Extrusion granulation production line whole-process material tracing method and system

Abstract

The application relates to a material tracing method and system for an extrusion granulation production line full-process, which comprises constructing a high-fidelity digital twin model synchronized with a physical production line, and establishing a real-time data communication link to realize virtual-real synchronization. In the virtual environment, the raw material batch switching event is converted into a unique material digital identifier and dynamically injected into the virtual material microelement at the corresponding time point; by constructing a material identifier propagation dynamic model, the time and space motion trajectory of each identifier in the virtual device is calculated in real time based on production data. Finally, the mixing state of the material flow is analyzed in real time at the virtual finished product station, the dynamic finished product batch is divided according to the uniformity threshold, and the material pedigree tracing map containing the complete raw material source and process path is automatically generated. The application solves the problems of material flow ambiguity and tracing difficulty in continuous manufacturing, realizes the upgrade from static recording to dynamic and accurate tracing, and provides digital guarantee for quality control and process optimization.

Description

Technical Field

[0001] This application relates to the field of material traceability technology, and in particular to a method and system for tracing materials throughout the entire process of an extrusion granulation production line. Background Technology

[0002] With the deepening of intelligent manufacturing and Industry 4.0, continuous and process-oriented production methods, represented by extrusion granulation, are increasingly becoming core processes in the chemical, pharmaceutical, and new materials fields. Compared to intermittent production, continuous manufacturing has significant advantages in improving efficiency, stabilizing quality, and reducing energy consumption. However, the continuous flow and mixing of materials during its production process makes it extremely difficult to accurately trace back from the final product to the original raw material batch. Achieving end-to-end material traceability is a key technological foundation for enterprises to conduct in-depth process optimization, quality control, and achieve refined cost management.

[0003] Currently, common existing technologies for material traceability in production processes mainly rely on two categories of methods. The first is "static batch tracking" based on timestamps and logistics documents, which records batch numbers and times at raw material entry points, feeding points, and finished product output points, achieving coarse-grained traceability through logical association. The second is "serial number" or "RFID tag tracking," widely used in discrete manufacturing, which assigns a unique identifier to each independent product unit and records it through scanning at each process stage. In continuous manufacturing scenarios, some research has also explored deploying online detection instruments (such as near-infrared spectroscopy) at the inlets and outlets of key equipment, combined with simple residence time distribution models, to estimate and match the overall material flow trend.

[0004] Regarding the aforementioned technologies, static batch tracking cannot depict the dynamic behaviors of materials such as intense mixing, diffusion, and back-mixing that occur inside equipment such as screws and dies, resulting in a serious disconnect between the traceability results and the actual material flow direction, leading to insufficient accuracy. Discrete tagging and tracking technologies cannot be applied to continuous and uninterrupted fluid or particle flows, and the tags themselves may interfere with the process or be unable to withstand high-temperature and high-shear environments.

[0005] Based on this, this application provides a method and system for tracing materials throughout the entire extrusion granulation production line. Summary of the Invention

[0006] To address the shortcomings of static batch tracking, which fails to capture the dynamic behaviors of materials such as intense mixing, diffusion, and backmixing within equipment like screws and dies, resulting in a significant disconnect between traceability results and actual material flow, and insufficient accuracy; and the limitations of discrete tagging and tracking technologies in continuous fluid or particle flow, where tags may interfere with the process or be unable to withstand high-temperature and high-shear environments, this application provides a method and system for tracing materials throughout the entire extrusion granulation production line.

[0007] Firstly, this application provides a method for tracing materials throughout the entire process of an extrusion granulation production line, employing the following technical solution: including: A digital twin model is constructed based on the extrusion granulation production line. The digital twin model includes the geometric, physical, and control logic definitions corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line. A real-time data communication link is established between the digital twin model and the physical production line control system. In the digital twin model, virtual material digital identifiers are defined for raw materials entering the production line; raw material batch switching events and corresponding batch information in the physical production line are used as trigger signals and content, and the associated material digital identifiers are dynamically injected into the material micro-elements at the corresponding time points in the virtual material flow; Based on the equipment structure and motion relationship described by the digital twin model, a dynamic model for material identification propagation is constructed. The dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data, as continuous inputs to calculate the dynamic trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment. Based on the dynamic change trajectory of the material digital identifier, the complete process of a specific virtual material micro-element from raw material to finished product is correlated and calculated in real time. When the virtual material flow arrives at the preset finished product station, the finished product batch is dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material lineage traceability map corresponding to the dynamic batch is generated. The material lineage traceability map includes the raw material batch information, process path and key process parameter history that lead to the composition of the corresponding batch of finished product.

[0008] Preferably, the digital twin model constructed based on the extrusion granulation production line includes geometric, physical, and control logic definitions corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line. A real-time data communication link is established between the digital twin model and the physical production line control system, including: Construct a three-dimensional geometric model corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line, and perform virtual assembly to form the overall three-dimensional layout of the production line; Define physical properties for key components in the three-dimensional geometric model. These physical properties include mass, inertia, collision body properties, and material friction coefficient. Based on the actual motion mechanism of each unit, kinematic pairs and constraints are defined in the three-dimensional geometric model. Rotary pairs are configured for the screw and cutter, and sliding pairs are configured for the pusher cylinder and valve. Virtual sensors and virtual actuators are configured at key process nodes in the three-dimensional geometric model. The virtual sensors are used to detect material positions and equipment status in the simulation, and the virtual actuators are used to receive control signals and drive the movement of virtual components. Based on the aforementioned three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, a digital twin model is integrated to form a digital twin model that can respond to control commands and simulate the continuous operation of the production line. The digital twin model is connected to the control system of the physical production line through a real-time data communication link. The physical control system sends real-time production data as input signals to the digital twin model to drive the virtual actuators. The digital twin model sends the status feedback of the virtual sensors as output signals back to the physical control system, thus achieving virtual-real synchronization.

[0009] Preferably, the integration of the three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators to form a digital twin model, which can respond to control commands and simulate the continuous operation of the production line, includes: Configure corresponding electromechanical signals for virtual sensors and virtual actuators. The electromechanical signals include input signals and output signals. The input signals are used to receive external control commands, and the output signals are used to provide feedback on the status of the virtual sensors. Based on the continuous process flow of the extrusion granulation production line, a simulation sequence is established in the digital twin model. The simulation sequence defines the flow logic of virtual materials between the feeding, mixing, extrusion and granulation units, the action sequence and coordination rules of each virtual actuator, and the triggering conditions of each virtual sensor. The three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, electromechanical signal mapping relationships and simulation sequence logic are systematically integrated and encapsulated to form an independently operable digital twin model. The digital twin model interacts with external systems through electromechanical signal mapping relationships. Input simulated or real-time data communication link control commands into the digital twin model to drive the model to run, verify whether the digital twin model can accurately simulate the complete process of material feeding to granulation according to the commands and simulation sequence logic, and provide feedback on the corresponding operating status through output signals.

