A digital twin management and control system for carbon fiber composite material production process

Abstract

The application relates to the technical field of digital twinning and intelligent manufacturing, and particularly discloses a carbon fiber composite material production process digital twinning management and control system. The application realizes dynamic digital mapping of the microstructure evolution of a physical entity and the formation process of macroscopic performance by collecting multi-process process parameters and environmental parameters in real time and constructing a multi-scale digital twinning model, solves the problems of serious black-boxing of a traditional manufacturing process, lagging process deviation discovery and broken quality traceability chain, and improves the full-life-cycle closed-loop management and control capability and manufacturing consistency of carbon fiber composite materials.

Description

Technical Field

[0001] This invention relates to the field of digital twin and intelligent manufacturing technology, and in particular to a digital twin control system for the production process of carbon fiber composite materials. Background Technology

[0002] Carbon fiber composites, due to their superior properties such as high specific strength, high specific modulus, and strong designability, have been widely used in high-end equipment manufacturing fields such as aerospace, rail transportation, and new energy vehicles. Their production quality directly determines the structural safety and service reliability of the final products. However, the production process of carbon fiber composites involves multiple complex steps, including prepreg laying, autoclave curing, and resin transfer molding. Process parameters are highly coupled, and many environmental factors influence the process. Current technologies mainly rely on manual experience to set process windows, supplemented by offline sampling and testing, resulting in a severe black box effect in the production process. It is difficult to perceive the dynamic correlation between the evolution of the material's internal microstructure and the formation of its macroscopic properties in real time. At the same time, traditional manufacturing execution systems (MES) focus primarily on production planning and equipment status monitoring, lacking the ability to digitally map all elements and processes of the physical entity. This leads to delayed detection of process deviations, broken quality traceability chains, and an inability to achieve closed-loop management throughout the entire lifecycle from raw materials to finished products, severely restricting the consistency and yield improvement of carbon fiber composite manufacturing.

[0003] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a digital twin control system for the production process of carbon fiber composite materials. The technical solution of this system is as follows: The data acquisition module is used to generate real-time manufacturing data by collecting process parameters and environmental parameters of multiple processes in the production of carbon fiber composite materials in real time. The digital twin modeling module is used to construct a multi-scale digital twin model based on material properties and process mechanisms, and to use the real-time manufacturing data to drive the multi-scale digital twin model to perform dynamic digital mapping, generating a dynamic digital twin that maps the evolution of the microstructure and the formation of the macroscopic properties of the physical material. The simulation analysis and decision-making module is used to perform process simulation and performance evolution simulation based on the dynamic digital twin, compare and analyze the simulation results with the predefined process specifications in real time to obtain process deviation data, and generate process parameter adjustment instructions based on the process deviation data. A precision execution control module is used to convert the process parameter adjustment command into a control signal for a physical entity actuator, and the control signal is used to drive the actuator to adjust the process parameter. The quality traceability and visualization module is used to integrate the real-time manufacturing data, the evolution data of the dynamic digital twin, and the process parameter adjustment instructions to generate and display a quality traceability chain covering the entire life cycle from raw materials to finished products.

[0005] The technical solution of this invention achieves dynamic digital mapping of the evolution of the microstructure and the formation of macroscopic properties of physical entities by real-time acquisition of multi-process parameters and environmental parameters and construction of multi-scale digital twin models. This solves the problems of severe black box nature in traditional manufacturing processes, delayed detection of process deviations, and broken quality traceability chains, and improves the closed-loop management capability and manufacturing consistency of the entire life cycle of carbon fiber composite material production.

[0006] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0008] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of an embodiment of a digital twin control system for the production process of carbon fiber composite materials according to the present invention. Detailed Implementation

[0009] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0010] Figure 1 A schematic diagram of an embodiment of a digital twin control system for carbon fiber composite material production processes provided by the present invention is shown. Figure 1 As shown, the system includes: The data acquisition module 110 is used to generate real-time manufacturing data by acquiring process parameters and environmental parameters of multiple processes in the production of carbon fiber composite materials in real time.

