Method and system for controlling sterile filling process for nitrogen filling protection of midazolam oral liquid

Abstract

The invention provides a sterile filling process control method and system for midazolam oral liquid nitrogen charging protection, and relates to the technical field of medicine filling process control. After midazolam oral liquid is subjected to nitrogen charging, a headspace gas multispectral image sequence in a bottle before sealing is collected through a transmission imaging system integrated with a filling line; extracting oxygen characteristic absorption wave band light intensity attenuation data through spectrum separation, and generating a bottleneck plane oxygen two-dimensional concentration diagram; calculating the local oxygen partial pressure of the pixel, and constructing a three-dimensional oxygen concentration field and marking an oxygen-enriched area by combining pre-stored geometric parameters of the bottle body; and then the model reference adaptive controller is used for comparing the model with an ideal nitrogen model to generate supplementary parameters containing purging angle, intensity and time, an execution mechanism is driven to perform fixed-point pulse purging, and after verification reaches the standard, a screw cap is used for sealing, so that accurate detection, directional processing and standard verification of the headspace oxygen-enriched area in the midazolam oral liquid bottle can be realized, and the safety of the midazolam oral liquid bottle is ensured. And the oxygen concentration before sealing meets the standard.

Description

Technical Field

[0001] This application relates to the field of pharmaceutical filling process control technology, and in particular to a method and system for aseptic filling process control of midazolam oral solution with nitrogen protection. Background Technology

[0002] In the pharmaceutical field, midazolam oral solution is a commonly used drug, and its quality and stability are of paramount importance. Currently, nitrogen purging is a common protective measure. During aseptic filling under nitrogen purging, to prevent oxidation and deterioration of the oral solution, extend its shelf life, and ensure efficacy, precise control of the oxygen concentration in the headspace region of the bottle is necessary. This necessitates an efficient, precise, and monitorable process control method to ensure that each bottle of oral solution is filled and sealed in a standard low-oxygen environment.

[0003] Currently, for aseptic filling of midazolam oral solution with nitrogen protection, some companies use a fixed-parameter nitrogen purging method. That is, during the filling process, the bottle is uniformly purged with nitrogen according to pre-set parameters such as nitrogen flow rate, time, and pressure. Afterwards, simple sampling tests are used to roughly assess whether the oxygen concentration inside the bottle meets the standard.

[0004] This fixed-parameter nitrogen purging method lacks flexibility and precision. Because the actual conditions during the filling process of each oral liquid may vary—such as the liquid level and headspace volume within the bottle—uniform nitrogen purging parameters cannot be adjusted according to the specific circumstances of each bottle. This can easily lead to some bottles failing to meet the expected oxygen concentration standards. Furthermore, sampling inspections can only reflect the situation of a portion of the products and cannot monitor and address every single bottle, making it difficult to guarantee the quality stability of all products. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for controlling the aseptic filling process of midazolam oral solution with nitrogen protection, in order to solve the problems of the existing technology that uses fixed parameter nitrogen purging and sampling inspection, which cannot adjust the nitrogen purging parameters according to the specific situation of each bottle, cannot monitor and process each bottle, and makes it difficult to ensure the stability of product quality.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a method for controlling the aseptic filling process of midazolam oral solution under nitrogen protection, comprising: After nitrogen filling of midazolam oral solution, a multispectral image sequence of headspace gas inside the bottle before sealing was acquired by a transmission imaging system integrated into the filling line. The spectral image sequence is subjected to spectral separation processing to extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane is generated based on the light intensity attenuation data. Based on the two-dimensional concentration map, the local oxygen partial pressure value corresponding to each pixel is calculated. Combined with the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, a three-dimensional oxygen concentration field of the headspace region inside the bottle is constructed. The oxygen-rich region with excessive concentration is identified according to the preset oxygen concentration threshold. The three-dimensional oxygen concentration field is spatially registered and compared with the ideal nitrogen coverage model by a model reference adaptive controller. Based on the distribution characteristics of the oxygen-rich area, supplementary purging parameters for each bottle are generated. The supplementary purging parameters include purging angle, airflow intensity and purging duration. Based on the supplementary purging parameters, the purging actuator controls the nozzle orientation according to the purging angle, and performs targeted pulse purging on the oxygen-rich area using the purging duration and airflow intensity. After targeted pulse purging, the treated headspace gas inside the bottle is quickly verified and tested. After confirming that the oxygen concentration reaches the predetermined standard, the bottle enters the capping and sealing process.

[0007] Optionally, the three-dimensional oxygen concentration field is spatially registered and compared with an ideal nitrogen coverage model using a model reference adaptive controller. Based on the distribution characteristics of the oxygen-rich region, supplementary purging parameters are generated for each bottle. These supplementary purging parameters include purging angle, airflow intensity, and purging duration, including: The model reference adaptive controller loads a pre-stored ideal nitrogen coverage model, which defines the target oxygen concentration value at each spatial location in the headspace region inside the bottle. The three-dimensional oxygen concentration field is spatially registered with the ideal nitrogen coverage model to establish a precise correspondence between the three-dimensional spatial coordinates of the two models. Based on the precise correspondence, the measured oxygen concentration values ​​at each location in the three-dimensional oxygen concentration field are compared with the target oxygen concentration values ​​at the corresponding locations in the ideal nitrogen coverage model to identify the concentration difference values. Based on the spatial coordinate range, volume, and maximum oxygen concentration deviation of the oxygen-rich region, supplementary purging parameters are generated for each bottle, combined with the concentration difference value.

[0008] Optionally, the supplementary purging parameters for each bottle are generated based on the spatial coordinate range, volume, and maximum oxygen concentration deviation of the oxygen-rich region, combined with the concentration difference value, including: For each bottle, the azimuth angle of the spatial coordinate range of the oxygen-enriched area relative to the central axis of the bottle is used as the purging angle. The airflow intensity is determined based on the concentration difference value and the maximum oxygen concentration deviation value. The weighted sum of the volume of the oxygen-enriched area and the concentration difference value is used as the purging duration. The combination of the purging angle, airflow intensity, and purging duration is used as supplementary purging parameters for the corresponding bottle.

[0009] Optionally, the step of driving the purging actuator based on the supplementary purging parameters to control the nozzle orientation according to the purging angle, using the purging duration and airflow intensity to perform targeted pulse purging on the oxygen-rich area, and after targeted pulse purging, rapidly verifying and detecting the treated headspace gas inside the bottle, and confirming that the oxygen concentration reaches the predetermined standard before proceeding to the capping and sealing process, includes: Based on the supplementary purging parameters, a control signal is sent to the purging actuator to drive the purging actuator to start operation; According to the purging angle in the supplementary purging parameters, the nozzle is rotated to the position corresponding to the purging angle by the angle adjustment component of the purging actuator, so that the nozzle is aligned with the spatial position of the oxygen-rich area. Adjust the nitrogen output pressure of the purging actuator according to the airflow intensity in the supplementary purging parameters, control the duration of nitrogen output according to the purging duration in the supplementary purging parameters, and inject nitrogen into the oxygen-rich area to complete the fixed-point pulse purging. After the fixed-point pulse purging is completed, the spectral image of the headspace region inside the bottle is acquired, the light intensity attenuation data of the characteristic absorption band of oxygen molecules is extracted, and the current oxygen concentration value is calculated. The current oxygen concentration value is compared with a preset oxygen concentration threshold. If the current oxygen concentration value reaches the preset oxygen concentration threshold, a sealing control signal is triggered to cause the bottle to enter the capping and sealing process.

