feature parameters of the inner bag based on the initial sealing reference data, the real-time gas composition data collected in real time, the real-time pressure difference data inside and outside the inner bag, and the current ambient temperature and humidity data. The sealing state feature parameters include the inert gas concentration decay rate, the pressure difference change rate, and the temperature and humidity compensation correction value. The micro-leakage determination module is used to input the sealing state characteristic parameters into a preset micro-leakage determination model to obtain the micro-leakage determination result of the inner bag of the target packaging bag; The early warning signal generation module is used to generate a corresponding leakage early warning signal when the micro-leakage determination result indicates the presence of a micro-leakage.

Images