Liquid medicine plastic bottle production control system based on blow-fill-seal sterile integration

Abstract

This invention relates to the field of automated control in pharmaceutical packaging, and discloses a production control system for aseptic blow-fill-seal integrated pharmaceutical liquid plastic bottles. The system includes a data integration management module, a collaborative optimization control module, a process distribution execution module, and a process online monitoring module. The process online monitoring module collects key quality attributes in real time and generates early warning signals; the collaborative optimization control module calculates and dynamically updates the optimal control setpoints and work cycle sequence based on the early warning signals; the process distribution execution module receives the setpoints, executes closed-loop control of the blow-fill-seal process, and provides feedback on actual measured values; the data integration management module collects underlying raw data, generates electronic batch records, and uploads them. This invention achieves dynamic adaptive adjustment of process parameters based on quality feedback, improves product quality consistency, and meets the requirements for complete data traceability throughout the entire process of aseptic drugs.

Description

Technical Field

[0001] This invention relates to the field of automated control of pharmaceutical packaging, specifically to a production control system for pharmaceutical liquid plastic bottles based on an integrated blow-fill-seal aseptic system. Background Technology

[0002] Currently, the production and packaging of sterile pharmaceutical solutions require high levels of cleanliness and production efficiency. Blow-fill-seal technology integrates plastic molding, pharmaceutical filling, and container sealing into a single, continuous process within a single sterile device. This integrated process reduces the risk of external environmental contamination. As pharmaceutical industry standards continue to evolve, regulatory authorities are imposing stricter standards on the process stability and data traceability capabilities of drug production.

[0003] For the aseptic liquid packaging process described above, conventional blow-fill-seal equipment typically relies on a basic logic controller combined with a human-machine interface for operation and control. Before starting the machine, operators manually input the initial settings for each station. During operation, the underlying controllers of each distributed process execute mechanical actions such as heating and plasticizing, mold opening and closing, liquid filling, and sealing according to fixed parameter instructions. After the production task is completed, staff summarize the process operation data through partial system export or manual transcription.

[0004] Traditional fixed-parameter execution modes are essentially passive control. Fluctuations in the external environment or raw material properties make it difficult for a single loop to adjust in a timely manner. This easily leads to process drift, resulting in uneven wall thickness or filling volume deviations. Temperature and pressure in the bottle-sealing process are clearly coupled; existing independent single loops lack multi-variable collaborative calculations, making it difficult to guarantee the final sealing strength of the container. Information barriers often exist between field equipment and management systems, and underlying process and quality data lack precise timestamp synchronization. Numerous isolated data silos cannot automatically generate tamper-proof structured electronic batch records. Therefore, the full-process traceability of sterile pharmaceuticals faces significant obstacles.

[0005] Therefore, the present invention provides a production control system for pharmaceutical plastic bottles based on blow-fill-seal aseptic integration to overcome the shortcomings of the prior art. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a blow-fill-seal aseptic integrated pharmaceutical bottle production control system, which solves the problems of poor process stability and insufficient data traceability and compliance caused by the lack of dynamic control and integrated data management based on real-time quality feedback in traditional blow-fill-seal production control systems.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A production control system for pharmaceutical plastic bottles based on blow-fill-seal aseptic integration includes:

[0009] The data integration and management module is used to receive production instruction data packets issued by the manufacturing execution system and collect raw data records from the process distribution execution module and the process online monitoring module. It is used to generate electronic batch records and upload them to the manufacturing execution system to complete the full-process data traceability and closed-loop management of production batches.

[0010] The collaborative optimization control module is used to parse the production instruction data packet to obtain the formula parameters, calculate the optimal control setpoint and generate the work cycle timing for distribution, and receive the early warning signal to update the optimal control setpoint and re-distribute it.

[0011] The process distribution execution module is used to receive the optimal control setpoint and the operation cycle sequence, execute the closed-loop control of the blow-fill-seal process of the liquid medicine plastic bottle in an aseptic environment, and feed back the actual measured values ​​of the process parameters to the collaborative optimization control module.

[0012] The process online monitoring module is used to collect raw data of key quality attributes in real time, generate the early warning signal and feed it back to the collaborative optimization control module, which is used to trigger the dynamic update of the optimal control setpoint, thereby dynamically adjusting the closed-loop control of the blow-fill-seal process.

[0013] By adopting the above technical solution, the system integrates a data integration management module, a collaborative optimization control module, a process distribution execution module, and a process online monitoring module in its logical architecture. The process online monitoring module captures quality attribute deviations in real time and generates early warnings. Based on this, the collaborative optimization control module performs forward-looking calculations and dynamically updates the control setpoints. The process distribution execution module performs specific parameter adjustments. Therefore, the traditional passive control with a single setting is transformed into a dynamic adaptive adjustment effect based on quality feedback. This continuously suppresses process drift within the closed loop and achieves complete traceability from the underlying production data to the upper-level management system.

[0014] Preferably, the collaborative optimization control module parses the production instruction data packet to obtain the formula parameters, specifically including:

[0015] The production instruction parsing unit in the collaborative optimization control module starts a parsing program on the received production instruction data packet, extracts the production batch number, product code and formula version number contained in the production instruction data packet and performs the parsing action;

[0016] The production instruction parsing unit verifies the format and data type of the extracted fields, generates a confirmation receipt, and sends it back to the manufacturing execution system.

[0017] The production instruction parsing unit uses the parsed product code and formula version number as composite query keywords to retrieve and read matching formula records from the formula database and load them into the system's dynamic memory to obtain the formula parameters that cover detailed process parameters.

[0018] By adopting the above technical solutions, the automatic parsing and verification of production instructions can be achieved, ensuring that the underlying control system can accurately call and load the process formula corresponding to the product specifications and versions, providing an accurate initial operating benchmark for multi-variety switching production.

[0019] Preferably, the process distribution execution module receives the optimal control setpoint and executes closed-loop control, and the feedback of actual measured values ​​of process parameters specifically includes:

[0020] The feeding and plasticizing control unit in the process distribution execution module collects the actual measured values ​​of the melt temperature of the extruder barrel and the actual measured values ​​of the screw speed in real time, and feeds them back to the collaborative optimization control module as the actual measured values ​​of the process parameters.

[0021] The feeding and plasticizing control unit receives the adjustment amount of the temperature setpoint or screw speed setpoint calculated by the collaborative optimization control module based on the deviation between the actual measured value and the target setpoint of the process parameters;

[0022] The feeding and plasticizing control unit uses the received adjustment amount as the updated optimal control setting value to control the heating zone of the extruder barrel and the screw drive motor to perform the closed-loop control to maintain the constant quality of the tubular preform.

[0023] By adopting the above technical solution, a closed-loop regulation mechanism for extruder temperature and speed is established to compensate for the effects of batch differences in polymer raw materials or fluctuations in ambient temperature in real time, and to maintain the consistency of the molten state and physical dimensions of the extruded tubular preform.

[0024] Preferably, the process distribution execution module receiving the closed-loop control of the job cycle timing execution specifically includes:

[0025] The blow-fill-seal core control unit in the process distribution execution module measures the current actual blow molding pressure and compares the actual blow molding pressure with the target pressure set value to calculate the pressure error;

[0026] The blow-fill-seal core control unit receives a series of millisecond-level operation cycle timing signals from the global timing arbitration unit within the collaborative optimization control module;

[0027] The blow-fill-seal core control unit automatically switches to the parameter group of the corresponding stage based on the received operation cycle timing signal, calculates the adjustment amount of the proportional control valve opening based on the pressure error, and executes the closed-loop control in the manner of tracking the pressure curve.

[0028] By adopting the above technical solution, millisecond-level timing signals are combined with dynamic pressure error calculation to achieve accurate tracking of blow molding pressure curves at different molding stages, ensuring the uniformity of the wall thickness distribution of the molded container.

[0029] Preferably, the process distribution execution module receiving the optimal control setpoint and executing closed-loop control specifically includes:

[0030] The blow-fill-seal core control unit in the process distribution execution module executes the liquid filling program and controls the injection of liquid into the formed container;

[0031] The blow-fill-seal core control unit receives the control compensation amount calculated by the collaborative optimization control module based on the deviation between the actual measured weight data and the target filling weight;

[0032] The blow-fill-seal core control unit uses the received control compensation amount as the updated optimal control setting value, and automatically adjusts the execution parameters of the next filling cycle to perform the closed-loop control.

[0033] By adopting the above technical solution, a dynamic compensation algorithm is constructed based on front-end online weighing feedback to effectively eliminate systematic filling drift caused by changes in liquid physical properties or wear of mechanical parts, thereby improving the control accuracy of drug filling volume.

[0034] Preferably, the collaborative optimization control module receiving the actual measured values ​​of the process parameters to update the optimal control setpoint specifically includes:

[0035] The advanced process control algorithm unit in the collaborative optimization control module receives the actual measured values ​​of the process parameters, such as sealing temperature and sealing pressure, which are monitored and uploaded in real time by the process distribution execution module.

