Sealing integrity detection system and method for sealed bottle

Abstract

The invention provides a sealing integrity detection system and method for a sealed bottle, and the method comprises the steps: transmitting a first laser beam to a collimator through an optical fiber after a controller receives a trigger signal; the collimator corrects the first laser beam and emits an obtained second laser beam to the to-be-detected sealed bottle so as to transmit a laser signal to the signal receiving unit through the to-be-detected sealed bottle; the signal receiving unit outputs a first signal according to the laser signal and sends the first signal to the controller so as to output a second signal through the controller and send the second signal to the upper computer; the upper computer calculates the first gas content in the sealed bottle to be detected according to the second signal, compares the first gas content with the preset standard gas content, and determines the sealing integrity detection result of the sealed bottle to be detected according to the obtained comparison result. According to the system, the sealing integrity of the to-be-detected sealing bottle can be detected online without damaging the to-be-detected sealing bottle, and full detection of sealing bottle products produced by a production line can be realized.

Description

Technical Field

[0001] This invention relates to the field of seal integrity testing technology, and in particular to a seal integrity testing system and method for sealed bottles. Background Technology

[0002] Currently, for pharmaceutical packaging vials or ampoules, the sealing performance of vials or ampoules requiring nitrogen filling or vacuum packaging is typically tested using a destructive method. This involves breaking the packaging, extracting the gas, and then introducing the gas into a sensor's measuring chamber for spectral or physicochemical measurements to analyze its composition. By comparing the gas composition content with that of a standard sample, it is determined whether a leak has occurred in the pharmaceutical packaging. This method requires destroying the pharmaceutical packaging to measure the gas inside, and destructive measurement methods can only be used for sampling inspections. Furthermore, the gas extraction and measurement process is time-consuming, making it difficult to perform full inspection of products produced on the production line. Summary of the Invention

[0003] The purpose of this invention is to provide a sealing integrity testing system and method for sealed bottles, so as to complete the online testing of the sealing integrity of sealed bottles and realize full inspection of sealed bottle products produced on the production line.

[0004] This invention provides a sealing integrity testing system for a sealed bottle. The system includes a controller, a collimator, a signal receiving unit, and a host computer. The controller, upon receiving a trigger signal, emits a first laser beam and transmits it to the collimator via an optical fiber. The collimator calibrates the first laser beam and directs the calibrated second laser beam toward the sealed bottle under test, transmitting a laser signal through the bottle to the signal receiving unit. The signal receiving unit outputs a first signal based on the laser signal and sends it to the controller. The controller processes the first signal, outputs a second signal, and sends the second signal to the host computer. The host computer calculates the content of a first gas in the sealed bottle based on the second signal, compares the first gas content with a preset standard gas content to obtain a comparison result, and determines the sealing integrity testing result of the sealed bottle based on the comparison result.

[0005] Furthermore, the controller includes: a trigger sensor, a signal conditioning module, and a laser connected in sequence via communication; the laser is connected to a collimator via an optical fiber; the trigger sensor is used to output a first control signal to the signal conditioning module after receiving a trigger signal; the signal conditioning module is used to output a laser emission signal according to the first control signal; and the laser is used to emit a first laser beam according to the laser emission signal.

[0006] Furthermore, the controller also includes: a signal amplifier and a communication module; the signal amplifier is communicatively connected to the signal receiving unit; the communication module is communicatively connected to the host computer; the signal amplifier is used to amplify the first signal to obtain the second signal, and the second signal is sent to the host computer through the communication module.

[0007] Furthermore, the host computer is also used to: determine that the sealed bottle under test has a leak if the comparison result indicates that the deviation between the content of the first gas and the content of the standard gas exceeds the preset threshold range; and determine that the sealed bottle under test does not have a leak if the comparison result indicates that the deviation between the content of the first gas and the content of the standard gas does not exceed the preset threshold range.

[0008] Furthermore, the host computer is used to calculate the difference between the second signal and the signal value corresponding to the second laser beam to obtain the laser attenuation value, and to confirm the content of the first gas in the sealed bottle to be tested through the laser attenuation value.

[0009] Furthermore, during the detection process of a single sealed bottle to be tested, multiple second signals are continuously acquired within a preset test time window to obtain multiple laser attenuation values, forming an attenuation value sequence; the attenuation value sequence is analyzed to identify invalid data segments whose attenuation values ​​are distorted due to the sealed bottle itself; the invalid data segments are removed, and the first gas content is calculated based on the remaining valid attenuation value sequence.

[0010] Furthermore, the analysis of the attenuation value sequence to identify invalid data segments where the attenuation value is distorted due to the nature of the sealed bottle itself includes: using a time-series decomposition algorithm to decompose the attenuation value sequence into a trend component sequence and a residual component sequence; calculating the first-order difference sequence of the residual component sequence; comparing the absolute value of the difference sequence with a dynamic threshold preset based on the overall data volatility; and identifying the attenuation value data points corresponding to the difference values ​​exceeding the threshold as distortion points caused by large optical path loss due to poor uniformity of the bottle surface shape. And / or, The attenuation value sequence is compared with the attenuation value range predicted by the model based on the normal passage of the sealed bottle under test; attenuation value data points that exceed the predicted range are identified as invalid data points caused by severe deflection or occlusion of the bottle.