[0010] Optionally, the step of calculating the local oxygen partial pressure value corresponding to each pixel based on the two-dimensional concentration map, constructing a three-dimensional oxygen concentration field in the headspace region inside the bottle by combining the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, and identifying oxygen-rich areas with excessive concentrations according to a preset oxygen concentration threshold includes: Based on the light intensity attenuation data of each pixel in the two-dimensional concentration map, the local oxygen partial pressure value corresponding to each pixel is calculated, and the pre-stored bottle geometry parameters of midazolam oral solution are obtained. Based on the local oxygen partial pressure value and the bottle body geometric parameters, a three-dimensional coordinate system with the bottle mouth plane as the reference is established, and the headspace region inside the bottle is divided into multiple layered planes at preset intervals along the height direction of the bottle body. On each layered plane, based on the variation law of the radial dimension of the bottle body, the oxygen partial pressure value of the bottle mouth plane is mapped to the corresponding position of each layered plane through the bilinear interpolation algorithm to obtain the oxygen partial pressure value of the three-dimensional spatial grid node; Based on the oxygen partial pressure values ​​of all three-dimensional spatial grid nodes, a continuous three-dimensional oxygen concentration field is generated using a three-dimensional data reconstruction algorithm. The oxygen concentration value of each node in the three-dimensional oxygen concentration field is compared with the preset oxygen concentration threshold point by point. The three-dimensional spatial regions where the oxygen concentration value exceeds the threshold are identified as oxygen-rich regions. The spatial coordinate range, volume, and maximum oxygen concentration deviation of each oxygen-rich region are recorded.

[0011] Optionally, the step of acquiring a multispectral image sequence of the headspace gas inside the bottle before sealing using a transmission imaging system integrated on the filling line after nitrogen filling of the midazolam oral solution includes: After nitrogen purging of midazolam oral solution, a transmission imaging system integrated on the filling line is activated. The transmission imaging system includes a hyperspectral linear array camera and a specific band of LED array light source. The LED array light source is controlled to sequentially illuminate the narrowband light-emitting units corresponding to the characteristic absorption bands of oxygen molecules in a preset order; The hyperspectral linear array camera is synchronously triggered to collect the transmitted light signal of the corresponding band of the head air gas inside the bottle when each narrowband light-emitting unit is lit, and convert the transmitted light signal into digital image data. The digital image data acquired in each band are combined in spectral band order to generate a multispectral image sequence containing information on the characteristic absorption bands of oxygen molecules.

[0012] Optionally, the step of performing spectral separation processing on the spectral image sequence, extracting light intensity attenuation data of the characteristic absorption bands of oxygen molecules, and generating a two-dimensional concentration map reflecting the oxygen distribution at the bottle opening plane based on the light intensity attenuation data includes: The multispectral image sequence is subjected to spectral separation processing to identify and extract digital image data corresponding to the characteristic absorption bands of oxygen molecules; The light intensity value of each pixel is obtained from the digital image data, and the attenuation of the light intensity value of each pixel relative to the initial light intensity value of the corresponding band is calculated to obtain the light intensity attenuation data of each pixel. The position of each pixel is mapped one-to-one with the two-dimensional coordinates of the bottle opening plane. The light intensity attenuation data corresponding to all the two-dimensional coordinates are arranged in the order of the coordinate distribution of the bottle opening plane to generate a two-dimensional concentration map reflecting the oxygen distribution of the bottle opening plane.

[0013] Secondly, this application provides a nitrogen-filled aseptic filling process control system for midazolam oral solution, comprising: The acquisition module is used to acquire multispectral image sequences of headspace gas inside the bottle before sealing by a transmission imaging system integrated on the filling line after nitrogen filling of midazolam oral solution. The processing module is used to perform spectral separation processing on the spectral image sequence, extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and generate a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane based on the light intensity attenuation data. The construction module is used to calculate the local oxygen partial pressure value corresponding to each pixel based on the two-dimensional concentration map, construct a three-dimensional oxygen concentration field in the headspace region inside the bottle by combining the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, and identify the oxygen-rich areas with excessive concentration according to the preset oxygen concentration threshold. The generation module is used to spatially register and compare the three-dimensional oxygen concentration field with the ideal nitrogen coverage model through a model reference adaptive controller, and generate supplementary purging parameters for each bottle based on the distribution characteristics of the oxygen-rich area. The supplementary purging parameters include purging angle, airflow intensity and purging duration. The verification module is used to drive the purging actuator to control the nozzle orientation according to the purging angle based on the supplementary purging parameters, and to perform fixed-point pulse purging on the oxygen-rich area using the purging duration and airflow intensity. After the fixed-point pulse purging, the treated headspace gas in the bottle is quickly verified and tested. After confirming that the oxygen concentration reaches the predetermined standard, the bottle enters the capping and sealing process.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the aseptic filling process control method for nitrogen protection of midazolam oral solution as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the aseptic filling process control method for nitrogen protection of midazolam oral solution as described in the first aspect above.

[0016] The aseptic filling process control method for nitrogen-filled protection of midazolam oral solution provided in this application acquires multispectral image sequences using a transmission imaging system integrated into the filling line after nitrogen filling of the midazolam oral solution. Oxygen characteristic absorption data is extracted through spectral separation to generate a two-dimensional concentration map. A three-dimensional oxygen concentration field is constructed by combining the bottle's geometric parameters and accurately identifying oxygen-rich areas. Then, a model reference adaptive controller is used to compare with an ideal nitrogen model to generate specific supplementary purging parameters. The actuator is driven to perform targeted pulse purging of the oxygen-rich areas according to these parameters. Finally, after rapid verification confirming that the oxygen concentration meets the standards, the solution proceeds to the sealing process. This method achieves precise monitoring of headspace oxygen concentration within the bottle, precise positioning of oxygen-rich areas, and personalized targeted treatment. It effectively avoids the limitations of fixed-parameter nitrogen filling, ensuring that each bottle of oral solution is in a standard low-oxygen environment before sealing, significantly improving the accuracy, flexibility, and product quality stability of the filling process.

[0017] Furthermore, a pre-stored ideal nitrogen coverage model is loaded using a model reference adaptive controller. The three-dimensional oxygen concentration field is spatially registered with this model to establish a precise correspondence between the three-dimensional coordinates. The measured oxygen concentration values ​​at corresponding locations are compared with the target oxygen concentration values ​​to obtain the concentration difference value. Then, combined with the spatial coordinate range, volume, maximum oxygen concentration deviation, and concentration difference value of the oxygen-enriched area, supplementary purging parameters, including purging angle, airflow intensity, and duration, are generated for each bottle. This step achieves accurate comparison between the three-dimensional oxygen concentration field and the ideal model, enabling the supplementary purging parameters to be personalized according to the specific characteristics of the oxygen-enriched area of ​​each bottle. This improves the targeting and accuracy of the supplementary purging parameters and provides a scientific basis for subsequent fixed-point pulse purging. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application 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 of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic flowchart illustrating a method for controlling the aseptic filling process of midazolam oral solution with nitrogen protection, provided in an embodiment of this application. Figure 2 A flowchart illustrating a specific embodiment of a method for controlling the aseptic filling process of midazolam oral solution with nitrogen protection, provided in this application. Figure 3 A schematic diagram of a specific embodiment of a method for controlling the aseptic filling process of midazolam oral solution with nitrogen protection, provided in this application. Figure 4 This is a schematic diagram of a nitrogen-filled aseptic filling process control system for midazolam oral solution, provided as an embodiment of this application. Detailed Implementation

[0020] In the pharmaceutical field, nitrogen-purging aseptic filling of midazolam oral solution plays a crucial role in product quality and stability. However, existing fixed-parameter nitrogen purging methods have significant shortcomings. Because the actual conditions at the time of filling vary for each bottle of oral solution, such as different liquid levels and headspace volumes, uniform nitrogen purging parameters cannot be applied to specific needs, resulting in some bottles having substandard oxygen concentrations. Furthermore, sampling inspections can only reflect the condition of a portion of the products and cannot monitor and address every single bottle, making it difficult to ensure that all products are filled and sealed in a compliant low-oxygen environment, thus compromising product quality and stability.

[0021] To address the aforementioned issues, this application proposes a method for controlling the aseptic filling process of midazolam oral solution under nitrogen-filled protection. After nitrogen filling, the method utilizes a transmission imaging system to acquire multispectral image sequences of the headspace gas inside the bottle. Through a series of processing steps, a two-dimensional concentration map and a three-dimensional oxygen concentration field are generated, accurately identifying the oxygen-rich areas. Then, based on the distribution characteristics of the oxygen-rich areas, supplementary purging parameters are generated for each bottle, driving the purging actuator to perform targeted pulse purging of the oxygen-rich areas, followed by rapid verification testing. This approach allows for precise adjustments based on the specific conditions of each bottle, monitoring and processing each bottle to ensure that each bottle of oral solution is filled and sealed in a standard low-oxygen environment, effectively improving the stability of product quality.