[0036] The advanced process control algorithm unit uses a pre-established multiple linear regression prediction model for sealing strength, substitutes the actual measured values ​​of the process parameters, and calculates the predicted value of the sealing strength in real time.

[0037] If the predicted value shows a trend of deviating from the target intensity, the advanced process control algorithm unit calculates the parameter instructions for multi-variable coordinated adjustment based on the model, and sends out the adjusted parameter instructions as the updated optimal control setpoint.

[0038] By adopting the above technical solution, the mutually coupled sealing temperature and sealing pressure are input into the prediction model for forward calculation, and multi-variable collaborative adjustment instructions are output, which breaks through the limitations of single-parameter control and ensures the stability of the molten seal strength at the bottle mouth.

[0039] Preferably, the process online monitoring module collects raw data of key quality attributes in real time, specifically including:

[0040] The key quality attribute sensing and acquisition unit in the process online monitoring module acquires weight data from a high-speed dynamic weighing sensor and temperature data from a non-contact infrared thermometer at a frequency synchronized with the production line cycle as the real-time acquisition action.

[0041] The key quality attribute sensing and acquisition unit internally converts the acquired data into digital values ​​with specific units;

[0042] The key quality attribute sensing and acquisition unit adds a unique production serial number and a unified timestamp to the digitized value to generate complete raw data of the key quality attribute.

[0043] By adopting the above technical solutions, a high-frequency sensing acquisition mechanism that is strictly synchronized with the operation of production equipment is established to ensure the originality, timeliness, and sequence correspondence of key quality attribute data, providing a reliable data source for subsequent trend analysis.

[0044] Preferably, the process online monitoring module generates an early warning signal and feeds it back to the collaborative optimization control module, specifically including:

[0045] The statistical process control trend analysis unit in the online process monitoring module receives continuous data streams to construct statistical process control charts and applies preset anomaly detection rules to perform real-time pattern matching on the data point sequence to identify non-random fluctuations.

[0046] When a data point is detected to exceed the preset control limit or violate the preset discrimination rule, the real-time early warning feedback unit in the online process monitoring module is activated to generate the structured early warning signal.

[0047] The real-time early warning feedback unit actively feeds back the early warning signal containing specific deviation information to the central coordinating controller in the collaborative optimization control module, providing a basis for the adaptive adjustment of the process parameters of the collaborative optimization control module.

[0048] By adopting the above technical solutions, the statistical process control principle is used to identify small deviations or amplifications in the mean of the production process, and early warnings are generated before quality indicators exceed the qualified limits, thus achieving preventive intervention in quality control.

[0049] Preferably, the data integration management module collects the raw data records from the process distribution execution module and the online process monitoring module, specifically including:

[0050] The production data acquisition and monitoring system data acquisition unit in the data integration management module continuously communicates data via the industrial bus, periodically acquiring the process parameters and equipment status of the process distribution execution module to perform the acquisition action;

[0051] The production data acquisition and monitoring system's data acquisition unit synchronously acquires key quality attribute data from the process online monitoring module and performs the acquisition action.

[0052] The production data acquisition and monitoring system's data acquisition unit adds a unified clock to all acquired data for time synchronization and stores it in a real-time database to form a complete record of the original data.

[0053] By adopting the above technical solution, multi-source heterogeneous data distributed across different nodes are centrally collected and unified globally in time, eliminating information silos and constructing a production process underlying dataset with a clear timeline.

[0054] Preferably, the data integration management module generates electronic batch records and uploads them to the manufacturing execution system, specifically including:

[0055] The electronic batch record automatic generation unit in the data integration management module retrieves the original data records related to the batch from the real-time database and integrates them according to a predefined structured template to generate the electronic batch record.

[0056] The manufacturing execution system bidirectional interface unit in the data integration management module packages the generated electronic batch records and summary reports containing key production indicators into a standardized data package;

[0057] The bidirectional interface unit of the manufacturing execution system sends the data packet to the application programming interface of the manufacturing execution system, and completes the upload to the manufacturing execution system after receiving the confirmation message.

[0058] By adopting the above technical solutions, tamper-proof structured electronic documents are automatically integrated and generated, opening up the data interaction link between the underlying production site and the upper manufacturing execution system, and realizing paperless circulation of compliant data and closed-loop management of the entire production cycle.

[0059] This invention provides a blow-fill-seal aseptic integrated production control system for pharmaceutical plastic bottles. It offers the following advantages:

[0060] 1. This invention uses an online process monitoring module to collect key quality attribute data in real time and generates early warning signals based on statistical process control trends. Subsequently, a collaborative optimization control module dynamically calculates and updates control setpoints based on this data, which are then sent to the distributed process execution module for specific parameter adjustments. This dynamic closed-loop mechanism based on quality data feedback changes the traditional passive operation mode with fixed parameters. It can compensate for minor deviations caused by external interference or changes in equipment status during production, suppressing process parameter drift and thus improving the quality consistency of the same batch of pharmaceutical liquid plastic bottles.

[0061] 2. The collaborative optimization control module of this invention can generate and issue millisecond-level work cycle timings. Simultaneously, it utilizes an advanced process control algorithm unit to establish a predictive model, collaboratively calculating variables with coupled relationships, such as sealing temperature and pressure. The process distribution execution module, based on high-precision timing signals and multi-variable adjustment instructions, dynamically tracks the blow molding pressure curve, performs weight compensation for the next filling cycle, and collaboratively adjusts sealing parameters. This overcomes the limitations of independent single-loop control, achieving precise control over container molding wall thickness, liquid filling volume, and bottle mouth sealing strength, ensuring the execution accuracy of all core blow-fill-seal processes.

[0062] 3. This invention incorporates a data integration and management module that unifies and centrally collects multi-source, heterogeneous process parameters and quality data using unified timestamps. It then automatically generates structured electronic batch records based on preset templates and uploads them to the manufacturing execution system. This establishes a seamless data exchange link between the field control layer and the enterprise management layer, eliminating data gaps in the production process. It enables paperless transfer and tamper-proof storage of batch production data, meeting the requirements for complete data traceability and compliant control throughout the sterile drug manufacturing process. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the architecture of the integrated blow-fill-seal aseptic pharmaceutical bottle production control system of the present invention.

[0064] Figure 2 This is a flowchart of the production control method for pharmaceutical plastic bottles based on blow-fill-seal aseptic integration according to the present invention;

[0065] Figure 3 This is a flowchart illustrating the system initialization and command interaction of the present invention.

[0066] Figure 4 This is a schematic diagram of the closed-loop control principle of the core process of blow-fill-seal in this invention;

[0067] Figure 5 This is a schematic diagram illustrating the key quality attribute trend analysis and early warning based on statistical process control of the present invention.

[0068] Figure 6 This is a flowchart of the data acquisition and electronic batch record generation process of the present invention;

[0069] Figure 7 A simulation diagram showing the effect comparison between the advanced process control of drug liquid filling weight of the present invention and the traditional control method;

[0070] Figure 8 This is a graph showing the statistical process control and exponentially weighted moving average analysis results of the final sealing temperature monitoring data of this invention. Detailed Implementation

[0071] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] See attached document Figure 1 This invention provides a production control system for pharmaceutical plastic bottles based on blow-fill-seal aseptic integration. The system includes: a collaborative optimization control module, a process distribution execution module, a process online monitoring module, and a data integration management module.

[0073] The collaborative optimization control module is the system's top-level decision-making and coordination mechanism, comprising a central coordination controller. This central coordination controller, in its implementation, can utilize a high-performance industrial computer or a large-scale programmable logic controller (PLC), possessing powerful data processing and computational capabilities. Internally, the central coordination controller is logically divided into a production instruction parsing unit, an advanced process control algorithm unit, and a global timing arbitration unit. These three units work collaboratively, forming the foundation for centralized management and intelligent optimization of the entire blow-fill-seal production process.

[0074] The production instruction parsing unit is responsible for communication between this system and the upper-level manufacturing execution system. It receives production work order data packets from the manufacturing execution system via an industrial Ethernet interface. This unit performs syntax parsing and semantic verification on the received data packets, extracting key production information such as production batch number, product code, formula version, and planned output. After parsing, this unit distributes the structured production instructions to the advanced process control algorithm unit and the global timing arbitration unit, providing initial settings for subsequent process control and timing coordination.

[0075] The advanced process control algorithm unit is the core functional unit for achieving optimized process control. This unit pre-stores advanced process control models for different product specifications, built upon historical production data and process mechanisms. During production, this unit receives real-time process parameters and key quality attribute data from the online process monitoring module, using these as inputs. By running the advanced process control algorithm, it calculates the optimal process parameter setpoints, such as target temperatures for each extruder zone, screw speed, blow molding pressure curve, filling time, and sealing temperature, and dynamically distributes these setpoints to the distributed process execution module.