[0011] Furthermore, the attenuation range predicted by the model based on the normal passage of the sealed bottle under test is an absolute boundary range determined based on the physical principle of optical transmission. The lower limit of the absolute boundary range is the theoretical minimum possible attenuation value, which corresponds to the case where the laser beam is not absorbed by the gas under test at all. The upper limit of the absolute boundary range is the theoretical maximum possible attenuation value, which corresponds to the theoretical limit of light intensity attenuation caused by factors such as absorption and reflection of the bottle body and the medium inside the bottle when the laser beam passes through the sealed bottle to be tested. The invalid data point refers to a data point whose attenuation value is lower than the minimum possible attenuation value or higher than the maximum possible attenuation value.

[0012] Furthermore, calculating the first gas content based on the remaining effective attenuation value sequence includes: Calculate a corresponding mass weight coefficient for each effective attenuation value in the remaining effective attenuation value sequence; Based on the quality weighting coefficient, the effective attenuation value is weighted and calculated to determine a final characteristic attenuation value; The content of the first gas is calculated based on the characteristic attenuation value; The calculation of the quality weighting coefficient includes at least one of the following methods: The stability weight factor is calculated based on the stability of the data points within the local time window where the effective decay value is located. The proximity weight factor is calculated based on the degree of deviation between the effective decay value and the statistical center value of the entire effective decay value sequence.

[0013] Furthermore, the system also includes a sorting module; the sorting module is used to: when the sealed bottle to be tested is moved to the preset rejection position, if the sealed bottle to be tested has a leak, reject the sealed bottle to be tested according to the preset rejection method; if the sealed bottle to be tested does not have a leak, allow the sealed bottle to be tested to pass.

[0014] Furthermore, the preset rejection methods include: blowing method, hitting method, or guiding diversion method.

[0015] Furthermore, the system also includes a display module; the display module is used to display the seal integrity test results of the sealed bottle to be tested.

[0016] This invention provides a method for detecting the seal integrity of a sealed bottle. The method includes: after receiving a trigger signal, a controller emits a first laser beam and transmits the first laser beam to a collimator via an optical fiber; the collimator calibrates the first laser beam and directs the calibrated second laser beam toward the sealed bottle to be tested, so as to transmit a laser signal through the sealed bottle to a signal receiving unit; the signal receiving unit outputs a first signal based on the laser signal and sends the first signal to the controller; the controller processes the first signal and outputs a second signal, which is then sent to a host computer; the host computer calculates the content of a first gas in the sealed bottle based on the second signal, compares the first gas content with a preset standard gas content to obtain a comparison result, and determines the seal integrity detection result of the sealed bottle based on the comparison result.

[0017] The present invention provides a sealing integrity detection system and method for sealed bottles. A controller, upon receiving a trigger signal, emits a first laser beam, which is transmitted to a collimator via an optical fiber. The collimator calibrates the first laser beam and directs the calibrated second laser beam toward the sealed bottle under test, transmitting a laser signal through the bottle to a signal receiving unit. The signal receiving unit outputs a first signal based on the laser signal and sends it to the controller. The controller processes the first signal, outputs a second signal, and sends the second signal to a host computer. The host computer calculates the content of a first gas in the sealed bottle based on the second signal, compares the first gas content with a preset standard gas content, obtains the comparison result, and determines the sealing integrity detection result of the sealed bottle based on the comparison result. This system can perform online detection of the sealing integrity of the sealed bottle under test without damaging it, and can achieve full inspection of sealed bottle products produced on the production line. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 A schematic diagram of a seal integrity detection system for a sealed bottle provided in an embodiment of the present invention; Figure 2 A schematic diagram of another sealing integrity detection system for a sealed bottle provided in an embodiment of the present invention; Figure 3 This is a flowchart of a method for detecting the seal integrity of a sealed bottle, provided as an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Existing technologies for testing the seal integrity of vials or ampoules requiring nitrogen filling or vacuum packaging necessitate damaging the packaging to measure the internal gas. Furthermore, destructive testing methods are limited to sampling inspections and cannot achieve 100% complete inspection. Additionally, the time required for gas extraction and measurement restricts the feasibility of online full-scale product inspection. Therefore, this invention provides a system and method for detecting the seal integrity of sealed bottles. This technology can be applied to applications requiring the detection of the seal integrity of sealed bottles.