[0022] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] The core of this application is to provide a method for controlling the aseptic filling process of midazolam oral solution with nitrogen protection, and a flowchart of one specific embodiment is shown below. Figure 1 As shown, the method includes: S101. After nitrogen filling of midazolam oral solution, a multispectral image sequence of headspace gas inside the bottle before sealing is acquired by a transmission imaging system integrated on the filling line. Optionally, step S101 may specifically include the following steps: S1011. After nitrogen filling treatment of midazolam oral solution, activate the transmission imaging system integrated on the filling line, the transmission imaging system comprising a hyperspectral linear array camera and a specific band LED array light source. S1012. Control the LED array light source to sequentially light up the narrowband light-emitting units corresponding to the characteristic absorption bands of oxygen molecules in a preset order; S1013. Synchronously trigger the hyperspectral linear array camera, and collect the transmitted light signal of the corresponding band of the head air gas inside the bottle when each narrowband light-emitting unit is lit, and convert the transmitted light signal into digital image data. S1014. Combine the digital image data acquired in each band according to the spectral band order to generate a multispectral image sequence containing information on the characteristic absorption bands of oxygen molecules.

[0024] In the above scheme, the transmission imaging system is integrated into the filling line and is used to acquire image information of the headspace gas inside the bottle. It consists of a hyperspectral linear array camera and an LED array light source with specific wavelengths. The hyperspectral linear array camera is a camera capable of acquiring spectral information of a specific area and converting the acquired transmitted light signal into digital image data. The LED array light source is a light source composed of multiple LED light-emitting units, including narrowband light-emitting units corresponding to the characteristic absorption band of oxygen molecules. The characteristic absorption band of oxygen molecules refers to the absorption phenomenon of oxygen molecules in a specific spectral band, which allows analysis of the oxygen content in the headspace gas inside the bottle. The narrowband light-emitting unit is a light-emitting unit in the LED array light source that emits light only in a specific narrow spectral band. The transmitted light signal is the light signal carrying gas information after light passes through the headspace gas inside the bottle. The digital image data is the digital image information obtained after the hyperspectral linear array camera converts the transmitted light signal. The multispectral image sequence is an image sequence containing information about the characteristic absorption band of oxygen molecules, composed of digital image data acquired in each band in spectral band order.

[0025] In this embodiment of the application, after the midazolam oral solution has undergone nitrogen filling treatment in step S1011, the transmission imaging system installed on the filling line is started. The system is equipped with a hyperspectral linear array camera that can capture fine spectral information and an LED array light source that can emit light in a specific wavelength band.

[0026] Secondly, in step S1012, the control system calls the pre-stored narrowband light-emitting unit lighting sequence and sends control signals to the LED array light source in this sequence, so that the narrowband light-emitting units corresponding to the characteristic absorption band of oxygen molecules in the light source are lit one after another, ensuring that each narrowband light-emitting unit works independently at a specified time and emits light of the corresponding band to illuminate the headspace area inside the bottle.

[0027] Next, in step S1013, while sending a signal to the LED array light source to light up a certain narrowband light-emitting unit, a collection signal is sent to the hyperspectral linear array camera to achieve synchronous operation between the camera and the light source. When the light emitted by the narrowband light-emitting unit passes through the headspace gas inside the bottle, the camera captures the light signal after it has been transmitted through the gas in a timely manner. Then, through the photoelectric conversion component inside the camera, the captured light signal is converted into digital image data composed of a series of numbers and temporarily stored.

[0028] Finally, in step S1014, the digital image data of each band acquired in step S1013 are retrieved. According to the preset order of the corresponding spectral bands of each image, these single-band digital images are arranged and integrated in sequence to form a continuous image set. This set is a multispectral image sequence containing the characteristic absorption band information of oxygen molecules.

[0029] In practical applications, on an oral liquid bottling production line in a pharmaceutical factory, after a bottle of midazolam oral solution completes nitrogen purging, the production line control system automatically activates the transmission imaging system. The control module in the LED array light source, following a preset sequence, first illuminates the narrowband light-emitting unit corresponding to a specific absorption band of oxygen molecules, simultaneously triggering a hyperspectral linear array camera to acquire the transmitted light signal of that band and convert it into digital image data. Then, other narrowband light-emitting units corresponding to the characteristic absorption bands of oxygen molecules are illuminated sequentially, and the hyperspectral linear array camera synchronously acquires and converts digital image data for each band. Finally, the data processing module combines these digital image data according to the spectral band order to generate a multispectral image sequence of the headspace gas inside the bottle.

[0030] The overall scheme of S101 described above activates the transmission imaging system, using an LED array light source to sequentially illuminate narrowband light-emitting units in specific wavelength bands, and simultaneously triggers a hyperspectral linear array camera to acquire transmitted light signals, ultimately generating a multispectral image sequence. This accurately acquires information on the oxygen molecule characteristic absorption band of the headspace gas inside the bottle, providing a detailed and accurate data foundation for subsequent analysis of the oxygen concentration distribution within the bottle. This helps improve the evaluation and control of nitrogen filling protection effects and ensures product quality.

[0031] S102. Perform spectral separation processing on the spectral image sequence, extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and generate a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane based on the light intensity attenuation data. Optionally, step S102 may specifically include the following steps: S1021. Perform spectral separation processing on the multispectral image sequence to identify and extract digital image data corresponding to the characteristic absorption bands of oxygen molecules; S1022. Obtain the light intensity value of each pixel from the digital image data, calculate the attenuation of the light intensity value of each pixel relative to the initial light intensity value of the corresponding band, and obtain the light intensity attenuation data of each pixel. S1023. Correspond the position of each pixel to the two-dimensional coordinates of the bottle mouth plane, arrange the light intensity attenuation data corresponding to all two-dimensional coordinates in the order of coordinate distribution of the bottle mouth plane, and generate a two-dimensional concentration map reflecting the oxygen distribution of the bottle mouth plane.

[0032] In the above scheme, spectral separation processing is the operation of distinguishing and extracting image data from different bands in the spectral image sequence. The characteristic absorption band of oxygen molecules refers to the absorption phenomenon of oxygen molecules in specific spectral bands, which can be used to analyze oxygen conditions. A pixel is the smallest unit in an image, and each pixel has a corresponding light intensity value. Light intensity value is a numerical value that measures the intensity of light. The initial light intensity value is the light intensity value of the corresponding band before the light is absorbed by the headspace gas inside the bottle. Light intensity attenuation data is the amount of attenuation of the light intensity value of each pixel compared to the initial light intensity value of the corresponding band, reflecting the degree of light absorption by oxygen at that location. Two-dimensional coordinates are used to represent the coordinates of each point on the bottle opening plane. The two-dimensional concentration map is generated based on the light intensity attenuation data, reflecting the oxygen distribution on the bottle opening plane.

[0033] In this embodiment of the application, the multispectral image sequence obtained in step S101 is first subjected to spectral separation processing in step S1021. For example, a spectral unmixing algorithm is used to identify and extract image data that corresponds only to the characteristic absorption band of oxygen molecules from images of multiple bands.

[0034] Secondly, in step S1022, the light intensity value of each pixel is read from the extracted image data of the band, and the initial light intensity value of the corresponding band is subtracted from the light intensity value of each pixel. The initial light intensity value is the light intensity when the light does not pass through the gas. The light intensity attenuation of each pixel is calculated to obtain the light intensity attenuation data. For example, if the light intensity value of a certain pixel is 80 and the initial light intensity value of the corresponding band is 100, then the attenuation is 20.

[0035] Finally, in step S1023, a one-to-one correspondence is established between the position of each pixel in the image and the two-dimensional coordinates of the bottle opening plane. For example, the row and column numbers of the image pixels correspond to the X and Y coordinates of the bottle opening plane. Then, all the light intensity attenuation data corresponding to the two-dimensional coordinates are arranged in the order of coordinate distribution of the bottle opening plane to generate a two-dimensional concentration map that reflects the oxygen distribution of the bottle opening plane.