[0076] The global timing arbitration unit is responsible for generating and broadcasting a unified, high-precision timing reference signal for the entire blow-fill-seal production line. Based on the product specification information provided by the production instruction parsing unit, it retrieves the corresponding work cycle time, such as 15 seconds, from its internal database. Using this cycle as a reference, the unit generates a series of timing control instructions accurate to the millisecond level. These instructions precisely coordinate the start / stop, sequence, and duration of actions of various units in the process distribution execution module, ensuring that every step from preform extrusion to finished product output is executed precisely within the predetermined time window, achieving seamless multi-axis, multi-station collaborative operation.

[0077] The distributed process execution module consists of a set of independent hardware control modules distributed across the production site, responsible for the execution and closed-loop control of specific process actions. This module includes a material feeding and plasticizing control unit, a blow-fill-seal core control unit, an online detection and rejection unit, and a sterile environment control unit. Each unit communicates bidirectionally with the central coordinating controller via a real-time industrial bus, receiving control commands and setpoints, and feeding back its own operating status and process parameters.

[0078] The feeding and plasticizing control unit is specifically designed to control the entire process of medical-grade polymer granules melting and plasticizing from the hopper to the extruder. This unit manages the automatic delivery of raw materials by controlling the start and stop of the vacuum feeder. Internally, it integrates multiple independent temperature control loops, precisely adjusting the power output of multiple heating zones on the extruder barrel via solid-state relays or SCRs, while simultaneously receiving temperature feedback from thermocouples, forming a closed-loop control system to strictly adhere to the multi-segment temperature curves set by the advanced process control algorithm unit. Furthermore, this unit precisely controls the speed of the extruder's main motor via a frequency converter, thereby stabilizing the screw speed and ensuring a high degree of consistency in the dimensions, temperature, and viscosity of the extruded tubular preform.

[0079] The blow-fill-seal core control unit is the direct controller that performs the three key steps of blow molding, aseptic filling, and melt sealing. This unit receives precise timing instructions from the global timing arbitration unit and sequentially drives the mold opening and closing mechanism, preform cutter, blow molding pneumatic solenoid valve, liquid filling pump or time-pressure filling system, and thermoforming sealing head, among other actuators, within a continuous work cycle. This unit performs precise closed-loop control of key process parameters such as blow molding pressure, filling volume, sealing temperature, sealing pressure, and cooling time, ensuring that the molding quality, liquid filling accuracy, and seal integrity of each plastic bottle meet quality standards.

[0080] The online inspection and rejection unit integrates and controls non-contact online quality inspection equipment located on the finished product conveying path. Specifically, this unit is directly connected to high-voltage discharge leak detection equipment and high-speed dynamic weighing equipment, collecting leak detection signals and weight data for each finished bottle in real time. The unit has internally set criteria for acceptable products, such as the absence of a high-voltage discharge breakdown signal and a weight value within preset upper and lower tolerance limits. Once any indicator is detected as not meeting the criteria, the unit immediately sends an action command to the rejection mechanism, such as a high-speed pneumatic pusher, to precisely remove the defective product from the main production line to a designated collection box.

[0081] The aseptic environment control unit is responsible for dynamically monitoring and maintaining a Class A clean environment in the core operating area. This unit connects to and processes real-time signals from wind speed sensors, differential pressure sensors, temperature and humidity sensors, and a particulate matter counter. By controlling the fan speed of the HEPA filter air supply unit, this unit maintains the wind speed and differential pressure in critical areas within set ranges. Simultaneously, before or after a production task, this unit automatically executes preset online cleaning and sterilization programs according to instructions from the central coordinating controller. By controlling relevant valves, pumps, and steam generators, it automates the cleaning and sterilization of components in contact with the pharmaceutical solution, and records key parameters of the entire process for auditing purposes.

[0082] The online process monitoring module utilizes process analysis technology to achieve real-time, non-destructive monitoring, analysis, and early warning of key quality attributes during production. This module includes a key quality attribute sensor acquisition unit, a statistical process control trend analysis unit, and a real-time early warning feedback unit. Through deep process insight, this module enables a shift from passive product inspection to proactive quality prevention.

[0083] The key quality attribute sensing and acquisition unit is the source of process data, integrating a series of high-precision, high-frequency sensors. For example, at the filling station, high-precision load cells or Coriolis mass flow meters with a resolution of 0.001 grams are installed to directly measure the filling volume of each container; at the heat-sealing station, non-contact infrared thermometers with millisecond-level response times are installed to capture the surface temperature of the mold or container at the moment of sealing. This unit acquires the raw data from these sensors at a sampling frequency far exceeding the process control cycle, and after digital processing, transmits it to the statistical process control trend analysis unit.

[0084] The Statistical Process Control Trend Analysis Unit is a software-based functional unit that receives continuous data streams from the sensor acquisition unit. This unit establishes a statistical process control model for each monitored key quality attribute, such as a mean-range control chart or a mean-standard deviation control chart. It calculates the statistics for each data point in real time and plots them on the corresponding control chart. Simultaneously, this unit incorporates various trend determination rules, such as Nielsen rules, to identify non-random trends in the data sequence, such as continuous multiple deviations from the centerline, continuous increases or decreases, and other potential signs of process drift.

[0085] The real-time early warning feedback unit executes early warning logic based on the analysis results of the statistical process control trend analysis unit. This unit has multiple preset alarm thresholds, including warning limits and control limits. When a monitored data point exceeds the warning limit but remains within the control limit, or when a specific trend judgment rule is triggered, the unit generates a low-level warning signal to alert the operator. When a data point directly exceeds the control limit, a high-level alarm is generated. These warning and alarm signals are not only displayed on the human-machine interface but are also packaged into signals containing specific numerical values ​​and status information, actively fed back to the central coordinating controller in the collaborative optimization control module, providing a decision-making basis for adaptive adjustment of process parameters.

[0086] The data integration and management module serves as the data hub of the entire system, responsible for the collection, integration, storage, and interaction of data throughout the entire production process. This module ensures data integrity and traceability and establishes an information channel between the production execution layer and the enterprise management layer. Specifically, it includes a production data acquisition and monitoring unit, an automatic electronic batch record generation unit, and a two-way interface unit with the manufacturing execution system.

[0087] The production data acquisition and monitoring unit is built on a standard production data acquisition and monitoring system platform. This unit uses a unified communication protocol to periodically poll or subscribe to all data points from the distributed process execution module and the online process monitoring module. It comprehensively collects all production-related information, including real-time and setpoint values ​​of process parameters, equipment operating status, alarm information, operator action records, monitoring data of key quality attributes, and online test results. All collected data is appended with precise timestamps and stored in a highly reliable real-time database, forming raw data records.

[0088] The automatic electronic batch record generation unit is triggered upon completion of each production batch. This unit retrieves all data records related to the specified batch number from the real-time database and integrates them in a structured manner according to a predefined electronic batch record template that conforms to Good Manufacturing Practice (GMP) requirements. During the integration process, the unit applies data integrity verification algorithms, such as generating hash signatures for critical data chains, to ensure the immutability of the records. Ultimately, the unit generates a comprehensive, formatted, and traceable electronic batch record file that details the entire process from production preparation to batch completion.

[0089] The Manufacturing Execution System (MES) bidirectional interface unit provides a standardized data exchange channel between this system and the enterprise's upper-level MES, typically employing industrial communication standards such as the OPC unified architecture. This unit possesses bidirectional communication capabilities: on one hand, it receives production instructions from the MES and transmits them to the collaborative optimization control module; on the other hand, it proactively uploads information such as complete electronic batch records generated by the electronic batch record automatic generation unit, real-time calculated equipment overall efficiency data, material consumption statistics, and batch production quality reports to the MES, achieving seamless data linkage and closed-loop management between the shop floor level and the enterprise management level.

[0090] See attached document Figure 2 The present invention also provides a method for controlling the production of pharmaceutical plastic bottles based on the integrated blow-fill-seal aseptic process.

[0091] The method first executes step S100: system initialization and production instruction parsing. In this step, the bidirectional interface unit of the manufacturing execution system in the data integration management module receives the production instruction data packet from the upper-level manufacturing execution system. The production instruction parsing unit in the collaborative optimization control module then parses the data packet, extracting information such as the production batch number, product formula identifier, and planned output. Subsequently, based on the formula identifier, the production instruction parsing unit retrieves and loads a formula file containing detailed process parameters from the local database. These parameters include extrusion temperature, blow molding pressure, filling volume setpoint, and sealing temperature. The central coordination controller of the collaborative optimization control module instructs the aseptic environment control unit in the process distribution execution module to automatically execute the pre-production preparation procedures, including equipment self-checks and initiating online cleaning and online sterilization programs. After the program execution is complete, the system confirms that the aseptic environment parameters and equipment status meet the preset startup conditions, completing the initialization.