[0022] To facilitate understanding of this embodiment, a sealing integrity detection system for a sealed bottle disclosed in this embodiment of the invention will first be introduced, such as... Figure 1 As shown, the system includes: a controller 10, a collimator 11, a signal receiving unit 12, and a host computer 13; The controller 10, upon receiving a trigger signal, emits a first laser beam and transmits it to the collimator 11 via optical fiber. The collimator 11 calibrates the first laser beam and directs the calibrated second laser beam toward the sealed bottle under test, transmitting a laser signal through the bottle to the signal receiving unit 12. The trigger signal can indicate that the sealed bottle under test has reached a preset designated position. The sealed bottle under test can be a vial or ampoule containing medicine. In practice, when the controller 10 receives the trigger signal, it can determine that the sealed bottle under test has reached the designated position, at which point it can control the emission of the first laser beam, which is then transmitted to the collimator 11 via optical fiber. The collimator 11 is an optical element used for input and output in optical fiber communication devices, capable of transmitting optical fiber signals. The divergent light is transformed into parallel light by a front-mounted convex lens, thereby correcting the first laser beam. The corrected second laser beam is then directed at the sealed bottle under test. In practical applications, to ensure consistency of results, the detection position of the sealed bottle under test is usually fixed. That is, for different sealed bottles under test, the second laser beam is directed at the same detection position. For example, if the sealed bottle under test is a vial containing medicine, the detection position can be the area above the medicine inside the vial. Since the detection position is not blocked by the medicine, the second laser beam can pass through the sealed bottle under test. The gas inside the sealed bottle under test usually absorbs or reflects the second laser beam, causing a change in the intensity of the second laser beam, thus obtaining a laser signal. The obtained laser signal can be transmitted to the signal receiving unit 12.

[0023] The signal receiving unit 12 is used to output a first signal based on the laser signal, send the first signal to the controller 10, process the first signal through the controller 10, output a second signal, and send the second signal to the host computer 13. In actual implementation, the signal receiving unit 12 can convert the received laser signal into a first signal, which is usually an analog electrical signal, and send the first signal to the controller 10. The controller 10 can demodulate, amplify, and process the first signal, output a second signal, and send the second signal to the host computer 13.

[0024] The host computer 13 is used to calculate the content of the first gas in the sealed bottle to be tested based on the second signal, compare the content of the first gas with the preset standard gas content, obtain the comparison result, and determine the sealing integrity test result of the sealed bottle to be tested based on the comparison result. The aforementioned first gas content may include the content of only one gas, or it may include the content of multiple gases. For example, if the sealed bottle to be tested is filled with nitrogen and the sealed bottle is intact, then the first gas content may be the nitrogen content. If the sealed bottle to be tested leaks, then the first gas content may include the nitrogen content, oxygen content, etc. The aforementioned standard gas content is usually a predetermined gas content value, representing the gas content inside the sealed bottle to be tested when it is intact. In actual implementation, the host computer 13 can deduce the first gas content in the sealed bottle to be tested based on the received second signal and compare it with the preset standard gas content. The integrity of the seal of the sealed bottle to be tested can be evaluated based on the comparison result. For example, the deviation value between the first gas content and the corresponding standard gas content can be calculated. If the deviation value exceeds the preset threshold range, it can be considered that the sealed bottle to be tested is leaking. At this time, it can continue to check whether the sealed bottle to be tested is damaged, etc.

[0025] The aforementioned sealing integrity detection system for sealed bottles includes a controller 10 that, upon receiving a trigger signal, emits a first laser beam, which is transmitted via optical fiber to a collimator 11. The collimator 11 calibrates the first laser beam and directs the calibrated second laser beam toward the sealed bottle under test, transmitting a laser signal through the bottle to a signal receiving unit 12. The signal receiving unit 12 outputs a first signal based on the laser signal and sends it to the controller 10. The controller 10 processes the first signal, outputs a second signal, and sends it to a host computer 13. The host computer 13 calculates the content of a first gas in the sealed bottle based on the second signal, compares the first gas content with a preset standard gas content, obtains the comparison result, and determines the sealing integrity detection result of the sealed bottle based on the comparison result. This system can perform online detection of the sealing integrity of the sealed bottle under test without damaging it, and can achieve full inspection of sealed bottle products produced on the production line.

[0026] Furthermore, such as Figure 2 The schematic diagram shows another sealing integrity detection system for a sealed bottle. The controller 10 includes: a trigger sensor 101, a signal conditioning module 102, and a laser 103, which are connected in sequence via optical fiber. The laser 103 is connected in communication with a collimator 11 via optical fiber. The trigger sensor 101 is used to output a first control signal to the signal conditioning module 102 after receiving a trigger signal. The signal conditioning module 102 is used to output a laser emission signal according to the first control signal. The laser 103 is used to emit a first laser beam according to the laser emission signal.

[0027] In practical implementation, the trigger sensor 101 can select a suitable triggering method according to actual needs. For example, it can use a mechanical triggering method (e.g., the sealed bottle under test triggers a limit switch), a photoelectric induction triggering method (e.g., the sealed bottle under test blocks light), or an internal program triggering method (e.g., the logic controller 10 automatically triggers the test by controlling the position of the sealed bottle under test). In practical implementation, to ensure the accuracy of the detection results, the trigger sensor 101 can receive the above-mentioned triggering signals. For example, when the sealed bottle under test triggers a limit switch, it sends a triggering signal, which the trigger sensor 101 can collect. At this time, it can output a first control signal to the signal conditioning module 102. The signal conditioning module 102 can generate a laser emission signal according to the first control signal. This laser emission signal can be used to instruct the laser 103 to emit a first laser beam. After receiving the laser emission signal, the laser 103 can emit a first laser beam. Typically, the laser 103 can use a tunable function to emit a narrow-band first laser beam directed at the detection position of the sealed bottle under test.