[0010] Preferably, in the digital twin model, a virtual material digital identifier is defined for the raw materials entering the production line; the raw material batch switching events and corresponding batch information in the physical production line are used as trigger signals and content, and the associated material digital identifier is dynamically injected into the material micro-element at the corresponding time point in the virtual material flow, including: In the digital twin model, a data structure for the digital identification of materials is defined, and the data structure includes at least the raw material batch number, the proportion of ingredients in the formula, and physical property parameters; In the digital twin model, an identification injection logic unit is configured corresponding to the raw material inlet position of the physical production line. The identification injection logic unit listens for and receives raw material batch switching event signals and associated batch information from the physical control system through a real-time data communication link. Establish a virtual time synchronization mapping to keep the simulation time in the digital twin model synchronized with the real-time clock of the physical production line, ensuring that the position of the material micro-element on the virtual time axis corresponds to the physical production time; When the identifier injection logic unit receives a raw material batch switching event signal, it creates a new material digital identifier based on the received batch information and binds it to the virtual timestamp of the event occurrence time. Based on the virtual time synchronization mapping, the virtual material micro-element corresponding to the virtual timestamp is located in the virtual material flow of the digital twin model, and the bound material digital identifier is dynamically injected into the corresponding material micro-element as a traceable virtual identity.

[0011] Preferably, a dynamic model for material identification propagation is constructed based on the equipment structure and motion relationship described by the digital twin model. This dynamic model uses the injected material digital identifiers and their injection time, as well as real-time production data, as continuous inputs to calculate the dynamic trajectory of the spatial position and temporal attributes of the material digital identifiers within the virtual production line equipment, including: Based on the geometric connection relationships and movement directions of the feeding, mixing, extrusion and granulation units described by the digital twin model, the directed network topology of the material flow of the virtual production line is abstracted and established, where network nodes correspond to equipment functional units and directed edges correspond to material transmission paths. An input interface and an output interface are defined for the dynamic model of material identification propagation. The input interface is used to continuously receive material digital identification, injection time and real-time production data, and the output interface is used to output the dynamic change trajectory. Configure material identifier propagation calculation rules for various nodes and directed edges in the directed network topology of material flow. Based on equipment mechanism or data-driven, calculate the spatial displacement, time delay and mixing state with other identifiers of material digital identifiers when they flow through corresponding nodes or edges according to real-time production data. During the operation of the material identification propagation dynamic model, the propagation calculation rule library is called according to the directed network topology of the material flow and in virtual time sequence. For each injected material digital identifier, its spatial position and residence time in the virtual production line equipment are calculated step by step based on its real-time production data, and integrated to form the dynamic change trajectory and data of the corresponding identifier, and output through the output interface.

[0012] Preferably, the dynamic change trajectory based on the material digital identifier is used to correlate and calculate the complete process of a specific virtual material micro-element from raw material to finished product in real time; when the virtual material flow arrives at the preset finished product station, the finished product batch is dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material genealogy traceability map corresponding to the dynamic batch is generated, including: The dynamic change trajectory output by the material identification propagation dynamic model is obtained, and the virtual material micro-elements carrying the same material digital identifier are spatiotemporally correlated. The complete virtual process of the raw material carrying the corresponding material digital identifier from the injection point through each virtual unit of feeding, mixing, extrusion, and granulation to the finished product station is reconstructed in reverse. In the digital twin model, the status of the virtual material flow when it arrives at the preset finished product station is continuously monitored, the types, proportions and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status are determined, and the corresponding mixed state quantitative indicators are calculated. The mixed state quantification index is compared with the uniformity threshold in real time. When the index fluctuation exceeds the threshold, it is determined as the batch boundary and used as a dynamic cutting point to divide the previously continuously produced virtual material flow into an independent dynamic finished product batch with internal uniformity. For each segmented independent dynamic finished product batch, the system automatically associates and aggregates all raw materials that constitute the complete virtual process of the corresponding batch; it then integrates the raw material batch information, process path, and key process parameter history contained in the complete virtual process to generate a structured material family traceability map.

[0013] Preferably, in the digital twin model, the continuous monitoring of the virtual material flow's status when it arrives at a preset finished product station, determining the types, proportions, and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status, and calculating the corresponding mixed-state quantification index, includes: In the digital twin model, a virtual material analysis window is set up corresponding to the finished product station. The material analysis window simulates the behavior of instantaneous sampling of continuous output on the physical production line, and is used to capture and analyze a virtual material flow segment passing through the corresponding station at a certain moment. The virtual material flow segment captured in the material analysis window is parsed to identify and extract all the different material digital identifiers contained therein; the number of virtual material micro-elements or the virtual material quality represented by each different material digital identifier is counted. Based on the statistical results, the mass or quantity proportion of each material digital identifier in the virtual material flow segment is calculated to obtain the proportion data; at the same time, the spatial or temporal distribution dispersion of the virtual material micro-elements associated with each material digital identifier in the virtual material flow segment is analyzed to obtain the distribution characteristics. Based on the calculated proportion data and the distribution characteristics, the mixed state quantification index is generated. The mixed state quantification index includes at least the concentration ratio of the main identifier, the number of identifier types, and the uniformity index of the mixing between different identifiers.

[0014] Secondly, this application discloses a material traceability device for the entire process of an extrusion granulation production line, which adopts the following technical solution, including: The data mapping module is used to build a digital twin model based on the extrusion granulation production line. The digital twin model contains the geometric, physical and control logic definitions corresponding to the feeding, mixing, extrusion and granulation units in the physical production line. A real-time data communication link is built between the digital twin model and the physical production line control system. The digital twin module is used to define virtual material digital identifiers for raw materials entering the production line in the digital twin model; it uses raw material batch switching events and corresponding batch information in the physical production line as trigger signals and content, and dynamically injects the associated material digital identifiers into the material micro-elements at the corresponding time points in the virtual material flow; The dynamic change module is used to construct a dynamic model for material identification propagation based on the equipment structure and motion relationship described by the digital twin model. The dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data as continuous inputs, to calculate the dynamic change trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment. The traceability graph module is used to associate and calculate the complete process of a specific virtual material micro-element from raw material to finished product in real time based on the dynamic change trajectory of the material digital identifier. When the virtual material flow arrives at the preset finished product station, the module dynamically divides the finished product batches according to the mixing state of the material digital identifier and the preset uniformity threshold, and generates a material lineage traceability graph corresponding to the dynamic batch. The material lineage traceability graph includes the raw material batch information, process path and key process parameter history that lead to the composition of the corresponding batch of finished product.