[0011] Carbon fiber composite material production refers to the manufacturing process of combining carbon fiber as reinforcement and resin as matrix through a series of processing steps to form high-performance materials. For example, in the manufacture of aircraft structural components, carbon fiber prepreg is laid up and then placed in an autoclave for high-temperature and high-pressure curing. Multiple processes refer to the sequential execution of several independent processing stages in the manufacturing of carbon fiber composite materials. For example, the production of a structural component may involve multiple processes such as automatic prepreg laying, autoclave curing, and possible resin transfer molding. Process parameters refer to adjustable variables preset to control the production process and material properties. For example, the process parameters for autoclave curing include heating rate, target temperature, pressure value, and holding time. Environmental parameters refer to physical condition variables in the production site that affect the process but are not directly set by the equipment. For example, the initial temperature, ambient humidity, and ambient atmospheric pressure inside the autoclave before curing begins. Real-time manufacturing data refers to the structured set of process parameters and environmental parameters collected and processed synchronously during the production process. For example, the system records the temperature, pressure, and ambient humidity inside the autoclave every second, and the processed data is time-series data with a unified timestamp.

[0012] The digital twin modeling module 120 is used to construct a multi-scale digital twin model based on material properties and process mechanisms, and to use the real-time manufacturing data to drive the multi-scale digital twin model to perform dynamic digital mapping, thereby generating a dynamic digital twin that maps the microstructure evolution and macroscopic performance formation process of the physical material.

[0013] Material properties refer to the inherent physical and chemical attributes of the components constituting the composite material; for example, the tensile modulus of carbon fiber and the viscosity-temperature characteristics of resin are material properties. Processing mechanism refers to the principles and laws governing the physical changes and chemical reactions that occur in the composite material during processing; for example, the phase transition process of resin from liquid to gel to solid under heating conditions and the accompanying internal stress generation principle. Multi-scale digital twin models refer to integrated computational models that can simultaneously characterize the material's microstructure, the process's mesoscopic physical field, and the product's macroscopic performance, with data interaction relationships between models at each scale; for example, a software model that couples a resin micro-curing model, a tank temperature field simulation model, and a component stiffness prediction model.

[0014] Dynamic digital mapping refers to the process of using real-time data to drive a digital twin model, allowing its state to evolve synchronously with the physical production process. For example, inputting real-time temperature data from an autoclave into the model drives the calculation of the current resin curing degree and the internal stress distribution of the component. The evolution of the microstructure of the physical material refers to the dynamic changes in the microscopic characteristics of the composite material during production, such as fiber orientation, pore morphology, and resin cross-linking networks. For example, changes in resin flowability during the curing stage lead to pore migration and the formation of resin molecular cross-linking networks. The macroscopic performance formation process refers to the gradual formation and determination of the final mechanical properties and defects of the composite material during manufacturing. For example, the bending stiffness of the component gradually reaches the design value as the resin curing degree increases and the porosity decreases. A dynamic digital twin is a virtual entity generated during the dynamic digital mapping process, corresponding in real-time to the physical production entity, and containing multi-scale state information. For example, a virtual model containing the current temperature field, resin curing degree, and predicted component deformation at a given moment.

[0015] The simulation analysis and decision module 130 is used to perform process simulation and performance evolution simulation based on the dynamic digital twin, compare and analyze the simulation results with the predefined process specifications in real time to obtain process deviation data, and generate process parameter adjustment instructions based on the process deviation data.

[0016] Among them, process simulation and performance evolution simulation refer to the calculation and analysis of subsequent processes based on dynamic digital twins to predict the performance of the final product; for example, simulating the subsequent cooling process based on the current curing state and predicting the residual stress and final strength of the component. Simulation results refer to a series of quantitative data obtained through process simulation and performance evolution simulation calculations; for example, the set of data on the final porosity, tensile strength, and maximum residual stress of the component output by the simulation. Predefined process specifications refer to the target range of process parameters and the qualified standards of product performance set in advance to ensure product quality; for example, the specification stipulates that the curing temperature is maintained at 180℃±5℃ and the tensile strength of the finished product is not less than 1750 MPa. Process deviation data refer to the quantitative description of the difference between the simulation results or measured results and the predefined process specifications; for example, if the simulated predicted strength is 1800 MPa, which is higher than the target value of 1750 MPa, a positive deviation of 50 MPa is the process deviation data. Process parameter adjustment commands refer to operation commands generated to correct or optimize the process and designed to change the set values ​​of process parameters; for example, when it is predicted that excessively rapid heating may lead to excessive residual stress, a command is generated to adjust the heating rate from 2℃ / min to 1.5℃ / min.