[0092] After system initialization, the method proceeds to step S200: Multivariate Collaborative Optimization and Closed-Loop Control of the Blow-Fill-Seal Process. This step is the core execution stage for continuous production of pharmaceutical plastic bottles. The global timing arbitration unit in the collaborative optimization control module generates and broadcasts a precise work cycle timing, such as 15 seconds, based on the loaded formula parameters. Simultaneously, the advanced process control algorithm unit continuously calculates and dynamically updates the optimal control settings for each process step based on real-time feedback process parameters and data provided by the online process monitoring module. These settings and timing instructions are then sent to the process distribution execution module. The feeding and plasticizing control unit precisely controls the supply of polymer granules and the temperature and screw speed of the extruder according to the instructions, stabilizing the extrusion of the tubular preform. The blow-fill-seal core control unit strictly follows the global timing, precisely coordinating a series of actions within one work cycle, including mold opening and closing, preform cutting, container blow molding, precise quantitative filling of the pharmaceutical liquid, and melt sealing of the bottle neck. After finished product output, the online detection and rejection unit performs 100% high-voltage discharge leak detection and dynamic weighing detection on each product, automatically rejecting any unit that does not meet the preset quality standards.

[0093] Parallel to step S200 is step S300: online monitoring and proactive early warning of key quality attributes. In this step, the online process monitoring module continuously monitors key quality attributes in the production process in real time. Specifically, the key quality attribute sensing and acquisition unit acquires raw data on filling volume and final sealing temperature at a high speed of several times per second using a high-precision weighing sensor installed at the filling station and a non-contact infrared thermometer installed at the sealing station. The statistical process control trend analysis unit receives these continuous data streams and constructs and updates its statistical process control chart in real time. This unit applies preset statistical rules to analyze data trends to identify small drifts or increases in variation in the process mean. Once the monitored data reaches the warning limit or shows a continuous deviation from the centerline, the real-time early warning feedback unit immediately generates an early warning signal and feeds back the signal containing specific deviation information to the advanced process control algorithm unit of the collaborative optimization control module for proactive fine-tuning of process parameters, thereby intervening proactively before quality defects occur.

[0094] Throughout the entire production process, the method continuously executes step S400: the full-process acquisition of production data and generation of electronic batch records. The production data acquisition and monitoring unit in the data integration management module collects and synchronizes in real time the operating parameters, equipment status, alarm information, and key quality attribute data collected by the process online monitoring module from all distributed execution units via the industrial bus. Each data record is appended with a unique timestamp and relevant contextual information and stored in the database to ensure data integrity and traceability. When a production batch is completed, the automatic generation unit of the electronic batch record is triggered. This unit retrieves all data related to that batch from the database and automatically integrates it according to a predefined structured template to generate a complete electronic batch record that conforms to data integrity principles. Finally, the manufacturing execution system bidirectional interface unit uploads this generated electronic batch record, along with the calculated overall equipment efficiency and other production reports, to the manufacturing execution system via a standard interface, completing the data-level closed loop for that production batch.

[0095] See attached document Figure 3 The method for controlling the production of pharmaceutical plastic bottles based on blow-fill-seal aseptic integration provided in this embodiment of the invention specifically performs the following sub-steps in its system initialization and production instruction parsing step S100:

[0096] First, sub-step S101 is executed to receive and parse the production instructions issued by the Manufacturing Execution System (MES). The MES bidirectional interface unit in the data integration management module listens to a preset communication port via an industrial Ethernet interface to receive production instruction data packets sent from the MES. These data packets use a predefined structured data format, such as a data structure based on Extensible Markup Language (EXPLAIN) or JavaScript object representation. Once the production instruction parsing unit in the collaborative optimization control module receives the complete data packet, it initiates the parsing process, extracting the key information fields contained within the data packet according to the defined data structure.

[0097] The production instruction parsing unit validates the extracted field content. These fields specifically include: a production batch number used as a unique batch identifier, such as a string consisting of letters and numbers; a product code used to associate with the process formula; the planned production quantity specifying the target quantity for this batch, such as 50,000 units; and a formula version number to ensure the use of the correct version of the process parameters. The production instruction parsing unit verifies whether the format and data type of these fields meet system requirements and confirms that all required fields exist and are not empty.

[0098] After parsing and verification are complete, the production instruction parsing unit generates a confirmation receipt containing the received production batch number and a status code indicating successful parsing. This receipt is sent back to the Manufacturing Execution System (MES) via the MES bidirectional interface unit. This interaction ensures that production instructions are received accurately and establishes a clear starting point for instruction execution between the control system and the enterprise management system. If parsing fails, a status code containing error details is returned, terminating the subsequent process and prompting operator intervention.

[0099] Next, sub-step S102 is executed to call and load the process formula parameters for the corresponding product specification. After successfully parsing the production instruction, the production instruction parsing unit of the collaborative optimization control module uses the product code and formula version number obtained from the instruction as composite query keywords to search the system's formula database. This database stores all verified production process formulas, each formula record corresponding to a unique product code and version number, and contains the complete set of parameters required to start and run the production of that product.

[0100] The central coordinating controller reads the matching formula record from the database and loads all its parameters into the system's dynamic running memory. These parameters are extremely detailed, including but not limited to: the target temperature setpoints for each heating zone of the extruder required by the feeding and plasticizing control unit, such as segmented settings from 180 degrees Celsius to 210 degrees Celsius, and the screw speed setpoints, such as 50 revolutions per minute; the blow molding gas pressure curve, the target filling weight of the liquid, such as 10.20 grams, the heating temperature of the sealing mold, such as 135 degrees Celsius, the pressure applied for sealing, and the time sequence of each action required by the blow-fill-seal core control unit.

[0101] Once loading is complete, the central coordinating controller distributes these parameters. It sends parameters related to material plasticization to the feeding plasticization control unit. It sends the set of parameters related to bottle forming, filling, and sealing to the blow-fill-seal core control unit. It sends the upper and lower limits of the product's acceptable weight, such as 10.15 to 10.25 grams, to the online detection and rejection unit. It sends environmental requirement parameters for the production area, such as suspended particulate concentration limits, to the aseptic environment control unit. This distribution and presetting of parameters ensures that, after production starts, each distributed execution unit can operate according to unified and accurate instructions.

[0102] Sub-step S103 is then executed to automate the pre-production preparation procedures. After the recipe parameters are loaded and distributed, the central coordinating controller automatically triggers a series of pre-production preparation procedures. First, it broadcasts a self-test command to all process distribution execution modules. Upon receiving the command, each unit executes its internal diagnostic procedures, checking whether all connected sensors are online and reading normally, whether motor drivers are ready, and whether the initial positions of all pneumatic or electric actuators are correct. Each unit returns a self-test result status code to the central coordinating controller.

[0103] After all equipment self-tests pass, if required by the formulation or production plan, the central coordinating controller will instruct the aseptic environment control unit to initiate the online cleaning program. Following preset program logic, the aseptic environment control unit precisely controls the opening and closing of a series of valves, sequentially pumping and circulating water for injection, cleaning agent, and final rinse water through all surfaces in contact with the drug solution, including the reservoir, piping system, and filling nozzles. Parameters such as conductivity, temperature, and flow rate are monitored throughout the cleaning process to ensure that the cleaning effect meets the set standards.

[0104] After the online cleaning process is completed, the system automatically enters the online sterilization process. The aseptic environment control unit again controls the relevant valves to introduce clean steam into the cleaned system pipelines. Multiple temperature sensors in the system, especially those located at the coldest point, must maintain their readings above the sterilization target temperature, such as 121.1 degrees Celsius, for a specified time, such as 20 minutes. The time and temperature data of the entire sterilization process are recorded in detail as part of the batch production record, proving that the equipment has reached a sterile state before production begins.

[0105] Finally, sub-step S104 is executed to confirm that the aseptic environment and equipment status meet the preset startup conditions. After the online sterilization process is completed and the system cools to the operating temperature, the aseptic environment control unit continuously monitors the environmental status of the core operating area. Its built-in sensors measure and report key environmental parameters in real time, including suspended particle concentration, unidirectional airflow velocity, and pressure difference with the surrounding area. The central coordinating controller compares these real-time monitoring values ​​with standards defined in the formulation document, such as the dust particle limits required by ISO 5 level of the International Organization for Standardization.

[0106] Simultaneously, the central coordinating controller performs a final system status poll. It confirms that all process distribution execution modules have reported readiness status, that online cleaning and sterilization procedures have been successfully completed and properly logged, and that there are no outstanding Class A or Class B critical alarms in the system. Furthermore, the system checks whether the supply pressure and temperature of utilities required for production, such as compressed air and cooling water, are within normal operating ranges.

[0107] The central coordinating controller will only determine that the system meets the preset startup conditions when all conditions, including aseptic environment parameters, all equipment units, successful execution of critical preparation procedures, and no unresolved serious faults, are met. At this point, the system's overall status will switch from "ready" to "standby," and the indicator light on the human-machine interface will illuminate, awaiting final confirmation from the operator to initiate the continuous production process in step S200. This series of rigorous checks and confirmation steps ensures that each production batch begins in a safe, compliant, and controlled manner.