[0028] Furthermore, such as Figure 2 As shown, the controller 10 also includes a signal amplifier 104 and a communication module 105; the signal amplifier 104 is communicatively connected to the signal receiving unit 12; the communication module 105 is communicatively connected to the host computer 13; the signal amplifier 104 is used to amplify the first signal to obtain the second signal, and the second signal is sent to the host computer 13 through the communication module 105.

[0029] The aforementioned signal amplifier 104 can receive the first signal from the signal receiving unit 12, amplify the first signal to improve communication quality, and transmit the obtained second signal to the host computer 13 through the communication module 105.

[0030] Furthermore, the host computer is used to calculate the difference between the second signal and the signal value corresponding to the second laser beam to obtain the laser attenuation value, and to confirm the content of the first gas in the sealed bottle to be tested through the laser attenuation value.

[0031] The host computer calculates the content of the first gas in the sealed bottle based on the received second signal. Its core function lies in accurately retrieving the gas concentration (in this embodiment, the first gas refers to oxygen) by analyzing the laser attenuation value. The specific implementation is as follows: S1: Calculation of laser attenuation value The host computer pre-stores the "signal value corresponding to the second laser beam," which serves as a reference (denoted as I_ref) and represents the theoretical light intensity of the laser before absorption by the gas inside the sealed bottle being tested. This value can be measured and stored during the system calibration phase by passing the laser through a standard bottle filled with nitrogen or known to contain no gas to be tested.

[0032] For each acquired second signal (which can be denoted as I_measure), the host computer calculates its laser attenuation value using the following formula: Laser attenuation value = I_ref - I_measure According to the Lambert-Beer law, this decay value is directly proportional to the concentration of the gas being measured.

[0033] S2: Construction of decay value sequences and identification of invalid data To improve the robustness of online detection, the host computer executes the following intelligent processing flow for a single sealed bottle to be tested: Sequence Construction: Within a preset test time window (e.g., 500 milliseconds), multiple second signals are continuously acquired (e.g., 1000 points per second), and the corresponding laser attenuation values ​​are calculated. These attenuation values ​​are arranged in chronological order to form an attenuation value sequence.

[0034] Invalid Data Identification: Due to production line vibration, the bottles may experience slight shaking or deflection, or the bottles themselves may have poor uniformity in local surface shape at the microscopic level, leading to severe distortion of individual attenuation values. The host computer uses one or more of the following combined strategies to identify and remove these invalid data segments: A: Instantaneous distortion identification based on rate of change Because the bottle itself may exhibit macroscopic, slow shape changes (such as a cylindrical bottle), directly performing anomaly detection on the original attenuation value sequence would misjudge these normal overall trends as local defects. Therefore, the core of this implementation lies in separating macroscopic trends from local defects, thereby accurately identifying distortion points caused by the true microscopic level of local surface shape uniformity.

[0035] First, a time-series decomposition algorithm is used to decompose the collected original decay value sequence into two main components: a trend component sequence and a residual component sequence. Locally weighted regression (LOESS) or seasonal trend decomposition (STL) algorithms are then used to effectively fit the nonlinear trend in the data without pre-setting a specific trend model, thus decomposing the original decay value sequence into two main components: a trend component sequence and a residual component sequence.

[0036] The trend component represents the baseline of optical path loss formed by the slow change of the overall shape of the bottle; while the residual component reflects the abnormal fluctuations caused by local and minute unevenness of the surface. These fluctuations are directly caused by the poor shape uniformity of local areas of the bottle surface (such as depressions, protrusions, scratches).

[0037] It should be noted that in this application, the purpose of using a time-series decomposition algorithm (such as Local Weighted Regression LOESS) is to effectively separate the macroscopic shape trend of the bottle from the microscopic surface defect signal. In specific applications, the algorithm takes the original attenuation value sequence as input and analyzes it point by point through a preset smoothing window. For each data point, the algorithm examines the weighted average of all points within its neighboring window to fit a smooth, slowly changing trend line, which is the trend component sequence. This trend component reflects the baseline of optical path loss caused by the overall, slow shape changes (such as curvature) of the bottle. Subtracting this trend component sequence from the original attenuation value sequence yields the residual component sequence. The residual sequence contains local, sharp fluctuations in the original signal that cannot be explained by the smoothing trend. These fluctuations are caused by the shape inhomogeneities (such as depressions, protrusions, or scratches) on the bottle surface along the laser scanning path. When the beam scans to these defect points, the surface smoothness is disrupted, causing abrupt optical path loss, thus forming significant peaks or abrupt changes in the residual sequence. This decomposition process successfully decouples the complex detection signal based on its physical origin. Subsequent distortion point identification algorithms can then precisely focus on analyzing the residual sequence, effectively avoiding misjudging normal overall shape changes of the bottle as defects, thus significantly improving the accuracy and reliability of the detection.