[0017] The precision execution control module 140 is used to convert the process parameter adjustment command into a control signal for the physical entity actuator, and the control signal is used to drive the actuator to adjust the process parameters.

[0018] Among them, physical actuators refer to components on production equipment that receive control signals and directly execute actions to change process parameters; for example, electric heaters, air inlet valves, and cooling water regulating valves on autoclaves. Control signals refer to instruction signals generated by the control system and sent to the actuators to drive them to produce specific actions; for example, voltage or current signals sent to electric heaters to adjust their power to 80% of their rated power.

[0019] The quality traceability and visualization module 150 is used to integrate the real-time manufacturing data, the evolution data of the dynamic digital twin, and the process parameter adjustment instructions to generate and display a quality traceability chain covering the entire life cycle from raw materials to finished products.

[0020] The full lifecycle quality traceability chain refers to a complete record of all data affecting quality throughout the entire process from raw materials to the final product, linked by causal relationships and time sequence. For example, the chain can be used to query the batch of carbon fiber used in a component, the data at each moment of the curing process, and all process adjustment instructions.

[0021] The technical solution in this embodiment achieves dynamic digital mapping of the evolution of the microstructure and the formation of macroscopic properties of physical entities by real-time acquisition of multi-process parameters and environmental parameters and construction of multi-scale digital twin models. This solves the problems of severe black box nature in traditional manufacturing processes, delayed detection of process deviations, and broken quality traceability chains, and improves the closed-loop management capability and manufacturing consistency of the entire life cycle of carbon fiber composite material production.

[0022] In one alternative embodiment, the data acquisition module 110 is specifically used for: The process parameters of the multiple steps in the production process of carbon fiber composite materials, including prepreg laying, autoclave curing and resin transfer molding, are collected in real time, and the environmental parameters of the multiple steps are collected in real time simultaneously.

[0023] Among them, the prepreg laying, autoclave curing and resin transfer molding process refers to the names of three typical processes in the production of carbon fiber composite materials and their sequential combination; for example, to manufacture a part with reinforcing ribs, the main body may be laid first, then the reinforcing ribs may be resin transferred molded, and finally the whole part may be co-cured in an autoclave.

[0024] The process parameters and environmental parameters are preprocessed and timestamped, and the processed process parameters and environmental parameters are then structurally integrated to generate the real-time manufacturing data.

[0025] Preprocessing and timestamp alignment refers to filtering, calibrating, and uniformly assigning standard time stamps to the collected raw data to ensure time sequence consistency; for example, after removing outliers from temperature and pressure sensor readings, all data are marked as "2025-10-27 10:30:25".

[0026] Among the above-mentioned optional methods, the process parameters and environmental parameters of multiple processes such as prepreg laying, autoclave curing and resin transfer molding are collected in real time, and the data is preprocessed, timestamp aligned and structured integrated, which solves the problem of spatiotemporal inconsistency of multi-source heterogeneous data and improves the quality and usability of real-time manufacturing data.

[0027] In one alternative embodiment, the digital twin modeling module 120 is specifically used for: Based on the aforementioned material properties of carbon fiber composites, a microscale state evolution model is established to characterize the internal fiber arrangement, pore formation, and resin curing degree evolution of the material.

[0028] The evolution of fiber arrangement, pore formation, and resin curing degree within a material refers to the core material state change process described by the microscale state evolution model. For example, the model calculates the micro-bending of fibers, the annihilation of pore formation, and the evolution of resin curing degree from 0% to 100% under specific process conditions. The microscale state evolution model refers to a mathematical model used to quantitatively describe and predict microscopic processes such as fiber arrangement, pore formation, and resin curing degree evolution within a material. For example, a calculation program based on the finite element method that can output the local curing degree and porosity of the resin based on input process conditions.

[0029] Based on the aforementioned process mechanism, a mesoscale process physics model is established to characterize the prepreg laying trajectory, the temperature and pressure field distribution inside the autoclave, and the resin flow and filling behavior.