[0036] In a practical application, on the oral liquid production line of a pharmaceutical factory, after acquiring the spectral image sequence of a bottle of oral liquid, professional spectral separation software was used to process it, identifying and extracting image data corresponding to the characteristic absorption bands of oxygen molecules. Through image analysis, the light intensity value of each pixel in the image data was obtained and compared with the pre-recorded initial light intensity value of the corresponding band to calculate the light intensity attenuation data of each pixel. Based on the calibration information of the imaging system, the pixel positions were mapped to the two-dimensional coordinates of the bottle opening plane. Using plotting software, the light intensity attenuation data were arranged according to the coordinate order of the bottle opening plane, generating a two-dimensional concentration map reflecting the oxygen distribution on the bottle opening plane.

[0037] The overall solution described in S102 accurately extracts image data of the characteristic absorption bands of oxygen molecules through spectral separation processing, calculates light intensity attenuation data, and correlates it with the two-dimensional coordinates of the bottle opening plane to ultimately generate a two-dimensional concentration map. This allows for a direct understanding of the oxygen distribution on the bottle opening plane, providing crucial information for subsequent assessment of oxygen concentration distribution in the headspace region, identification of oxygen-rich areas, and implementation of targeted nitrogen purging measures. This contributes to improving the effectiveness of nitrogen purging protection and product quality.

[0038] S103. Based on the two-dimensional concentration map, calculate the local oxygen partial pressure value corresponding to each pixel point, and construct a three-dimensional oxygen concentration field in the headspace region inside the bottle by combining the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, and identify the oxygen-rich areas with excessive concentration according to the preset oxygen concentration threshold. Optionally, step S103 may specifically include the following steps: S1031. Based on the light intensity attenuation data of each pixel in the two-dimensional concentration map, calculate the local oxygen partial pressure value corresponding to each pixel, and obtain the pre-stored bottle geometry parameters corresponding to midazolam oral solution. S1032. Based on the local oxygen partial pressure value and the bottle body geometric parameters, establish a three-dimensional coordinate system with the bottle mouth plane as the reference, and divide the headspace region inside the bottle into multiple layered planes at preset intervals along the height direction of the bottle body. S1033. On each layered plane, based on the radial dimension variation law of the bottle body, the oxygen partial pressure value of the bottle mouth plane is mapped to the corresponding position of each layered plane through the bilinear interpolation algorithm to obtain the oxygen partial pressure value of the three-dimensional spatial grid node. S1034. Based on the oxygen partial pressure values ​​of all three-dimensional spatial grid nodes, a continuous three-dimensional oxygen concentration field is generated using a three-dimensional data reconstruction algorithm. S1035. The oxygen concentration value of each node in the three-dimensional oxygen concentration field is compared with the preset oxygen concentration threshold point by point. The three-dimensional spatial regions where the oxygen concentration value exceeds the threshold are identified as oxygen-rich regions. The spatial coordinate range, volume and maximum oxygen concentration deviation of each oxygen-rich region are recorded.

[0039] In the above scheme, the local oxygen partial pressure value is calculated using light intensity attenuation data, reflecting the oxygen partial pressure at a specific location within the bottle. The bottle's geometric parameters are pre-stored information about the shape and size of the midazolam oral solution bottle, used to assist in constructing a three-dimensional oxygen concentration field. The three-dimensional coordinate system is established based on the bottle opening plane and is used to determine the spatial position of each point in the headspace region within the bottle. The layered planes are planes obtained by dividing the headspace region along the bottle's height at preset intervals. The bilinear interpolation algorithm is an algorithm used to extrapolate the value of unknown points on a plane based on the values ​​of known points; here, it is used to map the oxygen partial pressure value at the bottle opening plane to each layered plane. The three-dimensional spatial grid nodes are nodes in the grid divided in the three-dimensional coordinate system, and their oxygen partial pressure values ​​are obtained through the interpolation algorithm. The preset oxygen concentration threshold is a pre-set standard value for oxygen concentration. The oxygen-rich region is a three-dimensional spatial region in the three-dimensional oxygen concentration field where the oxygen concentration value exceeds the preset oxygen concentration threshold; its distribution characteristics include spatial coordinate range, volume size, and maximum oxygen concentration deviation.

[0040] In this embodiment, step S1031 first uses the light intensity attenuation data of each pixel in the two-dimensional concentration map, combined with the preset correspondence between light intensity attenuation and oxygen partial pressure, to calculate the local oxygen partial pressure value of each pixel. This relationship is obtained through previous experimental calibration. For example, the local oxygen partial pressure value = K × light intensity attenuation data, where K is a calibration coefficient. For example, if the calibration coefficient K = 0.4 and the light intensity attenuation data of a certain pixel is 25, then the local oxygen partial pressure value of that point = 0.4 × 25 = 10. At the same time, the pre-stored bottle geometry parameters corresponding to the midazolam oral solution are retrieved from the system to prepare for the subsequent construction of the three-dimensional concentration field.

[0041] Secondly, based on the calculated local oxygen partial pressure and bottle geometry parameters in step S1032, and combined with the bottle mouth center position in the bottle geometry parameters, a three-dimensional coordinate system is established: the bottle mouth plane is defined as the XY plane, the geometric center of the bottle mouth is set as the coordinate origin (0,0,0), and the direction from the bottle mouth to the bottom of the bottle is set as the positive Z-axis direction; then, along the Z-axis direction, the headspace region inside the bottle is divided into multiple layered planes parallel to the bottle mouth plane at preset intervals.

[0042] Next, in step S1033, the correspondence between each position and the pixel point on the bottle mouth plane is determined on each layered plane according to the radial dimension variation law of the bottle body. A bilinear interpolation algorithm is used, that is, the intermediate point value is calculated based on the known oxygen partial pressure values ​​of four adjacent pixels, and the oxygen partial pressure value of the bottle mouth plane is mapped to the corresponding position of each layered plane to obtain the oxygen partial pressure value of multiple grid nodes in three-dimensional space. For example, if the oxygen partial pressure values ​​of four pixels on the bottle mouth plane around a certain point on a certain layered plane are 8, 10, 9 and 11 respectively, then the oxygen partial pressure value of that point is calculated to be 9.5 through interpolation.

[0043] Then, in step S1034, the oxygen partial pressure values ​​of all three-dimensional spatial grid nodes are collected, and these nodes are arranged in an orderly manner according to their coordinate positions in the three-dimensional coordinate system. Next, a suitable three-dimensional data reconstruction algorithm (such as a volume rendering algorithm) is selected, and the oxygen partial pressure data of the arranged grid nodes is input into the algorithm. The algorithm will fill the discrete node data into continuous spatial data through data interpolation and rendering processing based on the difference in oxygen partial pressure values ​​between adjacent nodes, and present it in a visual way to form a continuous three-dimensional oxygen concentration field that can intuitively show the oxygen concentration at each position in the headspace region inside the bottle. For example, through the volume rendering algorithm, regions with different oxygen partial pressure values ​​are represented by different gray levels to form a three-dimensional concentration model with gray-scale gradient.

[0044] Finally, in step S1035, the oxygen concentration value of each node in the three-dimensional oxygen concentration field is compared point by point with the preset oxygen concentration threshold. Nodes with oxygen concentration values ​​greater than the preset threshold are marked, and the continuous spatial region formed by these marked nodes is the oxygen-rich region. Next, the start and end coordinates of the X-axis, Y-axis, and Z-axis of each oxygen-rich region in the three-dimensional coordinate system are determined, and its spatial coordinate range is calculated. Then, the volume of each oxygen-rich region is calculated based on the coordinate range and spatial grid density. Finally, the maximum oxygen concentration value in each oxygen-rich region is found, and the preset threshold is subtracted from this value to obtain the maximum oxygen concentration deviation value. The information of these oxygen-rich regions is recorded and stored.