[0038] Next, the residual component sequence is differentially analyzed to quantify the severity of its local variations; the first-order difference sequence is calculated: the first-order difference of the residual component sequence is calculated as difference value(i) = residual(i) - residual(i-1). This difference sequence visually reflects the rate of change of anomalous fluctuations between adjacent detection points. A sudden surface defect will produce a significant peak in this difference sequence.

[0039] Take the absolute value: Calculate the absolute value of the difference sequence to uniformly handle drastic changes in the positive direction (such as from a flat surface to a convex surface) and the negative direction (such as from a flat surface to a concave surface).

[0040] Finally, an adaptive dynamic threshold is used to determine the distortion point.

[0041] Dynamic threshold preset: The dynamic threshold is not a fixed value, but is preset based on the overall volatility of the residual component sequence. A preferred implementation uses the median absolute deviation (MAD) to measure volatility. The dynamic threshold can be set as Threshold = k*MAD, where k is an adjustable sensitivity coefficient (typically between 2.5 and 3.5). MAD is insensitive to outliers, making the threshold itself more robust. This design allows the threshold to adapt to the inherent signal volatility levels of different bottle types or production lines.

[0042] Distortion point identification: The absolute value of the difference sequence is compared with the dynamic threshold mentioned above. The original attenuation data points corresponding to the difference values ​​whose absolute values ​​exceed the dynamic threshold are identified as distortion points caused by large optical path losses due to poor uniformity of the bottle surface shape.

[0043] B: Absolute boundary recognition based on physical principles Determine the absolute boundary range: Based on the physical principles of optical transmission, determine a theoretical absolute boundary range.

[0044] Lower limit: The theoretical minimum possible attenuation value. When the laser is completely unabsorbed by the gas being measured, the attenuation value should approach zero (or a small positive value determined by background gas absorption). Therefore, the lower limit can be set to 0 or a value slightly greater than 0 (e.g., 0.05). Any calculated attenuation value below this lower limit is physically impossible and is considered invalid.

[0045] Upper limit: The theoretical maximum possible attenuation value. It is determined by the maximum theoretical attenuation the laser beam may encounter when passing through the sealed bottle under test. This takes into account the reflection and absorption of the laser by the bottle wall, the scattering by liquids or other media that may be present inside the bottle, and the saturation limit of the absorption capacity of the gas being tested itself. Any calculated attenuation value higher than this upper limit indicates that it is not caused by normal gas absorption, but is most likely due to severe deflection of the bottle completely blocking the light path or interference from adjacent objects, and is therefore determined to be an invalid data point.

[0046] S3: Calculation of effective gas content After identifying invalid data points, the host computer executes: Data removal: Remove all invalid data points identified in the above steps from the decay value sequence.

[0047] Content calculation: Calculate the arithmetic mean of the remaining effective attenuation value sequence to obtain a characteristic attenuation value that can robustly reflect the gas absorption in the bottle.

[0048] Concentration inversion: Based on the pre-calibrated "characteristic decay value - gas content" relationship curve (which is based on Lambert-Beer's law), the above characteristic decay value is converted into the first gas content in the sealed bottle to be tested.

[0049] Furthermore, after obtaining the sequence of remaining valid attenuation values ​​after removing invalid data, the host computer calculates the first gas content in the sealed bottle to be tested based on this sequence. In addition to using conventional methods of calculating the arithmetic mean or median, to further improve the accuracy and robustness of the detection results, this invention also provides an optimization algorithm based on signal quality weighting.

[0050] The core idea of ​​this algorithm is to recognize that the "quality" of each valid data point in the sequence is not equal. Different weights should be assigned to each data point based on the measurement stability and the reasonableness of its value at the time of measurement, so that more reliable data points will have a higher proportion in the final result.

[0051] The specific implementation method includes the following steps: W1: Mass weighting coefficient for calculating effective attenuation value Calculate a corresponding quality weight coefficient for each effective decay value in the remaining effective decay value sequence. This coefficient can be calculated by combining one or more of the following factors: Calculation of stability weighting factor: Objective: To evaluate the stability of the signal at a local time point. The smaller the vibration of the bottle, the smoother the local signal fluctuation at that point should be, and the higher its quality.

[0052] Operation: Centered on a data point, take a local time window encompassing N points before and after it (e.g., N=10). Calculate the standard deviation of all decay values ​​within this window. Assignment: The stability weighting factor is inversely proportional to the standard deviation.

[0053] One specific calculation method is as follows: W_{s_i} = 1 / (1 + σ_i) This means that the smaller the local standard deviation (i.e., the more stable), the larger the stability weight factor at that point.

[0054] Calculation of proximity weight factor ((W_{p_i} \)): Objective: To evaluate the degree to which the values ​​of data points deviate from the main distribution of the entire effective sequence. Data points that deviate from the mainstream distribution, even if they are stable themselves, may have poor representativeness and therefore their weight should be reduced.

[0055] Operation: First, calculate the statistical center of the entire remaining effective decay value sequence, such as the median of the sequence (the median is not sensitive to outliers and is better than the mean).

[0056] Assignment: Calculate the absolute deviation of this data point from the median. The proximity weighting factor is inversely proportional to this absolute deviation. One specific calculation method is as follows: W_{p_i} = 1 / (1 + |A_i - A_{median}|) This means that the closer the decay value of a point is to the median of the entire sequence, the larger its proximity weight factor.