[0030] Among them, the prepreg placement trajectory, the temperature and pressure field distribution inside the autoclave, and the resin flow and filling behavior refer to the process environment and material response physical phenomena described by the mesoscale process physics model; for example, the model simulates the movement path of the placement head, calculates the non-uniform temperature field inside the autoclave, and predicts the resin flow front. The mesoscale process physics model refers to the mathematical model used to simulate the physical environment of the process site, such as the prepreg placement trajectory, the temperature and pressure field distribution inside the autoclave, and the resin flow and filling behavior; for example, the computational fluid dynamics model is used to solve the three-dimensional temperature distribution inside the autoclave during the curing process.

[0031] The process physics data output from the mesoscale process physics model is used as boundary conditions and loads input into the microscale state evolution model. Based on the microstructure state data output from the microscale state evolution model, a macroscale product performance prediction model is established to predict the mechanical properties and defect distribution of the finished product.

[0032] Boundary conditions and loads refer to external constraints or excitations applied to the model boundaries in simulation calculations to define the external environment or interactions; for example, in a microscale model, the local temperature of the resin provided by the mesoscale model is used as the boundary condition, and the fluid pressure borne by the resin is used as the load. Microstructural state data refers to the quantitative results describing the microscopic characteristics of the material output by the microscale state evolution model; for example, the porosity, fiber volume fraction, and resin curing degree data at a certain location output by the model. Finished product mechanical properties and defect distribution refer to the overall properties and quality status of the final product predicted by the macroscale product performance prediction model; for example, the model predicts the tensile strength, bending stiffness, and the location and size of internal pores and delamination defects of the component. Macroscale product performance prediction models refer to mathematical models used to predict the mechanical properties and defect distribution of finished products based on the material's microstructure; for example, an analytical model that calculates the elastic modulus of laminates based on homogenization theory and inputting fiber orientation and porosity.

[0033] The multi-scale digital twin model is constructed by defining the data interface and calling relationship between the microscale state evolution model, the mesoscale process physics model, and the macroscale product performance prediction model.

[0034] Here, "data interface" refers to the format and protocol conventions followed for data exchange between models or modules of different scales; for example, specifying that when a mesoscale model transmits temperature data to a microscale model, a JSON format string containing coordinates and temperature values ​​should be used. "Calling relationship" refers to the logical sequence and control flow of startup, operation, data exchange, and collaboration among the sub-models in a multi-scale digital twin model; for example, specifying that in an iteration, the mesoscale model is run first, its output is used as a boundary condition to call the microscale model, and finally, the output of the microscale model is used to call the macroscale model.

[0035] Among the above-mentioned optional approaches, a microscale state evolution model, a mesoscale process physics field model, and a macroscale product performance prediction model are further constructed based on material properties and process mechanisms. By defining data interfaces and calling relationships between models, a multi-scale digital twin model is formed, which solves the problem that a single-scale model is difficult to characterize cross-scale physical mechanisms and improves the accuracy of the correlation mapping from material microstructure to macroscopic performance.

[0036] In one alternative embodiment, the digital twin modeling module 120 is specifically used for: The process parameters from the real-time manufacturing data are input into the mesoscale process physics model in the multi-scale digital twin model, driving the mesoscale process physics model to evolve in real time.

[0037] Real-time evolution refers to the digital twin model performing calculations and updating its state at a speed synchronized with the actual production process based on the input real-time manufacturing data; for example, the mesoscale model receives real temperature data once per second and updates the temperature field distribution inside the tank accordingly.

[0038] The process physics data output from the evolved mesoscale process physics model is used as real-time boundary conditions and loads input into the microscale state evolution model in the multi-scale digital twin model, driving the microscale state evolution model to evolve in real time.

[0039] Among them, process physics data refers to the quantitative results of the mesoscale process physics model output, which describes the physical environment of the process site; for example, the temperature, pressure and resin flow velocity at different spatial coordinate points inside the autoclave output by the model.

[0040] The microstructure state data output in real time from the microscale state evolution model is input into the macroscale product performance prediction model in the multiscale digital twin model, driving the macroscale product performance prediction model to perform real-time calculations.