[0108] See attached document Figure 4 After the system completes initialization and confirms the start-up conditions, the production control method provided in this embodiment of the invention enters step S200, namely the multi-variable collaborative optimization and closed-loop control step of the blow-fill-seal process. This step specifically executes the following sub-steps:

[0109] First, sub-step S201 is executed to stabilize the feeding and plasticizing stages. The feeding and plasticizing control unit in the process distribution execution module, based on the formula parameters issued by the central coordinating controller, initiates the automatic conveying system for medical-grade polymer granules and controls multiple heating zones on the extruder barrel and the screw drive motor. This control aims to produce tubular preforms with stable and uniform melt flow, which is fundamental to subsequent molding quality. The advanced process control algorithm unit of the central coordinating controller has a built-in correlation model between melt flow index and extrusion process parameters, the mathematical expression of which is:

[0110] ;

[0111] in, Melt index is a key indicator characterizing the flowability of polymer melts. This is a material constant, determined by the specific polymer grade; This refers to the screw speed of the extruder. The power-law exponent characterizes the shear-thinning behavior of the material; It represents the activation energy for flow, characterizing the sensitivity of melt viscosity to temperature; It is the ideal gas constant; This is the absolute temperature scale value of the melting temperature.

[0112] During continuous production, the feeding and plasticizing control unit collects real-time measurements of the melt temperature and screw speed using temperature sensors installed in different areas of the extruder barrel and encoders on the screw motor, and uploads this data to the central coordinating controller. The advanced process control algorithm unit then uses this real-time data to calculate the predicted melt index under the current operating conditions using the aforementioned correlation model.

[0113] The central coordinating controller compares the predicted melt index value with the target melt index set in the formulation. When a deviation occurs, the advanced process control algorithm unit calculates the adjustment amount for the melt temperature setpoint or screw speed setpoint based on the deviation and a preset optimization strategy. For example, if the predicted melt index value is too low, indicating insufficient melt fluidity, the algorithm unit calculates the required increase in temperature or screw speed and sends the updated setpoints to the feeding and plasticizing control unit. The feeding and plasticizing control unit executes the new setpoints, thus forming a closed-loop feedback control loop to compensate for the effects of batch-to-batch variations in raw materials or environmental fluctuations, maintaining the constant quality of the tubular preform.

[0114] Next, sub-step S202 is executed to dynamically control the pressure during the preform blow molding stage. Once the tubular preform is extruded to the predetermined length, the blow-fill-seal core control unit in the process distribution execution module controls the mold to close and cut the preform. Subsequently, this unit injects sterile compressed gas into the preform through a precise proportional control valve, causing it to expand and conform to the inner wall of the mold to form a container. To ensure uniform and defect-free container wall thickness, the gas pressure during the blow molding process needs to follow a precise time-pressure curve. This curve is defined in the formulation as a series of pressure setpoints corresponding to a series of time points.

[0115] To accurately track this pressure curve, the blow-fill-seal core control unit employs a piecewise proportional-integral-derivative (PDI-DE) control algorithm. During each control cycle of the blow molding process, a pressure sensor installed on the air supply line measures the current actual blow molding pressure and feeds it back to the blow-fill-seal core control unit. This unit compares the actual pressure measurement with the setpoint on the pressure curve at the current moment and calculates the pressure error. Based on this error, the control algorithm calculates the adjustment amount for the proportional control valve opening, and the calculation of its control output signal follows the formula:

[0116] ;

[0117] in, for The control output at any time, i.e., the control signal for adjusting the blow molding air pressure; This is the proportional gain coefficient; This is the integral gain coefficient; The differential gain coefficient; for Pressure error at any given time is the difference between the target pressure setpoint and the actual measured value.

[0118] Different stages of the blow molding process, such as the pre-blow stage and the main blow stage, have different pressure control response requirements. Therefore, the central coordinating controller applies different combinations of proportional gain, integral gain, and derivative gain coefficients to different stages based on the formula. During execution, the blow-fill-seal core control unit automatically switches to the corresponding parameter set based on the timing signal issued by the global timing arbitration unit. This segmented control strategy ensures rapid system response in stages requiring rapid pressure build-up and minimal overshoot in stages requiring stable pressure holding, thereby achieving accurate reproduction of the entire blow molding pressure curve and guaranteeing the repeatability and consistency of bottle molding.

[0119] Sub-step S203 is then executed for high-precision metering control during the drug filling stage. After the container is blow-molded, the blow-fill-seal core control unit immediately executes the drug filling procedure. The drug is precisely injected into the freshly formed container through a time and pressure control system or a peristaltic pump system. To achieve proactive control and compensation of the filling volume for each container, the key quality attribute sensing and acquisition unit in the process online monitoring module uses high-speed dynamic weighing sensors installed on the production line to perform online, 100% weight detection on each finished product that has been filled and sealed.

[0120] The weighing sensor transmits the actual measured weight data of each product to the central control unit in real time. Upon receiving this data, the advanced process control algorithm unit of the central control unit executes a dynamic filling volume compensation algorithm based on weighing feedback. This algorithm compares the actual weight measured this time with the target filling weight set in the formula, calculates the filling deviation, and updates the control compensation amount for the next filling based on this deviation. The mathematical model of this algorithm is as follows:

[0121] ;

[0122] in, For the first Control compensation amount for each filling; For the first The control compensation amount used for each filling; The compensation coefficient determines the corrected response speed; Set a value for the target filling weight; For the first The actual measured weight of each filling.

[0123] After calculating the new control compensation amount, the advanced process control algorithm unit sends it to the blow-fill-seal core control unit. Based on the received compensation amount, this unit automatically adjusts the execution parameters for the next filling cycle. For example, for time- and pressure-based filling systems, this compensation amount is used to fine-tune the opening time of the filling valve; for peristaltic pump systems, it is used to fine-tune the pump's rotational speed or number of revolutions. This successive compensation closed-loop control method effectively eliminates systematic filling deviations caused by factors such as changes in drug temperature, fluctuations in liquid level and pressure, or wear and tear from long-term pump operation, ensuring that the filling weight of the entire batch of product is highly concentrated near the target value.

[0124] Finally, sub-step S204 is executed to perform coordinated temperature and pressure control during the melt-sealing stage. After the drug solution is filled, the blow-fill-seal core control unit controls the top sealing mold to descend, performing a thermo-pressurized melt seal on the opening at the neck of the container. The sealing quality of this process directly determines the aseptic integrity of the product, and its key process parameters are sealing temperature, sealing pressure, and sealing time. The blow-fill-seal core control unit precisely controls the heating element within the sealing mold through an independent proportional-integral-derivative control loop to maintain a constant sealing temperature. This control is based on the following energy balance model:

[0125] ;

[0126] in, The heat power input to the mold by the heater; The heat power lost by the mold to the environment and materials; To ensure the quality of the heat-affected zone of the sealing mold; The specific heat capacity of the sealing mold material; This represents the rate of change of mold temperature over time.

[0127] Meanwhile, the advanced process control algorithm unit of the central coordinating controller utilizes a pre-established multiple linear regression prediction model for sealing strength to proactively control sealing quality. This model, based on extensive experimental data, correlates key process parameters with the final sealing strength; its mathematical expression is as follows:

[0128] ;

[0129] in, This is the predicted value of the sealing strength; , , , These are the regression coefficients of the model obtained by fitting experimental data; For sealing temperature; For sealing pressure; For sealing time; This represents the model residuals.

[0130] During the production process, the core control unit of the blow-fill-seal system monitors and uploads the actual values ​​of parameters such as sealing temperature and sealing pressure in real time. The advanced process control algorithm unit then inputs these actual values ​​into a prediction model to calculate the predicted sealing strength in real time. If the predicted value deviates from the target strength, the advanced process control algorithm unit can perform multi-variable collaborative adjustments. For example, when a slight decrease in sealing temperature is detected, the algorithm unit can calculate, based on the model, the required increase in sealing pressure or extension of sealing time to compensate for the impact of the temperature drop on sealing strength, and then send the adjusted parameter instructions to the core control unit for execution. This collaborative optimization control strategy elevates the originally independent single-parameter control to multi-variable comprehensive control, enhancing the stability and reliability of the final product's sealing quality.

[0131] See attached document Figure 5 While performing the multivariate collaborative optimization and closed-loop control steps of the blow-fill-seal process, the production control method provided in this embodiment of the invention executes step S300 in parallel, namely the online monitoring and proactive early warning step for key quality attributes. This step aims to monitor key quality attributes in the production process in real time to proactively identify potential trends in quality fluctuations. This step specifically executes the following sub-steps:

[0132] First, sub-step S301 is executed to collect real-time data on key quality attributes such as filling volume and final sealing temperature. The key quality attribute sensing and acquisition unit in the process online monitoring module obtains the measured values ​​of key quality attributes for each production unit by communicating with sensors at specific locations on the production line. For the key quality attribute of filling volume, the unit acquires the weight data of each container from a high-speed dynamic weighing sensor installed on the finished product conveying path. For the key quality attribute of final sealing temperature, the unit acquires the instantaneous temperature data of the sealing area of ​​each container from a non-contact infrared thermometer installed at the sealing station exit.