[0057] Finally, the overall quality weight coefficient of a data point can be obtained by multiplying its stability weight factor and proximity weight factor: W_i = W_{s_i} * W_{p_i} This multiplicative relationship ensures that only data points that simultaneously satisfy "local stability" and "global rationality" can obtain high weights.

[0058] W2: Weighted calculation to determine characteristic attenuation value After obtaining the mass weighting coefficient for each data point, the final characteristic attenuation value used to calculate the gas concentration is determined using one of the following methods: Method 1: Weighted Average Method Calculate the weighted average of all effective attenuation values, with their mass weights as coefficients: Characteristic attenuation value = ( Σ (A_i * W_i) ) / ( Σ W_i ) This method makes full use of the information from all valid data, but suppresses the contribution of low-quality data through weighting.

[0059] Method 2: Maximum Weight Method Identify the data point with the largest quality weight coefficient in the entire effective attenuation value sequence. Directly use the attenuation value corresponding to this point as the feature attenuation value.

[0060] This method assumes that the measurement value at the moment when the measurement conditions are most ideal and the data quality is highest within the entire detection time window is the most reliable representative. This method is computationally efficient and avoids the signal smoothing that averaging may cause.

[0061] W3: Calculate the content of the first gas By substituting the characteristic attenuation value obtained from the above steps into the pre-calibrated "attenuation value - gas content" conversion formula or lookup table, the first gas content in the sealed bottle to be tested can be obtained.

[0062] In practical applications, different gas molecules exhibit specific spectral absorption characteristics. For example, taking oxygen as an example, when a laser beam modulated by a pulse of a certain frequency irradiates oxygen molecules, it will produce obvious reflection or absorption of the laser at that specific frequency. By analyzing the intensity of the laser light at that specific frequency, the amount of oxygen present can be determined.

[0063] Typical spectral processing may include the following steps: baseline correction, noise reduction, normalization, standard normal variable transformation, derivative spectral processing, and deconvolution. Each step is briefly explained below: 1. Baseline Correction: Removes baseline shifts or drifts caused by instruments, sample containers, or other factors to ensure that the spectrum reflects the characteristics of the sample itself.

[0064] 2. Noise Reduction: Various filtering techniques, such as Moving Average, Savitzky-Golay filtering, and Wavelet Transform, are used to reduce random noise in spectral data.

[0065] 3. Normalization: Adjusting the intensity ratio of the spectrum to the same level helps in comparing data. Common methods include vector normalization, maximum normalization, norm normalization, etc.

[0066] 4. Standard Normal Variate (SNV): Used to correct scattering effects and perform spectral correction.

[0067] 5. Derivative Spectroscopy: By calculating the first, second, or higher derivatives, the details of the spectral curve are highlighted, which helps to separate overlapping peaks and eliminate the effects of baseline drift.

[0068] 6. Deconvolution: A mathematical process used to separate overlapping signals in a spectrum to obtain clearer peak information.

[0069] Furthermore, the host computer 13 is also used to: determine that the sealed bottle under test has a leak if the comparison result indicates that the deviation between the content of the first gas and the content of the standard gas exceeds the preset threshold range; and determine that the sealed bottle under test does not have a leak if the comparison result indicates that the deviation between the content of the first gas and the content of the standard gas does not exceed the preset threshold range.

[0070] The aforementioned preset threshold range can be set according to actual needs. In actual implementation, the host computer 13 can calculate the deviation between the content of the first gas and the content of the standard gas. If the deviation exceeds the preset threshold range, for example, the nitrogen content is much lower than the corresponding standard nitrogen content, it can be considered that the sealed bottle under test has a leak, and it is necessary to check whether the sealed bottle under test is broken, etc. If the deviation does not exceed the preset threshold range, for example, the nitrogen content is slightly lower than the corresponding standard nitrogen content, but within the allowable range, it can be considered that the sealed bottle under test has no leak, and the sealing integrity of the sealed bottle under test meets the requirements.

[0071] Furthermore, the system also includes a sorting module; the sorting module is used to: when the sealed bottle to be tested is moved to the preset rejection position, if the sealed bottle to be tested has a leak, reject the sealed bottle to be tested according to the preset rejection method; if the sealed bottle to be tested does not have a leak, allow the sealed bottle to be tested to pass.

[0072] The aforementioned rejection location can be understood as the location on the production line where defective products are rejected; the aforementioned preset rejection methods include: air blowing, impact, or guide diversion; wherein, the air blowing method can be understood as rejecting defective products by blowing gas; the impact method can be understood as rejecting defective products by using a preset impactor; the aforementioned guide diversion method can be understood as diverting defective products to another production line branch to separate defective products from qualified products. In practice, after full inspection of products, the rejection of defective products is one of the important means to fully utilize the function of online full inspection. In actual application, the sealed bottles to be tested on the production line pass through the inspection positions in sequence on the assembly line. Before the designated sealed bottle to be tested reaches the rejection position, the seal integrity test result corresponding to the designated sealed bottle to be tested is calculated and stored in advance. It should be noted that the seal integrity test result of each sealed bottle to be tested on the production line is stored sequentially according to the order of product arrival. When the designated sealed bottle to be tested reaches the rejection position, the sorting module can read the seal integrity test result of the designated sealed bottle to be tested. If the seal integrity test result indicates that the designated sealed bottle to be tested has a leak, it can be automatically rejected according to the preset rejection method. If the seal integrity test result indicates that the designated sealed bottle to be tested has no leak, it can be released normally.