[0041] Real-time calculation refers to the macroscopic product performance prediction model immediately performing calculations based on the latest microstructural state data and outputting the latest performance prediction results; for example, whenever updated micro porosity data is received, the macroscopic model immediately calculates the new component strength prediction value.

[0042] The evolution data of the mesoscale process physics field model, the evolution data of the microscale state evolution model, and the calculation results of the macroscale product performance prediction model are integrated in real time to generate the dynamic digital twin that maps the microstructure evolution and macroscopic performance formation process of the physical entity material.

[0043] Evolutionary data refers to the process data generated by the digital twin model during real-time evolution, reflecting changes in its internal state; for example, the temperature field data of the mesoscale model every second during the entire curing process, and the resin curing degree change curve recorded by the microscale model over time. Calculation results refer to the final performance prediction value output by the macroscale product performance prediction model after performing real-time calculations; for example, the ultimate tensile strength of the component calculated by the macroscale model in this instance is 1820 MPa.

[0044] In the above-mentioned optional methods, real-time manufacturing data is further used to sequentially drive the mesoscale process physical field model, the microscale state evolution model, and the macroscale product performance prediction model to perform real-time evolution and calculation. Simultaneously, multi-scale evolution data is integrated to generate a dynamic digital twin, which solves the problem of the digital twin and the physical entity being out of sync, and improves the real-time performance and accuracy of virtual-real mapping.

[0045] In an alternative embodiment, the simulation analysis and decision-making module 130 is specifically used for: Based on the dynamic digital twin, the process of carbon fiber composite materials is simulated and the performance evolution of the finished product is simulated to obtain the simulation results that include multiple scale states.

[0046] Among them, multi-scale state refers to the comprehensive state information of a dynamic digital twin at a certain moment, covering micro, meso and macro scales; for example, at time t, the dynamic digital twin simultaneously contains the micro-curing degree of the resin, the meso-temperature field distribution inside the tank and the macro-predicted deflection of the component.

[0047] Extract a set of key performance indicators corresponding to the predefined process specifications from the simulation results.

[0048] The set of key performance indicators refers to a set of core performance parameters selected from simulation results for comparison with process specifications; for example, for a certain component, the set of key performance indicators includes tensile strength, bending stiffness, and porosity.

[0049] Each index value in the set of key performance indicators is compared in real time with the corresponding target value in the predefined process specification, and the comprehensive process deviation data of the set of key performance indicators is calculated based on the comparison results.

[0050] Here, the target value refers to the specific numerical standard or range set for each key performance indicator in the predefined process specification; for example, the specification stipulates that the target value for tensile strength is ≥1750 MPa, and the target value for porosity is ≤1%. The comparison result refers to the qualitative or quantitative conclusion drawn by comparing the value of each indicator in the set of key performance indicators with its corresponding target value; for example, the comparison result shows that the tensile strength meets the standard, but the porosity exceeds the standard. The comprehensive process deviation data refers to the scalar value or vector obtained by comprehensively and quantitatively evaluating the deviations of multiple key performance indicators using a specific algorithm; for example, using the formula to calculate the positive deviation of strength and the negative deviation of porosity, the value D(t) = 0.15 is obtained to measure the overall degree of deviation.

[0051] Based on the comprehensive process deviation data and combined with the influence weights of process parameters on the key performance indicators in the mesoscale process physics model, the process parameter adjustment instructions are generated.

[0052] The influence weight refers to the quantitative proportion of the influence of different process parameters on key performance indicators during the decision-making process; for example, the influence weight of "curing temperature" on "tensile strength" is 0.6, and the influence weight of "pressure" is 0.3.

[0053] Among the above-mentioned optional methods, further process simulation and performance simulation are performed based on dynamic digital twins, key performance indicators are extracted and compared with predefined process specifications in real time to calculate comprehensive process deviations, and adjustment instructions are generated by combining the influence weights of process parameters. This solves the problems of lagging process deviation identification and insufficient control basis, and improves the scientificity and timeliness of process decision-making.