[0133] The critical quality attribute sensing and acquisition unit operates at a frequency synchronized with the production line cycle, ensuring that data is collected for every product coming off the line. The raw sensor signals collected are converted into digital values ​​with specific units within this unit, such as grams (g) for weight and degrees Celsius (°C) for temperature.

[0134] This unit assigns a unique production serial number and a unified timestamp to each collected data point, forming a complete record of key quality attribute data. These records are transmitted in real time and continuously to the statistical process control trend analysis unit within the online process monitoring module, serving as the data input source for subsequent analysis.

[0135] Next, sub-step S302 is executed to construct and update statistical process control charts for key quality attribute data in real time. After receiving the data stream from the key quality attribute sensor acquisition unit, the statistical process control trend analysis unit in the online process monitoring module independently constructs and maintains statistical process control charts for the two key quality attributes: filling volume and final sealing temperature. These control charts include a mean control chart for monitoring the data mean and a range control chart for monitoring the data fluctuation range.

[0136] At the start of each production batch, the statistical process control trend analysis unit uses a set of sample data collected in the initial phase to calculate the center line, upper control limit, and lower control limit of each control chart. During subsequent continuous production, whenever new data points are received, the unit plots them on the corresponding control chart and dynamically updates the charts. Simultaneously, the unit periodically calculates the process capability index to quantitatively assess the current production process's ability to meet preset quality specifications. The process capability index is calculated using the following mathematical expression:

[0137] ;

[0138] in, It is a process capability index, used to assess the ability of a process to meet specifications. This represents the upper limit of the quality attribute specification. This represents the lower limit of the quality attribute specification. The overall mean of all sample data; This is an estimate of the total standard deviation of the process.

[0139] The calculated process capability index value is compared with a preset acceptance standard, for example, a process capability index value greater than 1.33 is required. This calculation result is then compared with statistical process control (SPC) data. Figure 1 The data is stored and displayed on the system's user interface, providing operators and quality management personnel with an intuitive view and quantifiable indicators of the process status.

[0140] Then, sub-step S303 is executed to apply early warning rules for trend analysis. The statistical process control trend analysis unit not only determines whether a single data point exceeds the control limits, but also has a built-in set of rules for identifying non-random fluctuations in data. These rules include, but are not limited to: multiple consecutive points falling on the same side of the center line, multiple consecutive points showing a continuous upward or downward trend, and data points alternating near the control limits. This unit performs real-time pattern matching on the data point sequence on the control chart to detect early signals indicating potential deviations in the process.

[0141] To further improve the detection sensitivity of small but persistent process deviations, the statistical process control trend analysis unit also applies an exponentially weighted moving average model to smooth the data and predict trends. This unit calculates the corresponding exponentially weighted moving average statistic for each newly acquired observation, following the mathematical model below:

[0142] ;

[0143] in, For the first EWMA statistics at each time point; It is a smoothing constant, with a value range greater than 0 and less than or equal to 1, which determines the weight of historical data; For the first Actual observations at each time point; This is the EWMA statistic for the previous time point.

[0144] The statistical process control trend analysis unit plots the calculated exponentially weighted moving average statistic series on a dedicated exponentially weighted moving average control chart. This control chart has narrower control limits than the standard mean control chart. By monitoring whether the exponentially weighted moving average statistic exceeds its control limits, the system can identify small changes in the process mean earlier than by relying solely on standard outlier rules.

[0145] Finally, sub-step S304 is executed to trigger a tiered early warning system and provide an early warning signal to the central coordination controller. When the statistical process control trend analysis unit detects any data point exceeding the control limit, violating any preset anomaly detection rules, or exceeding the limit of the exponentially weighted moving average statistic, the real-time early warning feedback unit in the process online monitoring module is activated. This unit performs tiered early warning operations based on the severity of the triggering conditions.

[0146] The first-level warning, defined as the alert level, is triggered when a relatively slight non-random trend is detected. The real-time warning feedback unit will display a prompt message on the system's human-machine interface and change the background color of the relevant control chart to attract the operator's attention, but will not interrupt production. The second-level warning, defined as the alarm level, is triggered when a data point is detected to have clearly exceeded control limits or other serious conditions. This unit will trigger an audible alarm and a more prominent visual alarm.

[0147] Upon triggering the second-level warning, the real-time warning feedback unit immediately generates a structured warning signal and sends it to the central coordinating controller in the collaborative optimization control module via the internal data bus. This warning signal contains specific information, indicating which critical quality attribute has deviated, the direction and magnitude of the deviation, and the specific rules for triggering the alarm. Upon receiving this signal, the advanced process control algorithm unit of the central coordinating controller immediately initiates the corresponding automatic adjustment program, making compensatory adjustments to the relevant process parameters, or, if necessary, safely suspending the production process and issuing instructions to the operator requiring manual intervention. This series of actions constitutes a closed-loop feedback from quality monitoring to process control, enabling proactive prevention and rapid response to quality problems.

[0148] See attached document Figure 6 As the entire production process continues, the production control method provided in this embodiment of the invention synchronously executes step S400, namely the step of full-process production data collection and electronic batch record generation. This step aims to ensure the comprehensive, complete, and traceable recording of all production-related data. This step specifically executes the following sub-steps:

[0149] First, sub-step S401 is executed to collect and synchronize process parameters, equipment status, and alarm information from all distributed control units. The production data acquisition and monitoring system data acquisition unit in the data integration management module establishes a continuous data communication connection with all control units under the process distributed execution module and the process online monitoring module via high-speed industrial Ethernet. This unit periodically obtains data such as temperature and screw speed in each section of the extruder from the feeding and plasticizing control unit, key process parameters such as blow molding pressure, filling time, and sealing temperature and pressure from the blow-fill-seal core control unit, leak detection results and weight values ​​for each product from the online detection and rejection unit, and environmental data such as air velocity, pressure difference, and particle count in the clean area from the aseptic environment control unit.

[0150] In addition, the production data acquisition and monitoring system's data acquisition unit also collects the real-time operating status of all control units, such as running, standby, and fault status, as well as any alarm events occurring during production and their confirmation and clearing records. All acquired data is synchronized using a unified high-precision clock to ensure consistency in the time series of data from different sources. The acquired raw data stream is sent to a dedicated real-time database for initial storage, providing source data for subsequent data processing and record generation.

[0151] This unit is also responsible for collecting key quality attribute data from the online process monitoring module, including raw measurements collected by the key quality attribute sensing and acquisition unit, as well as statistics and early warning signals generated by the statistical process control trend analysis unit. By aggregating process parameters, equipment status, environmental data, and key quality attribute data, this unit constructs a comprehensive dataset covering the entire production process, laying the foundation for achieving comprehensive data traceability.

[0152] Next, sub-step S402 is executed to attach metadata to each record to ensure that the data meets the principles of attribution, clarity, synchronization, originality, and accuracy. The electronic batch record automatic generation unit in the data integration management module extracts data records from the real-time database and processes each record, attaching a set of structured metadata to it. This metadata includes the operator's identity identifier who performed the relevant operation, the unique identifier of the device or sensor that collected the data, the precise timestamp of data generation, and the record type identifier.

[0153] To ensure the long-term integrity and immutability of data, the electronic batch record automatic generation unit applies a secure hash function to generate a unique digital fingerprint for each record after appending metadata. This process involves concatenating the core data of the current record, the timestamp, the user's identity identifier, and the hash value of the previous related record, and then calculating its hash value. This method establishes a data record chain, where any modification to any record will cause a change in the hash values ​​of all subsequent records, making it easily detectable. The calculation process follows the mathematical expression below:

[0154] ;

[0155] in, A unique hash value generated for the current record; It is a secure hash algorithm function; The core data to be recorded; This is the timestamp of the record; The user ID that performed the operation; The hash value of the previous record forms a data chain; This is a data concatenation operator.

[0156] By attaching complete metadata to each record and protecting it with hash chain technology, this step ensures that the source of all production data is clearly traceable, the content is authentic and original, and it has not been tampered with during storage and transmission, fully complying with the stringent requirements of Good Manufacturing Practice (GMP) for data integrity.

[0157] Then, sub-step S403 is executed to automatically integrate the structured data by batch and generate electronic batch records. Upon receiving a batch production completion instruction from the Manufacturing Execution System (MES), or when the system detects that a preset production quantity has been completed, the electronic batch record automatic generation unit initiates the electronic batch record generation process. This unit retrieves all data records related to the current production batch that have undergone integrity verification from the database, based on the current production batch number, covering the entire process from production preparation, online cleaning and sterilization, formal production to completion.