[0015] Thirdly, this application also provides a control device, the device comprising: It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described above for the material traceability method for the entire extrusion granulation production line.

[0016] Fourthly, this application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above in the material traceability method for the entire extrusion granulation production line.

[0017] In summary, this application first constructs a high-fidelity digital twin model fully synchronized with the physical production line, covering the entire unit geometry, physical, and control logic from feeding to granulation, and achieving virtual-physical mapping through a real-time data link. In the virtual environment, the system converts each raw material batch switching event into an identifiable material digital identifier and dynamically injects it into the corresponding virtual material micro-element. Subsequently, based on the equipment structure and motion relationships defined by the twin model, a dynamic model for material identifier propagation is constructed. This model, driven by real-time process data, continuously calculates the dynamic trajectory of the spatial position and temporal attributes of each identifier within the virtual equipment. Finally, at the virtual finished product station, the system monitors the mixing state of different identifiers in the material flow, dynamically classifies internally uniform finished product batches through intelligent comparison with a preset uniformity threshold, and automatically generates a detailed material genealogy traceability map displaying the batch. This upgrades traditional static, discrete batch records to dynamic, computable, end-to-end digital traceability that evolves synchronously with the production process, thus providing digital assurance for precise quality control and process root cause analysis. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for tracing materials throughout the entire extrusion granulation production line.

[0019] Figure 2 This is a structural block diagram of a material traceability device for the entire process of an extrusion granulation production line. Detailed Implementation

[0020] The following combination Figures 1-2 This application will be described in further detail.

[0021] The core implementation path of the full-process material traceability method for extrusion granulation production lines proposed in this application lies in constructing a digital twin environment that deeply integrates the process mechanism, and in this environment, completing the definition, injection, propagation, and aggregation analysis of the material's digital genes. The implementation process follows a continuous logical chain from physical perception to virtual modeling, from event-driven to identifier propagation, and from dynamic calculation to decision output. First, based on the digital definition of the production line's geometry, physical, and control logic, a high-fidelity digital twin base is constructed and a real-time data channel is established; then, a material identification system and dynamic propagation model are established in this virtual model to achieve accurate identification and tracking of raw material batch information in the continuous process; finally, through monitoring and analysis of the virtual finished product status, batches are dynamically divided and a full-process material genealogy map is generated.

[0022] Reference Figure 1 The embodiments of this application include at least steps S10 to S40.

[0023] S10, a digital twin model is constructed based on the extrusion granulation production line. The digital twin model includes the geometric, physical and control logic definitions corresponding to the feeding, mixing, extrusion and granulation units in the physical production line. A real-time data communication link is established between the digital twin model and the physical production line control system.

[0024] S20, in the digital twin model, defines virtual material digital identifiers for raw materials entering the production line; uses raw material batch switching events and corresponding batch information in the physical production line as trigger signals and content, and dynamically injects the associated material digital identifiers into the material micro-elements at the corresponding time points in the virtual material flow.

[0025] S30, based on the equipment structure and motion relationship described by the digital twin model, constructs a dynamic model for material identification propagation; the dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data as continuous inputs, to calculate the dynamic trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment.

[0026] S40, based on the dynamic change trajectory of material digital identifiers, real-time association and calculation of the complete process of a specific virtual material micro-element from raw material to finished product; when the virtual material flow arrives at the preset finished product station, according to the mixing state of the material digital identifiers and the preset uniformity threshold, the finished product batch is dynamically divided, and a material lineage traceability map corresponding to the dynamic batch is generated. The material lineage traceability map includes the raw material batch information, process path and key process parameter history that lead to the composition of the corresponding batch of finished product.

[0027] Specifically, the system establishes a high-fidelity virtual mirror of the physical extrusion granulation production line and maintains synchronization through a real-time data link. In this virtual environment, the system transforms each raw material switching event into a unique material digital identifier and precisely injects it into the virtual material flow at the corresponding moment. By constructing a "dynamic model of material identifier propagation" that integrates mechanism and data, the system can calculate the movement and mixing trajectory of these identifiers within complex process equipment in real time. Finally, at the virtual finished product end, the system intelligently analyzes the mixing state of the identifiers, dynamically divides homogeneous product batches, and automatically generates a detailed phylogenetic map, fully revealing the raw material source, process history, and key parameters of that batch of products. This transforms the ambiguous material flow in continuous production into a clear, calculable, and visualized digital network, fundamentally solving the challenges of accurate traceability, root cause analysis of quality, and flexible batch management.

[0028] In some embodiments, step S10 specifically includes the following steps: constructing a three-dimensional geometric model corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line, and performing virtual assembly to form the overall three-dimensional layout of the production line; defining physical properties for key components in the three-dimensional geometric model, including mass, inertia, collision body properties, and material friction coefficient; defining kinematic pairs and constraints in the three-dimensional geometric model according to the actual motion mechanism of each unit, configuring rotary pairs for screws and cutters, and sliding pairs for pusher cylinders and valves; configuring virtual sensors and virtual actuators at key process nodes in the three-dimensional geometric model, whereby the virtual sensors are used in simulation. The system detects material position and equipment status. Virtual actuators receive control signals and drive the movement of virtual components. Based on a 3D geometric model, physical properties, kinematic pairs and constraints, virtual sensors, and virtual actuators, a digital twin model is integrated. The digital twin model can respond to control commands and simulate the continuous operation of the production line. The digital twin model is connected to the control system of the physical production line through a real-time data communication link. The physical control system sends real-time production data as input signals to the digital twin model to drive the virtual actuators. The digital twin model sends the status feedback from the virtual sensors as output signals back to the physical control system, achieving synchronization between the virtual and real systems.

[0029] Specifically, starting with establishing a three-dimensional geometric model and layout corresponding to the physical units, physical properties such as mass and friction are gradually assigned to them, and constraints such as revolute joints and sliding joints are configured based on real motion mechanisms. By configuring virtual sensors and actuators at key process nodes, the model acquires sensing and driving capabilities, ultimately integrating into a digital twin model capable of responding to commands and simulating continuous operation. This model is connected to the physical control system via a real-time data link, achieving bidirectional synchronization of control signals and status feedback. This creates a virtual production environment consistent with the behavior of the physical world, providing a computable digital foundation for subsequent full-process dynamic traceability of materials.