[0073] Typically, the seal integrity test results of the sealed bottle under test are recorded in a database. When the seal integrity test result indicates that the sealed bottle under test has no leakage, it can be represented by "OK"; when the seal integrity test result indicates that the sealed bottle under test has leakage, it can be represented by "NG". The seal integrity test results stored in the database usually need to meet the requirements of data integrity, data uniqueness and immutability.

[0074] Furthermore, the system also includes a display module 14; the display module 14 is used to display the seal integrity test results of the sealed bottle under test. For example... Figure 2 As shown, the display module 14 is usually connected to the host computer 13. The display module 14 can be a display screen, etc. In actual implementation, after the host computer 13 determines the sealing integrity test result of the sealed bottle to be tested, it can display the sealing integrity test result through the display module 14, which makes it easier for users to view and take corresponding measures based on the sealing integrity test result.

[0075] In addition, periodic calibration of the above-mentioned seal integrity testing system is a necessary way to ensure the accuracy of the test. Standard samples with different standard gas contents are continuously passed through the test position to measure the measured values ​​of different gas contents. The measured values ​​of different standard gas contents and the actual values ​​are used to fit the parameters to achieve the minimum deviation between the measured values ​​and the actual values. The specific fitting method can be referred to relevant technologies, which will not be elaborated here.

[0076] The aforementioned seal integrity testing system for sealed bottles uses a programmable laser beam to pass through the detection location of the sealed bottle under test. It detects the absorption and reflection of the gas spectrum within the packaging to measure the gas composition. Based on the difference in gas composition content compared to a standard sample, it determines whether a leak has occurred. This system, through a non-destructive gas spectroscopy laser rapid detection method, can achieve 100% online measurement of the seal integrity of nitrogen-filled or vacuum-packed pharmaceuticals, avoiding the need for comprehensive product quality testing that sampling methods cannot achieve.

[0077] In addition, the corresponding testing equipment for this system is small in size and can be directly installed on the packaging line or light inspection machine to achieve equipment integration. This enables real-time monitoring of the production process and online alarm and rejection of defective products, avoiding the waste of products caused by continuous generation of defective products.

[0078] As the sealed bottle passes through a pre-defined location, continuous sampling is triggered simultaneously to measure the gas cavity at the bottle head and analyze the gas composition within the packaging. During the analysis, sliding filtering and false point filtering are applied to the data, greatly eliminating interference from reflections and refractions of the laser beam as it passes through the glass packaging container. This system eliminates the need for vacuum or nitrogen purging of the target area. Improvements to the detection method increase the detection speed, meeting the production requirements of production lines.

[0079] This invention also provides a method for detecting the seal integrity of a sealed bottle, such as... Figure 3 As shown, the method includes: In step S302, after receiving the trigger signal, the controller emits the first laser beam and transmits the first laser beam to the collimator through an optical fiber. In step S304, the collimator corrects the first laser beam and directs the corrected second laser beam toward the sealed bottle under test, so as to transmit a laser signal to the signal receiving unit through the sealed bottle under test. In step S306, the signal receiving unit outputs a first signal based on the laser signal and sends the first signal to the controller. The controller processes the first signal, outputs a second signal, and sends the second signal to the host computer. In step S308, the host computer calculates the content of the first gas in the sealed bottle to be tested based on the second signal, compares the content of the first gas with the preset standard gas content, obtains the comparison result, and determines the sealing integrity test result of the sealed bottle to be tested based on the comparison result.

[0080] The above-mentioned method for testing the seal integrity of sealed bottles can complete the online testing of the seal integrity of the sealed bottles without damaging them, and can also achieve full inspection of sealed bottle products produced on the production line.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A system for detecting the seal integrity of a sealed bottle, characterized by, The system comprises a controller, a collimator, a signal receiving unit and a host computer; The controller is configured to emit a first laser beam after receiving a trigger signal, and transmit the first laser beam to the collimator through an optical fiber; The collimator is configured to correct the first laser beam, and emit a second laser beam obtained after correction to the sealed bottle to be measured, so as to emit a laser signal to the signal receiving unit through the sealed bottle to be measured; The signal receiving unit is configured to output a first signal according to the laser signal, send the first signal to the controller, process the first signal by the controller, output a second signal, and send the second signal to the host computer; The host computer is configured to calculate a first gas content in the sealed bottle to be measured according to the second signal, compare the first gas content with a preset standard gas content to obtain a comparison result, and determine a sealing integrity detection result of the sealed bottle to be measured according to the comparison result; wherein the host computer is configured to calculate a difference value between the second signal and a signal value corresponding to the second laser beam, obtain a laser attenuation value, and confirm the first gas content in the sealed bottle to be measured by the laser attenuation value.