[0158] The electronic batch record automatic generation unit formats and organizes this data according to a predefined, regulatory-compliant template, generating a well-structured and comprehensive electronic batch record document. This document includes batch information, material information, curves or lists showing the changes of all key process parameters over time, statistical process control charts for key quality attributes, equipment operation and shutdown records, a detailed list of all alarm events, and all operator operation logs. Simultaneously, based on the batch's operational data, the unit automatically calculates the overall equipment efficiency and writes it into the electronic batch record as part of the batch summary. The overall equipment efficiency is calculated using the following formula:

[0159] ;

[0160] in, To improve the overall efficiency of the equipment; The time utilization rate is the ratio of actual operating time to planned operating time. Performance utilization rate, which is the ratio of actual output to theoretical output; The pass rate is the ratio of the number of qualified products to the total output.

[0161] The final generated electronic batch record is stored in a secure, standardized electronic document format, such as encrypted PDF or XML. Once generated, the record is locked in a read-only state and accompanied by an electronic signature to confirm its finality and authority, thereby replacing traditional paper batch records and achieving paperless management of production records.

[0162] Finally, sub-step S404 is executed, uploading the generated electronic batch record and production report to the Manufacturing Execution System (MES) via a standard interface. The MES bidirectional interface unit in the Data Integration Management module is responsible for data exchange between this system and the upper-level MES. Once an electronic batch record is generated and finalized, this unit packages it with a summary report containing key production indicators, ready for uploading.

[0163] This unit organizes this information into a standardized data packet for interface transmission. Structurally, this data packet contains several well-defined data fields. For example, it includes a batch information field containing a unique batch number, product code, production start time, and end time; an electronic batch record document field specifying the document format, storage location, or reference identifier, as well as a checksum for data verification; and a production summary field containing key performance indicators calculated for this batch, such as overall equipment efficiency, total output, number of qualified products, number of unqualified products, and pass / fail rate.

[0164] This unit sends the data packet to the application programming interface (API) of the Manufacturing Execution System (MES) via a pre-configured secure network connection. After receiving and verifying the data, the MES returns an acknowledgment message. Upon receiving the acknowledgment, the MES bidirectional interface unit updates the local electronic batch record status to "uploaded." This step completes the automatic data flow from the production execution layer to the enterprise management layer, achieving transparent management of the production process and enterprise-level data integration.

[0165] Specific application examples:

[0166] The following is a detailed explanation of a blow-fill-seal production task for a 10mL sodium chloride injection with production batch number NS10V3-B001.

[0167] First, during the system initialization and production instruction parsing steps, the Manufacturing Execution System (MES) sends a production instruction to this control system containing the following information: production batch number NS10V3-B001, product code NaCl-10mL-V3, planned output of 100,000 units, and formula version number 3.2. After successfully parsing this instruction, the production instruction parsing unit retrieves the NaCl-10mL-V3-3.2 process formula from the formula database. This formula specifies the target temperatures for zones one through five of the extruder as 185℃, 195℃, 205℃, 205℃, and 200℃, respectively; the target filling weight as 10.20g, with a acceptable range of 10.15g to 10.25g; and a sealing temperature of 135℃, among other key parameters. Subsequently, the system automatically completes equipment self-inspection, online pipeline cleaning, and online sterilization procedures, and confirms that the suspended particle concentration and pressure differential in the core Class A zone meet ISO 5 standards. The system then enters standby mode.

[0168] Production officially commences during the multivariate collaborative optimization and closed-loop control steps of the blow-fill-seal process. In the feeding and plasticizing stages, the system sets the target melt flow index. The concentration is 2.5g / 10min. The advanced process control algorithm unit is based on the correlation model. To achieve stable control. Assume that at a certain moment, the system monitors the actual melting temperature. The screw speed is 478K (205℃). The rpm is 50 rpm, and the model parameters are known: material constants. =2.1x10 6 Power Law Index =0.85, activation energy of flow =65000 J / mol, ideal gas constant =8.314 J / (mol·K). The calculated prediction is... The value was 2.41 g / 10 min, lower than the target value. The algorithm unit then calculated the required melting temperature. The temperature was increased to 480K (207℃) to correct the deviation, and this new setting was issued, thus achieving closed-loop control of the molten flowability of the billet.

[0169] During the preform blow molding stage, the blow-fill-seal core control unit employs a segmented proportional-integral-derivative control algorithm to accurately reproduce the blow molding pressure curve. At a certain moment during the main blow stage... The target pressure setting was 0.350 MPa, while the actual value fed back by the pressure sensor was 0.345 MPa, resulting in a pressure error of 0.005 MPa. The control unit is based on the formula. Calculate the control output. If the gain coefficient is set to: proportional gain coefficient... =1.2, integral gain coefficient =0.5, differential gain coefficient If the error is 0.1, the system will combine the current error, the historical error accumulation, and the error change rate to calculate a precise adjustment signal P(t) for the proportional control valve, which will quickly eliminate the pressure deviation.

[0170] During the liquid filling stage, the system executes a dynamic compensation algorithm based on weighing feedback. Let the target filling weight be... It is 10.20g, and the compensation coefficient is... It is 0.15. When the serial number is =500 Product online weighing results When the concentration is 10.18g, the filling deviation is +0.02g. Assume the control compensation amount used for the 500th filling. If the value is -0.012s (indicating a reduction in the filling valve opening time), then the system calculates the compensation amount for the next cycle as follows: =−0.012+0.15×(10.20−10.18)=−0.009s. This new compensation amount was used to adjust the filling time of the 501st product, achieving successive precise correction of the filling volume.

[0171] During the molten sealing stage, the system employs a multivariate collaborative control strategy. The sealing strength prediction model is as follows: The regression coefficients of the model were experimentally calibrated as follows: =-25.8, =0.2, =15.5, =2.1. At a certain moment, the system monitored the sealing temperature. For some reason, the temperature dropped slightly from 135°C to 134°C, while the sealing pressure... The pressure is 0.40 MPa, and the sealing time is... The time is 1.5 seconds. Substituting this into the model, the predicted sealing strength will be lower than the target value. The advanced process control algorithm unit immediately calculates that, to maintain the original sealing strength target, the sealing pressure needs to be adjusted. The pressure was increased to 0.413 MPa, and this command was issued and executed, thereby actively compensating for the impact of temperature fluctuations on the final sealing quality.

[0172] Throughout the entire production process, the data collection and electronic batch record generation steps are synchronized. For example, for a serial number of... For a product with a fill weight of 10.18g, the electronic batch record generation unit will attach metadata to it and apply a hash function. A digital fingerprint is generated. The specific concatenation is: "10.18g||Time||Filling Unit||a1b2c3d4..." (the hash value of the previous record), ensuring the integrity and immutability of the data chain. When batch production ends, the system automatically integrates all data to generate an electronic batch record and calculates the overall equipment efficiency. If the planned operation time is 8 hours, the actual operation time is 7.5 hours (time utilization rate). =93.75%); theoretical output 120,000 bottles, actual output 105,000 bottles (performance utilization rate) =87.5%); 104,500 bottles of qualified products (qualified product rate) =99.52%). Therefore, the final calculated percentage for this batch is... for: =93.75% × 87.5% × 99.52% = 81.64%. This electronic batch record, which includes complete process data, trend charts, alarm records, and production summaries, is ultimately uploaded to the Manufacturing Execution System via a standard interface.

[0173] Reference Figure 7This figure illustrates a simulation comparison between the Advanced Process Control (APC) method of this invention and the traditional fixed-parameter control method for the key quality attribute of liquid filling, specifically the filling weight. The horizontal axis represents the production serial number, and the vertical axis represents the filling weight (g). The two dark gray horizontal dashed lines represent the upper limit (10.25g) and lower limit (10.15g) of the specification, respectively. When the simulation reached serial number 1500, a disturbance caused by a small change in systemic filling pressure due to external factors (such as a drop in the liquid level in the storage tank) was simulated. The gray curve represents the traditional control method; after the disturbance, the moving average of the filling weight showed a significant and prolonged upward shift, with a slow recovery. The black curve represents the Advanced Process Control method of this invention. Thanks to the successive dynamic compensation algorithm based on weighing feedback, although fluctuations occurred at the moment of the disturbance, the system quickly pulled the average back to the target value of 10.20g within a short time after detecting the deviation, and consistently maintained a smaller fluctuation range, demonstrating the significant advantages of this invention in suppressing process disturbances and improving product quality consistency.

[0174] Experimental verification and effect comparison:

[0175] To further verify the effectiveness of the online monitoring and proactive early warning steps for key quality attributes in this invention, a comparative experiment was conducted. The experimental subject was the final sealing temperature, and a slow, linear aging effect of the heating element was artificially introduced to simulate a subtle process drift.