[0030] Furthermore, considering the integration and verification process of the digital twin model from "static assembly" to "dynamic operation," the corresponding processing steps are as follows: Configure corresponding electromechanical signals for virtual sensors and virtual actuators. These electromechanical signals include input and output signals. Input signals are used to receive external control commands, and output signals are used to provide feedback on the virtual sensor status. Based on the continuous process flow of the extrusion granulation production line, establish a simulation sequence in the digital twin model. The simulation sequence defines the flow logic of virtual materials between feeding, mixing, extrusion, and granulation units, the action sequence and coordination rules of each virtual actuator, and the status of each virtual sensor. The system integrates the three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, electromechanical signal mapping relationships, and simulation sequence logic into a digital twin model that can operate independently. The digital twin model interacts with external systems through electromechanical signal mapping relationships. Simulated or real-time data communication links are input into the digital twin model to drive its operation. The system verifies whether the digital twin model can accurately simulate the complete process of material feeding to granulation according to the instructions and simulation sequence logic, and provides feedback on the corresponding operating status through output signals.

[0031] Specifically, the system establishes a "neural interface" for the model to interact with the outside world by configuring electromechanical signals, and then constructs a "simulation sequence" that defines the material flow and equipment coordination logic based on the actual process flow. Subsequently, all components, including geometry, physics, and control logic, are integrated with the aforementioned interface and sequence, encapsulated into a complete, independently operable digital twin model. Finally, the model is driven by injected instructions to verify the accuracy of its simulation of continuous production processes and the realism of its status feedback. Its core function is to ensure that the constructed virtual model is not merely a three-dimensional visualization shell, but a dynamic system with correct process logic, capable of real-time response, and high-fidelity simulation of the physical world, thereby providing a reliable and consistent digital sandbox environment for subsequent precise material traceability.

[0032] In some embodiments, step S20 specifically includes the following steps: In the digital twin model, a data structure for material digital identifiers is defined, the data structure including at least the raw material batch number, the proportion of formula components, and physical property parameters; In the digital twin model, an identifier injection logic unit is configured corresponding to the raw material inlet position of the physical production line, the identifier injection logic unit listens to and receives raw material batch switching event signals and associated batch information content from the physical control system through a real-time data communication link; A virtual time synchronization mapping is established to keep the simulation time in the digital twin model synchronized with the real-time clock of the physical production line, ensuring that the position of the material micro-element on the virtual time axis corresponds to the physical production time; When the identifier injection logic unit receives a raw material batch switching event signal, a new material digital identifier is created according to the received batch information content, and it is bound to the virtual timestamp of the event occurrence time; According to the virtual time synchronization mapping, the virtual material micro-element corresponding to the virtual timestamp is located in the virtual material flow of the digital twin model, and the bound material digital identifier is dynamically injected into the corresponding material micro-element as a traceable virtual identity.

[0033] It should be understood that the "material micro-element" is the bridge connecting "batch events" and "traceability calculations," and is the basic unit of algorithm processing. In the digital twin model, it is the smallest traceable virtual material unit defined to achieve accurate traceability of continuously flowing materials. It makes the traceability of continuous materials a feasible and programmable process for the identification propagation calculation of discrete virtual units.

[0034] Specifically, this involves defining a "material digital identifier" data structure that carries information such as batch and formula; setting up a logical unit at the virtual raw material entry point to monitor physical batch switching events; establishing a strict virtual-real time synchronization mechanism; instantly creating a new identifier containing batch information and binding a timestamp when an event is detected; and finally, precisely injecting this identifier into the smallest unit of the virtual material flow—the "material micro-element"—according to the time mapping. Its fundamental function is to transform discrete batch switching events in the physical world into discrete information markers on a continuous material flow in the virtual world, thus laying the crucial data foundation for accurately tracking the source, mixing, and destination of each material in the digital twin environment.

[0035] In practice, the material element at position p in the equipment space at time t is defined as a tuple carrying a set of identifiers: ; Where M represents a material element, t is the simulation timestamp, and p is the coordinate vector of the element in the three-dimensional virtual space. This represents the numerical identifier of the i-th material injected into this micro-element. It is itself a data structure containing attributes such as raw material batch number and formula. A micro-element may contain multiple identifiers, representing the mixing state of different batches of raw materials at this location.

[0036] In some embodiments, step S30 specifically includes the following steps: Based on the geometric connection relationships and movement directions of the feeding, mixing, extrusion, and granulation units described by the digital twin model, abstract and establish the material flow directed network topology of the virtual production line, wherein network nodes correspond to equipment functional units and directed edges correspond to material transmission paths; define input and output interfaces for the material identification propagation dynamic model, the input interface is used to continuously receive material digital identifications, injection time, and real-time production data, and the output interface is used to output dynamic change trajectories; configure material identification propagation calculation rules for various nodes and directed edges in the material flow directed network topology, and calculate the spatial displacement, time delay, and mixing state with other identifications of the material digital identifications when they flow through the corresponding nodes or edges based on equipment mechanisms or data-driven methods and real-time production data; when the material identification propagation dynamic model is running, call the propagation calculation rule library according to the material flow directed network topology and in virtual time order; for each injected material digital identification, driveed by its real-time production data, gradually calculate its spatial position and residence time in the virtual production line equipment, integrate them to form the dynamic change trajectory and data of the corresponding identification, and output them through the output interface.

[0037] To formally describe this dynamic process, a state equation for material identification propagation is established. It is assumed that at time t, the state of all material elements in the virtual production line (including position and the set of identifications they contain) can be represented as a state vector. Its evolution over time is determined by the state transition matrix. and input vector Jointly driven: ; in, It is a time-varying state transition matrix whose elements are determined by the real-time production data at the current moment (such as screw speed, temperature, and flow rate), and is used to calculate the position changes and mixing diffusion of the identifier in the next time step. It is an input vector; when a raw material switching event occurs, its component at the corresponding position will be injected with a new material numerical identifier. B is the input matrix, which determines how the new identifier is assigned to a specific material element.

[0038] Specifically, the system abstracts a material flow network topology from the digital twin model, with equipment as nodes and transmission paths as edges. It sets explicit data interfaces for the model to receive identifier, time, and process data and output trajectories. It configures mechanism- or data-based propagation calculation rules for each node and edge in the network to simulate the displacement, delay, and mixing of identifiers as they flow through the network. During model execution, these rules are driven sequentially according to the topology, ultimately calculating and outputting the complete spatiotemporal trajectory of each material digital identifier in the virtual space. This transforms the abstract material traceability requirement into a computable dynamic simulation model, enabling quantitative simulation and prediction of the diffusion, mixing, and evolution processes of different raw material batches in complex continuous processes, providing direct data support for batch segmentation and traceability genealogy generation.