[0045] In practical applications, on a pharmaceutical production line, light intensity attenuation data for each pixel is extracted from a two-dimensional concentration map. Local oxygen partial pressure values ​​are calculated based on pre-stored conversion relationships. Simultaneously, geometric parameters of the oral liquid bottle model are retrieved, including a headspace height of 5cm and a bottle mouth radius of 2cm. Then, a three-dimensional coordinate system is established with the bottle mouth plane as the XY plane, the bottle mouth center as (0,0,0), and the bottle height direction as the Z-axis. Combining a headspace height of 3.5cm and a preset interval of 1.75mm, 20 layered planes parallel to the bottle mouth plane are divided along the Z-axis from 0 to 35mm. On each layered plane, the correspondence between each position and the pixel on the bottle mouth plane is determined based on the radial dimension variation of the bottle body. For the layered plane with Z=4mm, grid node (2,3,4) is selected, and its four corresponding adjacent pixels are found on the bottle mouth plane. The oxygen partial pressure values ​​of the four pixels (1,2), (3,2), (1,4), and (3,4) are 6, 8, 7, and 9, respectively. The oxygen partial pressure value of this grid node is calculated to be 7.5 using bilinear interpolation. The oxygen partial pressure values ​​of all layered planar grid nodes are obtained in this way. The oxygen partial pressure values ​​of 5000 three-dimensional spatial grid nodes in the headspace region are collected, sorted by coordinates, and input into the volume rendering algorithm to generate a three-dimensional oxygen concentration field model. The preset oxygen concentration threshold is retrieved, and each node in the three-dimensional concentration field is compared point by point. The nodes with oxygen concentration values ​​greater than the preset oxygen concentration threshold are marked as two continuous oxygen-rich regions. The spatial coordinate range of the first region is X5-15mm, Y8-18mm, Z2-7mm, the volume is V, the maximum oxygen concentration deviation is H, and the corresponding information of the second region is recorded.

[0046] The aforementioned S103 solution, by combining a two-dimensional concentration map and bottle geometry parameters, can accurately construct a three-dimensional oxygen concentration field in the headspace region inside the bottle, comprehensively reflecting the distribution of oxygen concentration within the bottle. By comparing with a preset oxygen concentration threshold, oxygen-rich areas can be precisely identified and their distribution characteristics recorded. This facilitates subsequent implementation of effective nitrogen purging protection measures for oxygen-rich areas, improving the targeting and effectiveness of nitrogen purging and ensuring the quality and stability of the product in low-oxygen environments.

[0047] S104. The three-dimensional oxygen concentration field is spatially registered and compared with the ideal nitrogen coverage model by a model reference adaptive controller. Based on the distribution characteristics of the oxygen-rich area, supplementary purging parameters for each bottle are generated. The supplementary purging parameters include purging angle, airflow intensity and purging duration. Optionally, step S104 may specifically include the following steps: S1041. Load the pre-stored ideal nitrogen coverage model through the model reference adaptive controller. The ideal nitrogen coverage model defines the target oxygen concentration value at each spatial location in the headspace region inside the bottle. S1042. Spatial registration is performed between the three-dimensional oxygen concentration field and the ideal nitrogen coverage model to establish a precise correspondence between the three-dimensional spatial coordinates of the two models. S1043. Based on the precise correspondence, the measured oxygen concentration values ​​at each location in the three-dimensional oxygen concentration field are compared with the target oxygen concentration values ​​at the corresponding locations in the ideal nitrogen coverage model to identify the concentration difference values. S1044. Based on the spatial coordinate range, volume, and maximum oxygen concentration deviation of the oxygen-rich region, and combined with the concentration difference value, generate supplementary purging parameters for each bottle.

[0048] Specifically, step S1044 includes the following process: for each bottle, the azimuth angle of the spatial coordinate range of the oxygen-enriched area relative to the central axis of the bottle is taken as the purging angle; the airflow intensity is determined according to the concentration difference value and the maximum oxygen concentration deviation value; the weighted sum of the volume of the oxygen-enriched area and the concentration difference value is taken as the purging duration; and the combination of the purging angle, airflow intensity and purging duration is taken as the supplementary purging parameters for the corresponding bottle.

[0049] In the above scheme, the model reference adaptive controller is a controller with intelligent analysis and comparison functions, used to process relevant data from the three-dimensional oxygen concentration field and the ideal nitrogen coverage model. The ideal nitrogen coverage model is pre-stored and specifies the target oxygen concentration values ​​that should be achieved at various spatial locations in the headspace region within the bottle under ideal conditions. Spatial registration refers to establishing a precise correspondence between the three-dimensional oxygen concentration field and the ideal nitrogen coverage model on three-dimensional spatial coordinates for accurate comparison. The concentration difference value is obtained by comparing the measured oxygen concentration values ​​at each location in the three-dimensional oxygen concentration field with the target oxygen concentration values ​​at the corresponding locations in the ideal nitrogen coverage model. The oxygen-rich region is the part of the headspace region within the bottle where the oxygen concentration exceeds a preset threshold, and its distribution characteristics include spatial coordinate range, volume, and maximum oxygen concentration deviation. The bottle's central axis is a virtual straight line located at the center of the bottle in space. Supplementary purging parameters are used to further reduce the oxygen concentration in the oxygen-rich region within the bottle, including parameters such as purging angle, airflow intensity, and purging duration determined for each bottle.

[0050] In the embodiments of this application, such as Figure 2 As shown, firstly, the model reference adaptive controller loads the pre-stored ideal nitrogen coverage model in the system through step S1041. This model clarifies the target oxygen concentration value at each spatial location in the headspace inside the bottle. For example, the target oxygen concentration value at any location in the headspace inside the bottle does not exceed a predetermined value.

[0051] Secondly, feature points are extracted from the three-dimensional oxygen concentration field and the ideal nitrogen coverage model in step S1042. These feature points are fixed geometric feature positions of the bottle body, such as multiple points on the edge of the bottle mouth, the intersection of the bottle body axis and the bottle mouth plane, and the center of the cross-section at a specific height of the bottle body. Next, a spatial registration algorithm based on feature points is used to align the corresponding feature points extracted from the two models. For example, the center feature points of the bottle mouth of both models are located at spatial coordinates (0,0,0), and the feature lines of the bottle body axis are aligned with the Z-axis. Finally, through iterative optimization of the algorithm, the coordinate error of all corresponding feature points of the two models is controlled within a very small range, thereby establishing a precise correspondence between the three-dimensional spatial coordinates of the two models.

[0052] Next, in step S1043, all spatial locations in the three-dimensional oxygen concentration field are traversed to obtain the three-dimensional coordinates of each location and the corresponding measured oxygen concentration value. Based on the precise correspondence established in step S1042, a location that completely corresponds to the measured location coordinates is found in the ideal nitrogen coverage model, and the target oxygen concentration value of that location is read. Then, the concentration difference value of that location is calculated by subtracting the target oxygen concentration value of the corresponding location from the measured oxygen concentration value of each location.

[0053] Finally, in step S1044, for each bottle, the azimuth angle of the spatial coordinate range of the oxygen-enriched area relative to the central axis of the bottle is determined. This azimuth angle is the purging angle. For example, if the oxygen-enriched area is mainly distributed in the 30-degree direction to the right of the central axis of the bottle, then the purging angle is set to 30 degrees. Next, the airflow intensity is calculated based on the concentration difference value and the maximum oxygen concentration deviation value, and the basic airflow intensity is set as follows: The concentration difference coefficient is a, the maximum oxygen concentration deviation coefficient is b, and the airflow intensity is = +a×concentration difference value +b×maximum oxygen concentration deviation value; the larger both are, the greater the airflow intensity, for example, the base airflow intensity. =2, coefficients a=0.3, b=0.5, concentration difference value M=4, maximum oxygen concentration deviation value C=6, airflow intensity =2+0.3×4+0.5×6=6.2; then set the weighting coefficient of volume size as P and the weighting coefficient of concentration difference value as Q, calculate the product of oxygen-rich area volume size and P plus the product of concentration difference value and Q, the weighted sum is the purging duration, for example, when the volume is V and the concentration difference value is M, the purging duration = V×P+M×Q; finally, combine the calculated purging angle, airflow intensity and purging duration as the supplementary purging parameters for the corresponding bottle.