2. The system of claim 1, wherein, The controller comprises a trigger sensor, a signal conditioning module and a laser device which are sequentially connected in communication; the laser device is connected in communication with the collimator through an optical fiber; The trigger sensor is configured to output a first control signal to the signal conditioning module after receiving a trigger signal; The signal conditioning module is configured to output a laser emission signal according to the first control signal; The laser device is configured to emit a first laser beam according to the laser emission signal.

3. The system of claim 1, wherein, The controller further comprises a signal amplifier and a communication module; the signal amplifier is connected in communication with the signal receiving unit; and the communication module is connected in communication with the host computer; The signal amplifier is configured to amplify and process the first signal to obtain a second signal, and send the second signal to the host computer through the communication module.

4. The system of claim 1, wherein, The host computer is configured to: During detection of a single sealed bottle to be measured, receive a plurality of second signals continuously collected within a preset test time window, obtain a plurality of laser attenuation values, and form an attenuation value sequence; Analyze the attenuation value sequence to identify an invalid data segment in which an attenuation value is distorted due to a reason of the sealed bottle to be measured itself; Eliminate the invalid data segment, and calculate the first gas content based on a remaining valid attenuation value sequence.

5. The system of claim 4, wherein, The analysis of the attenuation value sequence to identify the invalid data segment in which the attenuation value is distorted due to the reason of the sealed bottle to be measured itself comprises: adopting a time series decomposition algorithm to decompose the attenuation value sequence into a trend component sequence and a residual component sequence; calculating a first-order difference sequence of the residual component sequence, comparing an absolute value of the difference sequence with a dynamic threshold value preset based on overall fluctuation of data, and identifying an attenuation value data point corresponding to a difference value exceeding the threshold value as a distortion point caused by large light path loss due to poor uniformity of a bottle surface shape; and / or comparing the attenuation value sequence with an attenuation value range predicted according to a normal passing model of the sealed bottle to be measured. ​ The attenuation value data points beyond the predicted range are identified as invalid data points caused by serious deflection or shielding of the bottle body.

6. The system of claim 5, wherein, The attenuation value range predicted according to the normal passing model of the to-be-tested sealed bottle is an absolute boundary range determined based on the physical principle of light transmission. The lower limit value of the absolute boundary range is a theoretically minimum possible attenuation value, which corresponds to a case that the laser beam is not absorbed by the to-be-tested gas at all. The upper limit value of the absolute boundary range is a theoretically maximum possible attenuation value, which corresponds to a theoretical limit value of light intensity attenuation caused by the bottle body and the medium in the bottle, absorption and reflection, etc. when the laser beam passes through the to-be-tested sealed bottle. The invalid data points refer to data points with attenuation values lower than the minimum possible attenuation value or higher than the maximum possible attenuation value.

7. The system of claim 4, wherein, The calculation of the first gas content based on the remaining effective attenuation value sequence comprises: a corresponding quality weight coefficient is calculated for each effective attenuation value in the remaining effective attenuation value sequence; a final characteristic attenuation value is determined by weighted calculation of the effective attenuation values according to the quality weight coefficients; the first gas content is calculated according to the characteristic attenuation value; The calculation of the quality weight coefficient comprises at least one of the following modes: a stability weight factor is calculated based on the stability of the data points in the local time window in which the effective attenuation value is located; a proximity weight factor is calculated based on the deviation of the effective attenuation value from the statistical central value of the entire effective attenuation value sequence.

8. The system of claim 1, wherein, The host computer is further configured to: if the comparison result indicates that the deviation value between the first gas content and the standard gas content exceeds the preset threshold range, it is determined that the to-be-tested sealed bottle has a leak; if the comparison result indicates that the deviation value between the first gas content and the standard gas content does not exceed the preset threshold range, it is determined that the to-be-tested sealed bottle has no leak.

9. The system of claim 1, wherein, The system further comprises a sorting module, which is configured to: if the to-be-tested sealed bottle has a leak when it is moved to a preset rejection position, reject the to-be-tested sealed bottle in a preset rejection mode; the preset rejection mode comprises a blowing mode, a hitting mode or a guide diversion mode; if the to-be-tested sealed bottle has no leak, allow the to-be-tested sealed bottle to pass.

10. A method of detecting the seal integrity of a sealed bottle, characterized by, The method comprises: after the controller receives a trigger signal, the first laser beam is emitted and transmitted to the collimator through an optical fiber; the collimator corrects the first laser beam and emits the second laser beam obtained after correction to the to-be-tested sealed bottle, so as to emit a laser signal to the signal receiving unit through the to-be-tested sealed bottle; the signal receiving unit outputs a first signal according to the laser signal, sends the first signal to the controller, processes the first signal through the controller, outputs a second signal, and sends the second signal to the host computer; the host computer calculates the first gas content in the to-be-tested sealed bottle according to the second signal, compares the first gas content with a preset standard gas content to obtain a comparison result, and determines the sealing integrity detection result of the to-be-tested sealed bottle according to the comparison result. A display module is configured to display the sealing integrity detection result of the sealed bottle to be detected.

Images