[0039] In some embodiments, step S40 specifically includes the following steps: acquiring the dynamic change trajectory output by the material identification propagation dynamic model, spatiotemporally associating virtual material micro-elements carrying the same material digital identifier, and reconstructing the complete virtual process of the raw material carrying the corresponding material digital identifier from the injection point through each virtual unit of feeding, mixing, extrusion, and granulation to the finished product station; continuously monitoring the state of the virtual material flow when it arrives at the preset finished product station in the digital twin model, determining the types, proportions, and distribution of multiple material digital identifiers contained in the virtual material flow based on the current state, and calculating the corresponding mixed state quantification index; comparing the mixed state quantification index with the uniformity threshold in real time, and when the index fluctuation exceeds the threshold, determining it as a batch boundary and using it as a dynamic cutting point to divide the previously continuously produced virtual material flow into an independent dynamic finished product batch with internal uniformity; automatically associating and aggregating the complete virtual process of all raw materials constituting the corresponding batch for each divided independent dynamic finished product batch; and integrating the raw material batch information, process path, and key process parameter history contained in the complete virtual process to generate a structured material genealogy traceability map.

[0040] To achieve adaptive dynamic batch partitioning and avoid misjudgment due to instantaneous fluctuations, the decision model in this embodiment is constructed as follows: Let the mixing degree index sequence obtained from the finished product station sampling window be... The model evaluates in real time the posterior probability of a batch switch (i.e., the point of change) occurring at the current time t. : ; in, It is a discrete variable representing the current data point. The corresponding batch / regime. It is the state transition probability, which encodes the prior expectation of the batch duration (e.g., a longer stable running segment has a higher probability). It is the observational likelihood, assuming that within the same batch segment, It follows a Gaussian distribution centered at the mean of this segment. It is the posterior belief from the previous moment. When the posterior probability... When a dynamic confidence threshold is exceeded, the system determines that a valid batch switch has occurred at time t, thus creating a new batch. This model can distinguish between systematic trend changes and random noise, achieving robust and adaptive dynamic batch segmentation.

[0041] Specifically, the system first uses dynamic trajectory data of material identifiers to reconstruct the complete virtual processing history of each raw material. Then, it monitors the composition of the material flow in real time at the virtual finished product station, calculates mixing state indicators and compares them with preset thresholds to intelligently identify batch cutting points, thus dynamically dividing the continuously flowing virtual material flow into uniform finished product batches. Finally, the system automatically aggregates the complete process information of all materials constituting each batch, generating a structured genealogical traceability map. This transforms the microscopic material movement data obtained from the preceding steps into batch conclusions that are directly applicable to production management and quality analysis.

[0042] Furthermore, considering the following processing steps: In the digital twin model, a virtual material analysis window is set up corresponding to the finished product station. The material analysis window simulates the instantaneous sampling behavior of continuous output on the physical production line, used to capture and analyze the virtual material flow segment passing through the corresponding station at a certain moment; the virtual material flow segment captured in the material analysis window is parsed to identify and extract all the different material digital identifiers contained therein; the number of virtual material micro-elements associated with each different material digital identifier or the virtual material mass they represent is counted; based on the statistical results, the mass ratio or quantity ratio of each material digital identifier in the virtual material flow segment is calculated to obtain the ratio data; at the same time, the spatial or temporal distribution dispersion of the virtual material micro-elements associated with each material digital identifier in the virtual material flow segment is analyzed to obtain the distribution characteristics; based on the calculated ratio data and distribution characteristics, a mixed state quantification index is comprehensively generated. The mixed state quantification index includes at least the concentration ratio of the main identifier, the number of identifier types, and the uniformity index of the mixing between different identifiers.

[0043] To comprehensively evaluate mixing uniformity, a mixing degree index H based on information entropy is introduced. Assume that in the material flow segment captured by the analysis window, there are N material micro-elements, containing m different material numerical identifiers. Let... Let H be the mass fraction (or infinitesimal fraction) of the j-th identifier. Then the mixing degree H of this segment is defined as: ; Among them, when the material consists entirely of a single identifier ( When m = 1), H = 0, indicating that they are completely unmixed; when all m identifiers are uniformly distributed ( When H reaches its maximum value (=1 / m), H achieves its maximum value. The more uniform the mixing, the better. By calculating the H value in real time and comparing it with a preset uniformity threshold, the mixing state can be quantitatively determined, providing an accurate basis for dynamic batch division.

[0044] Specifically, by setting up a virtual material analysis window that simulates physical sampling, the system can capture material flow segments at specific moments and accurately analyze the types of all material digital identifiers within them, statistically determine the mass or quantity proportion of each identifier, and analyze the uniformity of these identifiers' distribution within the segment. Based on the proportion and distribution characteristics, the system comprehensively calculates key quantitative indicators of the mixing state. This transforms the previously difficult-to-quantify continuous material mixing process into a set of precise data (such as concentration ratio and uniformity index) that can be measured in real time and compared with thresholds. This provides a reliable quantitative basis for subsequent adaptive dynamic batch segmentation decisions, achieving a crucial improvement in traceability accuracy from qualitative to quantitative.

[0045] The implementation principle of the full-process material traceability method for an extrusion granulation production line according to this application embodiment is as follows: A high-fidelity digital twin model fully synchronized with the physical production line is constructed, covering the entire unit geometry, physical, and control logic from feeding to granulation, and achieving virtual-real mapping through a real-time data link. In the virtual environment, the system converts each raw material batch switching event into an identifiable material digital identifier and dynamically injects it into the corresponding virtual material micro-element. Subsequently, based on the equipment structure and motion relationships defined by the twin model, a dynamic model for material identifier propagation is constructed. This model, driven by real-time process data, continuously calculates the dynamic change trajectory of the spatial position and temporal attributes of each identifier within the virtual equipment. Finally, at the virtual finished product station, the system monitors the mixing state of different identifiers in the material flow, dynamically divides the internally uniform finished product batches through intelligent comparison with a preset uniformity threshold, and automatically generates a detailed material genealogy traceability map displaying the batch. This upgrades the traditional static, discrete batch records to a dynamic, computable, full-process digital traceability that evolves synchronously with the production process, thus providing digital assurance for precise quality control and process root cause analysis.

[0046] Figure 1 This is a flowchart illustrating a full-process material traceability method for an extrusion granulation production line in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0047] Based on the same technical concept, referring to Figure 2 This application also provides a material traceability device for the entire process of an extrusion granulation production line, which adopts the following technical solution: The device includes: The data mapping module is used to build a digital twin model based on the extrusion granulation production line. The digital twin model contains the geometric, physical and control logic definitions corresponding to the feeding, mixing, extrusion and granulation units in the physical production line. A real-time data communication link is built between the digital twin model and the physical production line control system. The digital twin module is used to define virtual material digital identifiers for raw materials entering the production line in the digital twin model; it uses raw material batch switching events and corresponding batch information in the physical production line as trigger signals and content, and dynamically injects the associated material digital identifiers into the material micro-elements at the corresponding time points in the virtual material flow; The dynamic change module is used to construct a dynamic model for material identification propagation based on the equipment structure and motion relationship described by the digital twin model. The dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data as continuous inputs, to calculate the dynamic change trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment. The traceability graph module is used to associate and calculate the complete process of a specific virtual material micro-element from raw material to finished product in real time based on the dynamic change trajectory of the material digital identifier. When the virtual material flow arrives at the preset finished product station, the finished product batch is dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material lineage traceability graph corresponding to the dynamic batch is generated. The material lineage traceability graph includes the raw material batch information, process path and key process parameter history that led to the composition of the corresponding batch of finished product.