[0054] In practical applications, in the oral liquid filling workshop of a pharmaceutical factory, a model reference adaptive controller loads a pre-stored ideal nitrogen coverage model. Fixed geometric feature points are extracted from the three-dimensional oxygen concentration field and the ideal nitrogen coverage model, including six evenly distributed points at the bottle mouth edge, the intersection of the bottle body axis and the bottle mouth plane, and the center of the cross-section at Z=15mm. A spatial registration algorithm based on these feature points is used to establish a precise correspondence between the three-dimensional spatial coordinates. Then, all spatial positions in the three-dimensional oxygen concentration field are traversed, and the three-dimensional coordinates and measured oxygen concentration value N of position (8,8,8) are obtained. Based on the precise correspondence, the position (8,8,8) is found in the ideal nitrogen coverage model, and the target oxygen concentration value K is read. The concentration difference value at that position is calculated by subtracting K from N. Similar calculations are performed for the entire area. For a specific bottle, the azimuth angle of the oxygen-rich area relative to the bottle's central axis is determined to be 20 degrees. Therefore, the purging angle is set to 20 degrees. The airflow intensity is calculated based on the concentration difference value M=5 and the maximum oxygen concentration deviation value C=7, and a basic airflow intensity is set. =3, concentration difference coefficient a=0.4, maximum oxygen concentration deviation coefficient b=0.6, airflow intensity=3+0.4×5+0.6×7=9.2, set volume size weight coefficient P=0.4, concentration difference weight coefficient Q=0.6, when the oxygen-rich area volume is V, the purging duration=V×0.4+5×0.6, and finally combine the purging angle of 20 degrees, airflow intensity of 9.2 and the calculated purging duration as the supplementary purging parameters for this bottle.

[0055] The overall solution described in S104 accurately identifies differences in oxygen concentration within the bottle by precisely comparing the three-dimensional oxygen concentration field with an ideal nitrogen coverage model. It generates supplementary purging parameters based on the distribution characteristics of the oxygen-enriched area, enabling personalized processing for each bottle. This facilitates precise purging of the oxygen-enriched area, effectively reducing the oxygen concentration within the bottle, improving the effectiveness of nitrogen filling protection, and ensuring that each bottle of oral liquid is filled in a standard low-oxygen environment, thereby enhancing the stability and reliability of product quality.

[0056] S105. Based on the supplementary purging parameters, drive the purging actuator to control the nozzle orientation according to the purging angle, use the purging duration and airflow intensity to perform fixed-point pulse purging on the oxygen-rich area, and after fixed-point pulse purging, quickly verify and detect the treated headspace gas in the bottle, and after confirming that the oxygen concentration reaches the predetermined standard, proceed to the capping and sealing process.

[0057] Optionally, step S105 may specifically include the following steps: S1051. Based on the supplementary purging parameters, a control signal is sent to the purging actuator to drive the purging actuator to start operation; S1052. According to the purging angle in the supplementary purging parameters, the nozzle is rotated to a position corresponding to the purging angle by the angle adjustment component of the purging actuator, so that the nozzle is aligned with the spatial position of the oxygen-rich area. S1053. Adjust the nitrogen output pressure of the purging actuator according to the airflow intensity in the supplementary purging parameters, control the duration of nitrogen output according to the purging duration in the supplementary purging parameters, and inject nitrogen into the oxygen-rich area to complete the fixed-point pulse purging. S1054. After the fixed-point pulse purging is completed, acquire the spectral image of the headspace region inside the bottle after processing, extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and calculate the current oxygen concentration value. S1055. Compare the current oxygen concentration value with the preset oxygen concentration threshold. If the current oxygen concentration value reaches the preset oxygen concentration threshold, trigger the sealing control signal to make the bottle enter the capping and sealing process.

[0058] In the above scheme, the purging actuator is the device used to perform the purging operation; the control signal is generated based on supplementary purging parameters and is used to drive the purging actuator to start operation. The angle adjustment component is the part in the purging actuator used to adjust the nozzle orientation. The nozzle is the device in the purging actuator that injects nitrogen gas. The nitrogen output pressure is related to the airflow intensity, and the airflow intensity can be controlled by adjusting this pressure. Point-to-point pulse purging refers to accurately injecting nitrogen gas into the oxygen-rich area, and the injection process has pulse characteristics. The spectral image is an image containing different spectral information, used to detect the headspace gas condition inside the bottle. The oxygen molecule characteristic absorption band is the absorption phenomenon of oxygen molecules in a specific spectral band, which can be used to analyze the oxygen condition. The light intensity attenuation data is the data on the reduction of light intensity due to oxygen absorption when light passes through the headspace gas inside the bottle. The current oxygen concentration value is the current oxygen concentration inside the bottle calculated by analyzing the light intensity attenuation data. The preset oxygen concentration threshold is a pre-set qualified oxygen concentration standard. The sealing control signal is triggered after the oxygen concentration reaches the predetermined standard, used to make the bottle enter the capping and sealing process.

[0059] In this embodiment of the application, firstly, based on the supplementary purging parameters generated in step S104, a start control signal is sent to the purging actuator in step S1051 to drive the actuator to start running. For example, when the parameters are a purging angle of 15 degrees, an airflow intensity of 5, and a duration of 3 seconds, a start signal containing these parameters is sent to the actuator.

[0060] Secondly, in step S1052, based on the blowing angle in the supplementary blowing parameters, the transmission structure is activated to drive the nozzle to start rotating. During the rotation, the current azimuth angle of the nozzle is detected by the angle sensor. When the angle sensor detects that the nozzle azimuth angle has reached the extracted blowing angle, the servo motor stops rotating. At this time, the nozzle is exactly aligned with the spatial position of the oxygen-rich area. For example, when the blowing angle is 15 degrees, the adjustment component drives the nozzle to rotate to 15 degrees to the right of the central axis of the bottle.

[0061] Next, in step S1053, the airflow intensity and purging duration are extracted from the supplementary purging parameters. For example, if the airflow intensity is 6 and the duration is 3 seconds, the actuator adjusts the internal pressure regulating valve according to the airflow intensity, adjusting the nitrogen output pressure to the value corresponding to that airflow intensity. The higher the airflow intensity, the higher the output pressure. Then, the actuator's timer starts timing, and simultaneously opens the nitrogen injection valve, injecting nitrogen into the oxygen-rich area at the adjusted pressure. When the injection time displayed by the timer reaches the extracted purging duration, the nitrogen injection valve is closed, completing the fixed-point pulse purging operation. For example, an airflow intensity of 5 corresponds to adjusting the pressure to a certain value, and a duration of 3 seconds controls the injection for 3 seconds.

[0062] Then, after the fixed-point pulse purging is completed in step S1054, the spectral image of the headspace region inside the bottle is acquired using a spectral acquisition device. The image is processed to extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules. The current oxygen concentration value is calculated using a pre-established relationship model between light intensity attenuation and oxygen concentration.

[0063] Finally, in step S1055, the current oxygen concentration value is compared with the preset qualified oxygen concentration standard, i.e., the preset oxygen concentration threshold. If the current oxygen concentration value is less than or equal to the preset threshold, it means that the oxygen concentration has reached the qualified standard. Then, the control system sends a sealing control signal to the capping and sealing device. After receiving the control signal, the capping and sealing device starts the working process, tightens the cap on the current bottle body, and makes the bottle body enter the capping and sealing process.

[0064] In practical applications, in the oral liquid bottling workshop of Pharmaceutical Factory A, after determining the supplementary purging parameters for a particular bottle, the control system generates a control signal and sends it to the purging actuator, which then initiates the purging process. The angle adjustment component, based on a purging angle of 30°, rotates the nozzle to a 30° position, aligning it with the oxygen-rich area. The nitrogen output pressure is adjusted according to the required airflow intensity, and nitrogen injection is controlled for 5 seconds according to the purging duration parameter, completing the targeted pulse purging. After purging, the spectral acquisition device acquires a spectral image of the headspace region inside the bottle, extracts the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and calculates the current oxygen concentration. This value is compared with a preset oxygen concentration threshold; if the standard is met, a sealing control signal is triggered, and the bottle enters the capping and sealing process.

[0065] The overall solution of S105 described above, through precise control of the purging actuator, can accurately inject nitrogen into the oxygen-enriched area, effectively reducing the oxygen concentration in the oxygen-enriched area by utilizing appropriate airflow intensity and purging time. The rapid verification and detection mechanism can promptly confirm whether the oxygen concentration inside the bottle meets the standard, ensuring that only bottles with the standard oxygen concentration enter the capping and sealing process, thereby improving product quality and stability, and guaranteeing the storage and quality of oral liquids in low-oxygen environments.