[0176] Reference Figure 8 This figure illustrates the results of statistical process control (SPC) and exponentially weighted moving average (EWMA) analysis of the monitoring data for the final sealing temperature during the experiment. The horizontal axis represents the production serial number, and the vertical axis represents the sealing zone temperature (°C). The light gray data points in the figure represent the original measured sealing temperature for each product. The figure shows that from serial number 1000 onwards, the data exhibits a slow upward trend. The two horizontal dark gray dashed lines represent the upper and lower control limits (UCL / LCL) calculated according to standard SPC theory. The black solid line is the EWMA trend line calculated using the exponentially weighted moving average model. A smoothing constant is set when calculating the EWMA statistic. =0.2. For example, when the first... Actual observed values ​​at each point =135.3℃, while the EWMA statistic at the previous time point When the temperature is 135.2℃, the current EWMA statistic is: =0.2×135.3+(1−0.2)×135.2=135.22℃.

[0177] from Figure 8 As can be seen, although the original data points (light gray dots) only first touched the upper control limit (UCL) when the sequence number approached 2800, potentially triggering a traditional SPC alarm, the trend line (solid black line) smoothed by the EWMA model is more sensitive to the cumulative effect of small offsets, making the process drift trend more obvious. Based on this trend line, the system can determine that a significant process deviation has occurred when the sequence number is around 1800, triggering a proactive warning. This indicates that the trend analysis method combined with EWMA used in this invention can identify potential process drift approximately 1000 production cycles earlier than the traditional SPC method. In this example, the process capability index was calculated for the stable process before the experiment began. Set upper limit for specifications =138℃, lower limit =132℃, the average value of the process was measured. =135.0℃, estimated process standard deviation =0.3℃, calculated as follows: =3.33, indicating a very high process capability. Early warning and intervention can ensure quality throughout the entire production process. The value was consistently maintained at a level far above 1.33, effectively preventing the generation of defective products due to process drift, thus verifying the superior performance of this invention in proactive quality prevention.

Claims

1. A production control system for aseptic blow-fill-seal integrated pharmaceutical plastic bottles, characterized in that: include: The data integration and management module is used to receive production instruction data packets issued by the manufacturing execution system and collect raw data records from the process distribution execution module and the process online monitoring module. It is used to generate electronic batch records and upload them to the manufacturing execution system to complete the full-process data traceability and closed-loop management of production batches. The collaborative optimization control module is used to parse the production instruction data packet to obtain the formula parameters, calculate the optimal control setpoint and generate the work cycle timing for distribution, and receive the early warning signal to update the optimal control setpoint and re-distribute it. The process distribution execution module is used to receive the optimal control setpoint and the operation cycle sequence, execute the closed-loop control of the blow-fill-seal process of the liquid medicine plastic bottle in an aseptic environment, and feed back the actual measured values ​​of the process parameters to the collaborative optimization control module. The process online monitoring module is used to collect raw data of key quality attributes in real time, generate the early warning signal and feed it back to the collaborative optimization control module to trigger the dynamic update of the optimal control setpoint, thereby dynamically adjusting the closed-loop control of the blow-fill-seal process.

2. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The collaborative optimization control module parses the production instruction data packet to obtain formula parameters, specifically including: The production instruction parsing unit in the collaborative optimization control module starts a parsing program on the received production instruction data packet, extracts the production batch number, product code and formula version number contained in the production instruction data packet and performs the parsing action; The production instruction parsing unit verifies the format and data type of the extracted fields, generates a confirmation receipt, and sends it back to the manufacturing execution system. The production instruction parsing unit uses the parsed product code and formula version number as composite query keywords to retrieve and read matching formula records from the formula database and load them into the system's dynamic memory to obtain the formula parameters that cover detailed process parameters.

3. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The process distribution execution module receives the optimal control setpoint and executes closed-loop control, specifically feeding back the actual measured values ​​of process parameters, including: The feeding and plasticizing control unit in the process distribution execution module collects the actual measured values ​​of the melt temperature of the extruder barrel and the actual measured values ​​of the screw speed in real time, and feeds them back to the collaborative optimization control module as the actual measured values ​​of the process parameters. The feeding and plasticizing control unit receives the adjustment amount of the temperature setpoint or screw speed setpoint calculated by the collaborative optimization control module based on the deviation between the actual measured value and the target setpoint of the process parameters; The feeding and plasticizing control unit uses the received adjustment amount as the updated optimal control setting value to control the heating zone of the extruder barrel and the screw drive motor to perform the closed-loop control to maintain the constant quality of the tubular preform.

4. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The process distribution execution module receiving the job cycle timing execution closed-loop control specifically includes: The blow-fill-seal core control unit in the process distribution execution module measures the current actual blow molding pressure and compares the actual blow molding pressure with the target pressure set value to calculate the pressure error; The blow-fill-seal core control unit receives a series of millisecond-level operation cycle timing signals from the global timing arbitration unit within the collaborative optimization control module; The blow-fill-seal core control unit automatically switches to the parameter group of the corresponding stage based on the received operation cycle timing signal, calculates the adjustment amount of the proportional control valve opening based on the pressure error, and executes the closed-loop control in the manner of tracking the pressure curve.

5. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The process distribution execution module receives the optimal control setpoint and executes closed-loop control, specifically including: The blow-fill-seal core control unit in the process distribution execution module executes the liquid filling program and controls the liquid to be injected into the formed container; The blow-fill-seal core control unit receives the control compensation amount calculated by the collaborative optimization control module based on the deviation between the actual measured weight data and the target filling weight; The blow-fill-seal core control unit uses the received control compensation amount as the updated optimal control setting value, and automatically adjusts the execution parameters of the next filling cycle to perform the closed-loop control.

6. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The collaborative optimization control module receives the actual measured values ​​of the process parameters to update the optimal control setpoint, specifically including: The advanced process control algorithm unit in the collaborative optimization control module receives the actual measured values ​​of the process parameters, such as sealing temperature and sealing pressure, which are monitored and uploaded in real time by the process distribution execution module. The advanced process control algorithm unit uses a pre-established multiple linear regression prediction model for sealing strength, substitutes the actual measured values ​​of the process parameters, and calculates the predicted value of the sealing strength in real time. If the predicted value shows a trend of deviating from the target intensity, the advanced process control algorithm unit calculates the parameter instructions for multi-variable coordinated adjustment based on the model, and sends out the adjusted parameter instructions as the updated optimal control setpoint.

7. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The online process monitoring module collects raw data of key quality attributes in real time, specifically including: The key quality attribute sensing and acquisition unit in the process online monitoring module acquires weight data from a high-speed dynamic weighing sensor and temperature data from a non-contact infrared thermometer at a frequency synchronized with the production line cycle as the real-time acquisition action. The key quality attribute sensing and acquisition unit internally converts the acquired data into digital values ​​with specific units; The key quality attribute sensing and acquisition unit adds a unique production serial number and a unified timestamp to the digitized value to generate complete raw data of the key quality attribute.

8. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The process online monitoring module generates an early warning signal and feeds it back to the collaborative optimization control module, specifically including: The statistical process control trend analysis unit in the online process monitoring module receives continuous data streams to construct statistical process control charts and applies preset anomaly detection rules to perform real-time pattern matching on the data point sequence to identify non-random fluctuations. When a data point is detected to exceed the preset control limit or violate the preset discrimination rule, the real-time early warning feedback unit in the online process monitoring module is activated to generate the structured early warning signal. The real-time early warning feedback unit actively feeds back the early warning signal containing specific deviation information to the central coordinating controller in the collaborative optimization control module, providing a basis for the adaptive adjustment of the process parameters of the collaborative optimization control module.

9. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The data integration and management module collects the raw data records from the process distribution execution module and the online process monitoring module, specifically including: The production data acquisition and monitoring system data acquisition unit in the data integration management module continuously communicates data via the industrial bus, periodically acquiring the process parameters and equipment status of the process distribution execution module to perform the acquisition action; The production data acquisition and monitoring system's data acquisition unit synchronously acquires key quality attribute data from the process online monitoring module and performs the acquisition action. The production data acquisition and monitoring system's data acquisition unit adds a unified clock to all acquired data for time synchronization and stores it in a real-time database to form a complete record of the original data.

10. The integrated blow-fill-seal aseptic pharmaceutical bottle production control system according to claim 1, characterized in that, The data integration and management module generates electronic batch records and uploads them to the manufacturing execution system, specifically including: The electronic batch record automatic generation unit in the data integration management module retrieves the original data records related to the batch from the real-time database and integrates them according to a predefined structured template to generate the electronic batch record. The manufacturing execution system bidirectional interface unit in the data integration management module packages the generated electronic batch records and summary reports containing key production indicators into a standardized data package; The bidirectional interface unit of the manufacturing execution system sends the data packet to the application programming interface of the manufacturing execution system, and completes the upload to the manufacturing execution system after receiving the confirmation message.

Images