[0048] In some embodiments, the data mapping module is specifically used to construct a three-dimensional geometric model corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line, and to perform virtual assembly to form the overall three-dimensional layout of the production line. Define physical properties for key components in the 3D geometric model. These physical properties include mass, inertia, collision body properties, and material friction coefficient. Based on the actual motion mechanism of each unit, kinematic pairs and constraints are defined in the three-dimensional geometric model. Rotary pairs are configured for the screw and cutter, and sliding pairs are configured for the pusher cylinder and valve. Virtual sensors and virtual actuators are configured at key process nodes in the three-dimensional geometric model. Virtual sensors are used to detect material positions and equipment status in the simulation, while virtual actuators are used to receive control signals and drive the movement of virtual components. Based on a three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, a digital twin model is integrated to form a digital twin model that can respond to control commands and simulate the continuous operation of the production line. The digital twin model is connected to the control system of the physical production line through a real-time data communication link. The physical control system sends real-time production data as input signals to the digital twin model to drive the virtual actuators. The digital twin model sends the status feedback of the virtual sensors as output signals back to the physical control system, thus achieving virtual-real synchronization.

[0049] In some embodiments, the data mapping module is specifically used to configure corresponding electromechanical signals for virtual sensors and virtual actuators. The electromechanical signals include input signals and output signals. The input signals are used to receive external control commands, and the output signals are used to provide feedback on the virtual sensor status. Based on the continuous process flow of the extrusion granulation production line, a simulation sequence is established in the digital twin model. The simulation sequence defines the flow logic of virtual materials between the feeding, mixing, extrusion and granulation units, the action sequence and coordination rules of each virtual actuator, and the triggering conditions of each virtual sensor. The three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, electromechanical signal mapping relationships and simulation sequence logic are systematically integrated and encapsulated to form a digital twin model that can run independently. The digital twin model interacts with external systems through electromechanical signal mapping relationships. Input simulated or real-time data communication link control commands into the digital twin model to drive the model to run, verify whether the digital twin model can accurately simulate the complete process of material feeding to granulation according to the commands and simulation sequence logic, and provide feedback on the corresponding operating status through output signals.

[0050] In some embodiments, the digital twin module is specifically used to define a data structure for the digital identifier of materials in the digital twin model. The data structure includes at least the raw material batch number, the proportion of ingredients in the formula, and physical property parameters. In the digital twin model, an identification injection logic unit is configured corresponding to the raw material inlet position of the physical production line. The identification injection logic unit listens for and receives raw material batch switching event signals and associated batch information from the physical control system through a real-time data communication link. Establish a virtual time synchronization mapping to keep the simulation time in the digital twin model synchronized with the real-time clock of the physical production line, ensuring that the position of the material micro-element on the virtual time axis corresponds to the physical production time; When the identifier injection logic unit receives a raw material batch switching event signal, it creates a new material digital identifier based on the received batch information and binds it to the virtual timestamp of the event occurrence time. Based on the virtual time synchronization mapping, the virtual material micro-element corresponding to the virtual timestamp is located in the virtual material flow of the digital twin model, and the bound material digital identifier is dynamically injected into the corresponding material micro-element as a traceable virtual identity.

[0051] In some embodiments, the dynamic change module is specifically used to abstract and establish the directed network topology of material flow in the virtual production line based on the geometric connection relationship and movement direction of the feeding, mixing, extrusion and granulation units described by the digital twin model, where network nodes correspond to equipment functional units and directed edges correspond to material transport paths. The input and output interfaces are defined for the dynamic model of material identification propagation. The input interface is used to continuously receive material digital identification, injection time and real-time production data, and the output interface is used to output the dynamic change trajectory. Configure material identifier propagation calculation rules for various nodes and directed edges in the directed network topology of material flow. Based on equipment mechanism or data-driven, calculate the spatial displacement, time delay and mixing state with other identifiers of material digital identifiers when they flow through corresponding nodes or edges according to real-time production data. When the material identification propagation dynamic model is running, the propagation calculation rule library is called according to the directed network topology of the material flow and in virtual time sequence. For each injected material digital identifier, its spatial position and residence time in the virtual production line equipment are calculated step by step, driven by its real-time production data, and integrated to form the dynamic change trajectory and data of the corresponding identifier, and output through the output interface.

[0052] In some embodiments, the traceability map module is specifically used to obtain the dynamic change trajectory output by the material identification propagation dynamic model, and to spatiotemporally correlate virtual material micro-elements carrying the same material digital identifier, and to reversely reconstruct the complete virtual process of the raw material carrying the corresponding material digital identifier from the injection point through each virtual unit of feeding, mixing, extrusion, and granulation to the finished product station. In the digital twin model, the status of the virtual material flow when it arrives at the preset finished product station is continuously monitored, the types, proportions and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status are determined, and the corresponding mixed state quantitative indicators are calculated. The mixed state quantification index is compared with the uniformity threshold in real time. When the index fluctuation exceeds the threshold, it is determined as the batch boundary and used as a dynamic cutting point to divide the previously continuously produced virtual material flow into an independent dynamic finished product batch with internal uniformity. For each segmented independent dynamic finished product batch, the system automatically associates and aggregates all raw materials that constitute the complete virtual process of the corresponding batch; it integrates the raw material batch information, process path and key process parameter history contained in the complete virtual process to generate a structured material genealogy traceability map.

[0053] In some embodiments, the traceability mapping module is specifically used to set up a virtual material analysis window in the digital twin model corresponding to the finished product station. The material analysis window simulates the behavior of instantaneous sampling of continuous outputs on the physical production line, and is used to capture and analyze a virtual material flow segment passing through the corresponding station at a certain moment. The virtual material flow segment captured in the material analysis window is parsed to identify and extract all the different material digital identifiers contained therein; the number of virtual material micro-elements or the virtual material quality represented by each different material digital identifier is counted. Based on the statistical results, the mass or quantity proportion of each material digital identifier in the virtual material flow segment is calculated to obtain the proportion data; at the same time, the spatial or temporal distribution dispersion of the virtual material micro-elements associated with each material digital identifier in the virtual material flow segment is analyzed to obtain the distribution characteristics. Based on the calculated proportion data and distribution characteristics, a mixed state quantitative index is generated. The mixed state quantitative index includes at least the concentration ratio of the main markers, the number of marker types, and the uniformity index of the mixing between different markers.