[0066] The following is a complete example for steps 101-105, such as Figure 3 As shown, in a midazolam oral solution aseptic filling line, after a bottle of oral solution completes nitrogen filling, the transmission imaging system integrated on the filling line is activated. The system includes a hyperspectral linear array camera and a specific band of LED array light source. The LED array light source sequentially illuminates three narrowband light-emitting units corresponding to the characteristic absorption band of oxygen molecules in a preset order. Each time a unit is illuminated, the hyperspectral linear array camera is simultaneously triggered to collect the transmitted light signal of the corresponding band of headspace gas inside the bottle and convert it into digital image data. Finally, the digital image data of the three bands are combined in spectral order to generate a multispectral image sequence containing information on the characteristic absorption band of oxygen.

[0067] Next, the generated multispectral image sequence is processed by a linear spectral separation algorithm to identify and extract the image data of the corresponding oxygen molecule characteristic absorption band. The light intensity value of each pixel is read point by point from the image data. Combined with the initial light intensity value of 100 stored in the system for that band, the light intensity attenuation of each pixel is calculated. For example, if the light intensity value of a certain pixel is 70, the attenuation is 30. Then, the position of each pixel is mapped one-to-one with the two-dimensional coordinates of the bottle mouth plane. All light intensity attenuation data are arranged in the order of coordinate distribution to generate a two-dimensional concentration map reflecting the oxygen distribution on the bottle mouth plane.

[0068] Subsequently, based on the light intensity attenuation data of each pixel in the two-dimensional concentration map, the local oxygen partial pressure value of each pixel is calculated, while the pre-stored geometric parameters of this type of oral liquid bottle are retrieved. A three-dimensional coordinate system is established with the bottle mouth plane as the XY plane, the bottle mouth center as (0,0,0), and the bottle height direction as the Z-axis. The headspace region is divided into 25 layered planes at 2mm intervals along the Z-axis. On each layered plane, according to the radial dimension variation law of the bottle body, the oxygen partial pressure value of the bottle mouth plane is mapped to the corresponding position of each layered plane using a bilinear interpolation algorithm to obtain the oxygen partial pressure value of the three-dimensional spatial grid node. All grid node data are collected, and a continuous three-dimensional oxygen concentration field is generated through a volume rendering algorithm. Finally, the oxygen concentration value of each node is compared with the preset oxygen concentration threshold point by point to identify the oxygen-rich areas with excessive oxygen concentration, and the spatial coordinate range, volume size, and maximum oxygen concentration deviation of the oxygen-rich areas are recorded.

[0069] Then, the model reference adaptive controller loads the pre-stored ideal nitrogen coverage model, extracting feature points such as four uniformly distributed points at the bottle mouth edge and the bottle mouth center from both the 3D oxygen concentration field and the ideal model. A feature point-based registration algorithm is used to align the feature points of the two models, establishing a precise coordinate correspondence. Based on this correspondence, the measured oxygen concentration values ​​at each location in the 3D oxygen concentration field are compared with the target values ​​at the corresponding locations in the ideal model to calculate the concentration difference. The airflow intensity is then calculated based on the concentration difference and the maximum oxygen concentration deviation, with a base airflow intensity set as... The concentration difference coefficient is a, the maximum oxygen concentration deviation coefficient is b, and the airflow intensity is = +a×concentration difference value+b×maximum oxygen concentration deviation value, the larger both are, the greater the airflow intensity; then set the weighting coefficient of volume size as P and the weighting coefficient of concentration difference value as Q, calculate the product of oxygen-rich area volume size and P plus the product of concentration difference value and Q, the weighted sum is the purging duration, and combine these three parameters as supplementary purging parameters.

[0070] Finally, a control signal is sent to the purging actuator based on the supplementary purging parameters, driving the actuator to start. The actuator extracts the purging angle from the signal, and the servo motor of the angle adjustment component drives the nozzle to rotate to that position, aiming at the oxygen-rich area. Then, the airflow intensity and purging duration are extracted, the nitrogen output pressure is adjusted to the corresponding value, and nitrogen is sprayed according to the purging duration to complete the fixed-point pulse purging. After purging, the transmission imaging system is triggered to collect the headspace spectrum image again, and the light intensity attenuation calculation process is repeated to obtain the current oxygen concentration value. The current oxygen concentration value is compared with the preset oxygen concentration threshold. After confirming that the current oxygen concentration value is less than or equal to the preset oxygen concentration threshold, the sealing control signal is triggered, so that the bottle enters the capping and sealing process.

[0071] Figure 4 This is a schematic diagram of a specific embodiment of a nitrogen-filled aseptic filling process control system for midazolam oral solution provided in this application, referring to... Figure 4 The system may include: The acquisition module 41 is used to acquire a multispectral image sequence of the headspace gas inside the bottle before sealing by a transmission imaging system integrated on the filling line after the midazolam oral solution has been nitrogen-filled. Processing module 42 is used to perform spectral separation processing on the spectral image sequence, extract light intensity attenuation data of the characteristic absorption band of oxygen molecules, and generate a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane based on the light intensity attenuation data. The construction module 43 is used to calculate the local oxygen partial pressure value corresponding to each pixel based on the two-dimensional concentration map, construct a three-dimensional oxygen concentration field in the headspace region inside the bottle by combining the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, and identify the oxygen-rich areas with excessive concentration according to the preset oxygen concentration threshold. The generation module 44 is used to spatially register and compare the three-dimensional oxygen concentration field with the ideal nitrogen coverage model through a model reference adaptive controller, and generate supplementary purging parameters for each bottle based on the distribution characteristics of the oxygen-rich area. The supplementary purging parameters include purging angle, airflow intensity and purging duration. Drive module 45 is used to drive the purging actuator to control the nozzle orientation according to the purging angle based on the supplementary purging parameters, and to perform fixed-point pulse purging on the oxygen-rich area using the purging duration and airflow intensity. After fixed-point pulse purging, the treated headspace gas in the bottle is quickly verified and tested. After confirming that the oxygen concentration reaches the predetermined standard, the bottle enters the capping and sealing process.

[0072] The aseptic filling process control system for nitrogen-filled protection of midazolam oral solution in this application embodiment is used to implement the aforementioned aseptic filling process control method for nitrogen-filled protection of midazolam oral solution. Therefore, the specific implementation of the aseptic filling process control system for nitrogen-filled protection of midazolam oral solution can be found in the embodiment section of the aseptic filling process control method for nitrogen-filled protection of midazolam oral solution mentioned above. The specific implementation can be referred to the description of the corresponding embodiments, and will not be repeated here.

[0073] This application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the aseptic filling process control method for nitrogen protection of midazolam oral solution as described above.

[0074] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the aseptic filling process control method for nitrogen protection of midazolam oral solution described above.

[0075] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0076] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the embodiments of the aseptic filling process control method for nitrogen protection of midazolam oral solution.

[0077] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0078] The above provides a detailed description of the aseptic filling process control method and system for nitrogen-protected midazolam oral solution provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for controlling the aseptic filling process of midazolam oral solution under nitrogen protection, characterized in that, include: After nitrogen filling of midazolam oral solution, a multispectral image sequence of headspace gas inside the bottle before sealing was acquired by a transmission imaging system integrated into the filling line. The spectral image sequence is subjected to spectral separation processing to extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane is generated based on the light intensity attenuation data. Based on the two-dimensional concentration map, the local oxygen partial pressure value corresponding to each pixel is calculated. Combined with the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, a three-dimensional oxygen concentration field of the headspace region inside the bottle is constructed. The oxygen-rich region with excessive concentration is identified according to the preset oxygen concentration threshold. The three-dimensional oxygen concentration field is spatially registered and compared with the ideal nitrogen coverage model by a model reference adaptive controller. Based on the distribution characteristics of the oxygen-rich area, supplementary purging parameters for each bottle are generated. The supplementary purging parameters include purging angle, airflow intensity and purging duration. Based on the supplementary purging parameters, the purging actuator controls the nozzle orientation according to the purging angle, and performs targeted pulse purging on the oxygen-rich area using the purging duration and airflow intensity. After targeted pulse purging, the treated headspace gas inside the bottle is quickly verified and tested. After confirming that the oxygen concentration reaches the predetermined standard, the bottle enters the capping and sealing process.