[0054] This application also discloses a control device.

[0055] Specifically, the control device includes a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed to perform the material traceability method for the entire extrusion granulation production line described above.

[0056] This application also discloses a computer-readable storage medium.

[0057] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed as described above in the material traceability method for the entire extrusion granulation production line. The computer-readable storage medium includes, for example, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0058] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for tracing materials throughout the entire process of an extrusion granulation production line, characterized in that, include: A digital twin model is constructed based on the extrusion granulation production line. The digital twin model includes the geometric, physical, and control logic definitions corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line. A real-time data communication link is established between the digital twin model and the physical production line control system. In the digital twin model, virtual material digital identifiers are defined for raw materials entering the production line; raw material batch switching events and corresponding batch information in the physical production line are used as trigger signals and content, and the associated material digital identifiers are dynamically injected into the material micro-elements at the corresponding time points in the virtual material flow; Based on the equipment structure and motion relationship described by the digital twin model, a dynamic model for material identification propagation is constructed. The dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data, as continuous inputs to calculate the dynamic trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment. Based on the dynamic change trajectory of the material digital identifier, the complete process of a specific virtual material micro-element from raw material to finished product is correlated and calculated in real time; when the virtual material flow arrives at the preset finished product station, the finished product batch is dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material lineage traceability map corresponding to the dynamic batch is generated. The material lineage traceability map includes the raw material batch information, process path and key process parameter history that lead to the composition of the corresponding batch of finished product. Specifically, based on the equipment structure and motion relationship described by the digital twin model, a dynamic model for material identification propagation is constructed. This dynamic model uses the injected material digital identifiers and their injection time, as well as real-time production data, as continuous inputs to calculate the dynamic trajectory of the spatial position and temporal attributes of the material digital identifiers within the virtual production line equipment. This includes: Based on the geometric connection relationships and movement directions of the feeding, mixing, extrusion and granulation units described by the digital twin model, the directed network topology of the material flow of the virtual production line is abstracted and established, where network nodes correspond to equipment functional units and directed edges correspond to material transmission paths. An input interface and an output interface are defined for the dynamic model of material identification propagation. The input interface is used to continuously receive material digital identification, injection time and real-time production data, and the output interface is used to output the dynamic change trajectory. Configure material identifier propagation calculation rules for various nodes and directed edges in the directed network topology of material flow. Based on equipment mechanism or data-driven, calculate the spatial displacement, time delay and mixing state with other identifiers of material digital identifiers when they flow through corresponding nodes or edges according to real-time production data. When the material identification propagation dynamic model is running, the propagation calculation rule library is called according to the material flow directed network topology and virtual time sequence. For each injected material digital identifier, its spatial position and residence time in the virtual production line equipment are calculated step by step, driven by its real-time production data, and integrated to form the dynamic change trajectory and data of the corresponding identifier, and output through the output interface. The dynamic change trajectory based on the material digital identifier is used to correlate and calculate the complete process of a specific virtual material micro-element from raw material to finished product in real time. When the virtual material flow arrives at the preset finished product station, the finished product batches are dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material genealogy traceability map corresponding to the dynamic batches is generated, including: The dynamic change trajectory output by the material identification propagation dynamic model is obtained, and the virtual material micro-elements carrying the same material digital identifier are spatiotemporally correlated. The complete virtual process of the raw material carrying the corresponding material digital identifier from the injection point through each virtual unit of feeding, mixing, extrusion, and granulation to the finished product station is reconstructed in reverse. In the digital twin model, the status of the virtual material flow when it arrives at the preset finished product station is continuously monitored, the types, proportions and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status are determined, and the corresponding mixed state quantitative indicators are calculated. The mixed state quantification index is compared with the uniformity threshold in real time. When the index fluctuation exceeds the threshold, it is determined as the batch boundary and used as a dynamic cutting point to divide the previously continuously produced virtual material flow into an independent dynamic finished product batch with internal uniformity. For each segmented independent dynamic finished product batch, the system automatically associates and aggregates all raw materials that constitute the complete virtual process of the corresponding batch; it then integrates the raw material batch information, process path, and key process parameter history contained in the complete virtual process to generate a structured material family traceability map.

2. The method for full-process material traceability in an extrusion granulation production line according to claim 1, characterized in that, The digital twin model constructed based on the extrusion granulation production line includes geometric, physical, and control logic definitions corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line. A real-time data communication link is established between the digital twin model and the physical production line control system, including: Construct a three-dimensional geometric model corresponding to the feeding, mixing, extrusion, and granulation units in the physical production line, and perform virtual assembly to form the overall three-dimensional layout of the production line; Define physical properties for key components in the three-dimensional geometric model. These physical properties include mass, inertia, collision body properties, and material friction coefficient. Based on the actual motion mechanism of each unit, kinematic pairs and constraints are defined in the three-dimensional geometric model. Rotary pairs are configured for the screw and cutter, and sliding pairs are configured for the pusher cylinder and valve. Virtual sensors and virtual actuators are configured at key process nodes in the three-dimensional geometric model. The virtual sensors are used to detect material positions and equipment status in the simulation, and the virtual actuators are used to receive control signals and drive the movement of virtual components. Based on the aforementioned three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, a digital twin model is integrated to form a digital twin model that can respond to control commands and simulate the continuous operation of the production line. The digital twin model is connected to the control system of the physical production line through a real-time data communication link. The physical control system sends real-time production data as input signals to the digital twin model to drive the virtual actuators. The digital twin model sends the status feedback of the virtual sensors as output signals back to the physical control system, thus achieving virtual-real synchronization.

3. The method for full-process material traceability in an extrusion granulation production line according to claim 2, characterized in that, The digital twin model, based on the aforementioned three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, is integrated to form a digital twin model. This digital twin model can respond to control commands and simulate the continuous operation of the production line, including: Configure corresponding electromechanical signals for virtual sensors and virtual actuators. The electromechanical signals include input signals and output signals. The input signals are used to receive external control commands, and the output signals are used to provide feedback on the status of the virtual sensors. Based on the continuous process flow of the extrusion granulation production line, a simulation sequence is established in the digital twin model. The simulation sequence defines the flow logic of virtual materials between the feeding, mixing, extrusion and granulation units, the action sequence and coordination rules of each virtual actuator, and the triggering conditions of each virtual sensor. The three-dimensional geometric model, physical properties, kinematic pairs and constraints, virtual sensors and virtual actuators, electromechanical signal mapping relationships and simulation sequence logic are systematically integrated and encapsulated to form an independently operable digital twin model. The digital twin model interacts with external systems through electromechanical signal mapping relationships. Input simulated or real-time data communication link control commands into the digital twin model to drive the model to run, verify whether the digital twin model can accurately simulate the complete process of material feeding to granulation according to the commands and simulation sequence logic, and provide feedback on the corresponding operating status through output signals.