2. The method according to claim 1, characterized in that, The process involves spatially registering and comparing the three-dimensional oxygen concentration field with an ideal nitrogen coverage model using a model reference adaptive controller. Based on the distribution characteristics of the oxygen-rich region, supplementary purging parameters are generated for each bottle. These supplementary purging parameters include purging angle, airflow intensity, and purging duration. The model reference adaptive controller loads a pre-stored ideal nitrogen coverage model, which defines the target oxygen concentration value at each spatial location in the headspace region inside the bottle. The three-dimensional oxygen concentration field is spatially registered with the ideal nitrogen coverage model to establish a precise correspondence between the three-dimensional spatial coordinates of the two models. Based on the precise correspondence, the measured oxygen concentration values ​​at each location in the three-dimensional oxygen concentration field are compared with the target oxygen concentration values ​​at the corresponding locations in the ideal nitrogen coverage model to identify the concentration difference values. Based on the spatial coordinate range, volume, and maximum oxygen concentration deviation of the oxygen-rich region, supplementary purging parameters are generated for each bottle, combined with the concentration difference value.

3. The method according to claim 2, characterized in that, Based on the spatial coordinate range, volume, and maximum oxygen concentration deviation of the oxygen-rich region, combined with the concentration difference value, supplementary purging parameters are generated for each bottle, including: For each bottle, the azimuth angle of the spatial coordinate range of the oxygen-enriched area relative to the central axis of the bottle is used as the purging angle. The airflow intensity is determined based on the concentration difference value and the maximum oxygen concentration deviation value. The weighted sum of the volume of the oxygen-enriched area and the concentration difference value is used as the purging duration. The combination of the purging angle, airflow intensity, and purging duration is used as supplementary purging parameters for the corresponding bottle.

4. The method according to claim 1, characterized in that, The purging actuator, driven by the supplementary purging parameters, controls the nozzle orientation according to the purging angle, performs targeted pulse purging on the oxygen-rich area using the purging duration and airflow intensity, and after targeted pulse purging, performs rapid verification testing on the treated headspace gas inside the bottle. Once the oxygen concentration is confirmed to meet the predetermined standard, the bottle proceeds to the capping and sealing process, including: Based on the supplementary purging parameters, a control signal is sent to the purging actuator to drive the purging actuator to start operation; According to the purging angle in the supplementary purging parameters, the nozzle is rotated to the position corresponding to the purging angle by the angle adjustment component of the purging actuator, so that the nozzle is aligned with the spatial position of the oxygen-rich area. Adjust the nitrogen output pressure of the purging actuator according to the airflow intensity in the supplementary purging parameters, control the duration of nitrogen output according to the purging duration in the supplementary purging parameters, and inject nitrogen into the oxygen-rich area to complete the fixed-point pulse purging. After the fixed-point pulse purging is completed, the spectral image of the headspace region inside the bottle is acquired, the light intensity attenuation data of the characteristic absorption band of oxygen molecules is extracted, and the current oxygen concentration value is calculated. The current oxygen concentration value is compared with a preset oxygen concentration threshold. If the current oxygen concentration value reaches the preset oxygen concentration threshold, a sealing control signal is triggered to cause the bottle to enter the capping and sealing process.

5. The method according to claim 1, characterized in that, Based on the two-dimensional concentration map, the local oxygen partial pressure value corresponding to each pixel is calculated. A three-dimensional oxygen concentration field is constructed in the headspace region inside the bottle, combining this with the pre-stored bottle geometry parameters corresponding to the midazolam oral solution. Oxygen-rich areas exceeding the concentration limit are identified according to a preset oxygen concentration threshold, including: Based on the light intensity attenuation data of each pixel in the two-dimensional concentration map, the local oxygen partial pressure value corresponding to each pixel is calculated, and the pre-stored bottle geometry parameters of midazolam oral solution are obtained. Based on the local oxygen partial pressure value and the bottle body geometric parameters, a three-dimensional coordinate system with the bottle mouth plane as the reference is established, and the headspace region inside the bottle is divided into multiple layered planes at preset intervals along the height direction of the bottle body. On each layered plane, based on the variation law of the radial dimension of the bottle body, the oxygen partial pressure value of the bottle mouth plane is mapped to the corresponding position of each layered plane through the bilinear interpolation algorithm to obtain the oxygen partial pressure value of the three-dimensional spatial grid node; Based on the oxygen partial pressure values ​​of all three-dimensional spatial grid nodes, a continuous three-dimensional oxygen concentration field is generated using a three-dimensional data reconstruction algorithm. The oxygen concentration value of each node in the three-dimensional oxygen concentration field is compared with the preset oxygen concentration threshold point by point. The three-dimensional spatial regions where the oxygen concentration value exceeds the threshold are identified as oxygen-rich regions. The spatial coordinate range, volume, and maximum oxygen concentration deviation of each oxygen-rich region are recorded.

6. The method according to claim 1, characterized in that, After nitrogen filling of the midazolam oral solution, a multispectral image sequence of the headspace gas inside the bottle before sealing is acquired using a transmission imaging system integrated into the filling line, including: After nitrogen purging of midazolam oral solution, a transmission imaging system integrated on the filling line is activated. The transmission imaging system includes a hyperspectral linear array camera and a specific band of LED array light source. The LED array light source is controlled to sequentially illuminate the narrowband light-emitting units corresponding to the characteristic absorption bands of oxygen molecules in a preset order; The hyperspectral linear array camera is synchronously triggered to collect the transmitted light signal of the corresponding band of the head air gas inside the bottle when each narrowband light-emitting unit is lit, and convert the transmitted light signal into digital image data. The digital image data acquired in each band are combined in spectral band order to generate a multispectral image sequence containing information on the characteristic absorption bands of oxygen molecules.

7. The method according to claim 1, characterized in that, The step of performing spectral separation processing on the spectral image sequence to extract light intensity attenuation data of the characteristic absorption bands of oxygen molecules, and generating a two-dimensional concentration map reflecting the oxygen distribution at the bottle opening plane based on the light intensity attenuation data, includes: The multispectral image sequence is subjected to spectral separation processing to identify and extract digital image data corresponding to the characteristic absorption bands of oxygen molecules; The light intensity value of each pixel is obtained from the digital image data, and the attenuation of the light intensity value of each pixel relative to the initial light intensity value of the corresponding band is calculated to obtain the light intensity attenuation data of each pixel. The position of each pixel is mapped one-to-one with the two-dimensional coordinates of the bottle opening plane. The light intensity attenuation data corresponding to all the two-dimensional coordinates are arranged in the order of the coordinate distribution of the bottle opening plane to generate a two-dimensional concentration map reflecting the oxygen distribution of the bottle opening plane.

8. A nitrogen-filled aseptic filling process control system for midazolam oral solution, characterized in that, include: The acquisition module is used to acquire multispectral image sequences of headspace gas inside the bottle before sealing by a transmission imaging system integrated on the filling line after nitrogen filling of midazolam oral solution. The processing module is used to perform spectral separation processing on the spectral image sequence, extract the light intensity attenuation data of the characteristic absorption band of oxygen molecules, and generate a two-dimensional concentration map reflecting the oxygen distribution in the bottle mouth plane based on the light intensity attenuation data. The construction module is used to calculate the local oxygen partial pressure value corresponding to each pixel based on the two-dimensional concentration map, construct a three-dimensional oxygen concentration field in the headspace region inside the bottle by combining the pre-stored bottle geometry parameters corresponding to the midazolam oral solution, and identify the oxygen-rich areas with excessive concentration according to the preset oxygen concentration threshold. The generation module is used to spatially register and compare the three-dimensional oxygen concentration field with the ideal nitrogen coverage model through a model reference adaptive controller, and generate supplementary purging parameters for each bottle based on the distribution characteristics of the oxygen-rich area. The supplementary purging parameters include purging angle, airflow intensity and purging duration. The verification module is used to drive the purging actuator to control the nozzle orientation according to the purging angle based on the supplementary purging parameters, and to perform fixed-point pulse purging on the oxygen-rich area using the purging duration and airflow intensity. After the fixed-point pulse purging, the treated headspace gas in the bottle is quickly verified and tested. After confirming that the oxygen concentration reaches the predetermined standard, the bottle enters the capping and sealing process.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the aseptic filling process control method for nitrogen protection of midazolam oral solution as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables a method for controlling the aseptic filling process of midazolam oral solution under nitrogen protection as described in any one of claims 1 to 7.

Images