4. The method for full-process material traceability in an extrusion granulation production line according to claim 1, characterized in that, In the digital twin model, virtual material digital identifiers are defined for raw materials entering the production line; raw material batch switching events and corresponding batch information in the physical production line are used as trigger signals and content, and the associated material digital identifiers are dynamically injected into the material micro-elements at the corresponding time points in the virtual material flow, including: In the digital twin model, a data structure for the digital identification of materials is defined, and the data structure includes at least the raw material batch number, the proportion of ingredients in the formula, and physical property parameters; In the digital twin model, an identification injection logic unit is configured corresponding to the raw material inlet position of the physical production line. The identification injection logic unit listens for and receives raw material batch switching event signals and associated batch information from the physical control system through a real-time data communication link. Establish a virtual time synchronization mapping to keep the simulation time in the digital twin model synchronized with the real-time clock of the physical production line, ensuring that the position of the material micro-element on the virtual time axis corresponds to the physical production time; When the identifier injection logic unit receives a raw material batch switching event signal, it creates a new material digital identifier based on the received batch information and binds it to the virtual timestamp of the event occurrence time. Based on the virtual time synchronization mapping, the virtual material micro-element corresponding to the virtual timestamp is located in the virtual material flow of the digital twin model, and the bound material digital identifier is dynamically injected into the corresponding material micro-element as a traceable virtual identity.

5. The method for full-process material traceability in an extrusion granulation production line according to claim 1, characterized in that, In the digital twin model, the status of the virtual material flow when it arrives at the preset finished product station is continuously monitored. The types, proportions, and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status are determined, and the corresponding mixed-state quantification indicators are calculated, including: In the digital twin model, a virtual material analysis window is set up corresponding to the finished product station. The material analysis window simulates the behavior of instantaneous sampling of continuous output on the physical production line, and is used to capture and analyze a virtual material flow segment passing through the corresponding station at a certain moment. The virtual material flow segment captured in the material analysis window is parsed to identify and extract all the different material digital identifiers contained therein; the number of virtual material micro-elements or the virtual material quality represented by each different material digital identifier is counted. Based on the statistical results, the mass or quantity proportion of each material digital identifier in the virtual material flow segment is calculated to obtain the proportion data; at the same time, the spatial or temporal distribution dispersion of the virtual material micro-elements associated with each material digital identifier in the virtual material flow segment is analyzed to obtain the distribution characteristics. Based on the calculated proportion data and the distribution characteristics, the mixed state quantification index is generated. The mixed state quantification index includes at least the concentration ratio of the main identifier, the number of identifier types, and the uniformity index of the mixing between different identifiers.

6. A material traceability device for the entire process of an extrusion granulation production line, characterized in that, The device includes: The data mapping module is used to build a digital twin model based on the extrusion granulation production line. The digital twin model contains the geometric, physical and control logic definitions corresponding to the feeding, mixing, extrusion and granulation units in the physical production line. A real-time data communication link is built between the digital twin model and the physical production line control system. The digital twin module is used to define virtual material digital identifiers for raw materials entering the production line in the digital twin model; it uses raw material batch switching events and corresponding batch information in the physical production line as trigger signals and content, and dynamically injects the associated material digital identifiers into the material micro-elements at the corresponding time points in the virtual material flow; The dynamic change module is used to construct a dynamic model for material identification propagation based on the equipment structure and motion relationship described by the digital twin model. The dynamic model for material identification propagation uses the injected material digital identification and its injection time, as well as real-time production data as continuous inputs, to calculate the dynamic change trajectory of the spatial position and temporal attributes of the material digital identification within the virtual production line equipment. The traceability graph module is used to associate and calculate the complete process of a specific virtual material micro-element from raw material to finished product in real time based on the dynamic change trajectory of the material digital identifier; when the virtual material flow arrives at the preset finished product station, the finished product batch is dynamically divided according to the mixing state of the material digital identifier and the preset uniformity threshold, and a material lineage traceability graph corresponding to the dynamic batch is generated. The material lineage traceability graph includes the raw material batch information, process path and key process parameter history that lead to the composition of the corresponding batch of finished product. Specifically, the dynamic change module is used to abstract and establish the directed network topology of material flow in the virtual production line based on the geometric connection relationship and movement direction of the feeding, mixing, extrusion and granulation units described by the digital twin model. The network nodes correspond to the equipment functional units and the directed edges correspond to the material transmission paths. An input interface and an output interface are defined for the dynamic model of material identification propagation. The input interface is used to continuously receive material digital identification, injection time and real-time production data, and the output interface is used to output the dynamic change trajectory. Configure material identifier propagation calculation rules for various nodes and directed edges in the directed network topology of material flow. Based on equipment mechanism or data-driven, calculate the spatial displacement, time delay and mixing state with other identifiers of material digital identifiers when they flow through corresponding nodes or edges according to real-time production data. When the material identification propagation dynamic model is running, the propagation calculation rule library is called according to the material flow directed network topology and virtual time sequence. For each injected material digital identifier, its spatial position and residence time in the virtual production line equipment are calculated step by step, driven by its real-time production data, and integrated to form the dynamic change trajectory and data of the corresponding identifier, and output through the output interface. Specifically, the traceability map module is used to obtain the dynamic change trajectory output by the material identification propagation dynamic model, and to spatiotemporally associate virtual material micro-elements carrying the same material digital identifier, and to reversely reconstruct the complete virtual process of raw materials carrying the corresponding material digital identifier from the injection point through each virtual unit of feeding, mixing, extrusion, and granulation to the finished product station. In the digital twin model, the status of the virtual material flow when it arrives at the preset finished product station is continuously monitored, the types, proportions and distribution of multiple material digital identifiers contained in the virtual material flow based on the current status are determined, and the corresponding mixed state quantitative indicators are calculated. The mixed state quantification index is compared with the uniformity threshold in real time. When the index fluctuation exceeds the threshold, it is determined as the batch boundary and used as a dynamic cutting point to divide the previously continuously produced virtual material flow into an independent dynamic finished product batch with internal uniformity. For each segmented independent dynamic finished product batch, the system automatically associates and aggregates all raw materials that constitute the complete virtual process of the corresponding batch; it then integrates the raw material batch information, process path, and key process parameter history contained in the complete virtual process to generate a structured material family traceability map.

7. A control device, characterized in that, The device includes: A memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1 to 5.

Images