[0054] In one alternative approach, the formula for calculating the comprehensive process deviation data is: in, This represents the comprehensive process deviation data at time t. The total number of process parameters to be controlled. This represents the deviation between the actual value and the target setpoint of the j-th process parameter at time t. This represents the reference value for the j-th process parameter. This represents the influence coefficient of the deviation of the j-th process parameter. The total number of indicators in the set of key performance indicators. This represents the value of the i-th key performance indicator in the simulation results at time t. This represents the target value of the i-th critical performance indicator in the predefined process specification. This represents the weighting coefficient for the independent deviation of the i-th key performance indicator. The interaction coefficient represents the coupling deviation between the i-th and k-th key performance indicators. This is the scale coupling adjustment factor.

[0055] It should be noted that the above formula quantifies the deviations in process parameter control and the deviations in final product performance indicators separately, and then weights and merges them. The performance indicator deviation component specifically introduces higher-order penalty terms for independent deviations and interaction terms for coupled deviations between different performance indicators to comprehensively characterize the complex deviation states of the manufacturing process. The function of this formula is to provide the simulation analysis and decision-making module with a comprehensive quantitative evaluation basis that can simultaneously reflect process stability and result conformity, and sensitively capture the correlational deviations of multiple performance indicators, thereby supporting accurate process adjustment decisions.

[0056] Among the above-mentioned optional methods, a comprehensive process deviation calculation formula that includes process parameter deviation items and key performance index deviation items is further adopted. Independent deviation weighting coefficient, coupled deviation interaction coefficient and scale coupling adjustment factor are introduced to solve the problem of difficulty in quantifying multi-parameter and multi-index coupled deviation, and improve the comprehensiveness and accuracy of process deviation assessment.

[0057] In one alternative embodiment, the precise execution control module 140 is specifically used for: The process parameter adjustment command is parsed to obtain the target process parameter to be adjusted and the corresponding target set value.

[0058] The target process parameter refers to the specific process parameter that needs to be adjusted, as explicitly stated in the process parameter adjustment instruction; for example, the target process parameter in an instruction might be "heating power of the third zone of the autoclave". The target setpoint refers to the new value specified by the process parameter adjustment instruction for the target process parameter; for example, the instruction requires adjusting "heating power of the third zone of the autoclave" to "85% of the rated power".

[0059] The target process parameters are mapped to the corresponding physical entity actuators.

[0060] Based on the target setpoint, the current state feedback of the physical entity actuator, and the pre-established dynamic response characteristics of the actuator, a control signal is calculated and generated to drive the actuator to adjust the process parameters.

[0061] Among them, the dynamic response characteristics of the actuator refer to the dynamic relationship model of the output state of the physical entity actuator changing with time after receiving the control signal; for example, the inertial time constant model describing the lag of the actual temperature rise in the tank after the heater receives a new power signal compared with the set value.

[0062] The control signal is sent to the corresponding physical entity actuator.

[0063] In the above-mentioned optional methods, the process parameter adjustment command is further parsed to obtain the target parameter and set value, mapped to the corresponding actuator, and the control signal is calculated based on the dynamic response characteristics. This solves the problem of improper command conversion and actuator adaptation, and improves the accuracy and response speed of process parameter adjustment.

[0064] In one alternative approach, the formula for calculating the control signal is: in, Indicates in The control signal value sent to the nth physical entity actuator at any given time. This represents the deviation between the target setpoint and the current actual value of the nth target process parameter at time t. This represents the basic response gain coefficient of the nth actuator with respect to its own parameter deviation. This represents the cross-domain coupling compensation coefficient between the action of the m-th actuator and the control signal of the n-th actuator. This represents the coupling effect of the state of the m-th actuator at time t on the n-th control signal. This represents the differential response coefficient of the nth actuator to changes in the control command. Indicates the deviation amount The rate of change per unit time.

[0065] It should be noted that the above formula establishes a look-ahead coupling control law applicable to multi-actuator collaborative operation scenarios. This control law not only includes a basic proportional adjustment term for its own parameter deviations, but also innovatively introduces a cross-domain coupling compensation term reflecting the influence of other actuator actions, as well as a differential response term based on the command change trend. The function of this formula is to calculate an optimized control signal that can compensate for inter-mechanism interference in advance and suppress process inertia and disturbances based on process parameter adjustment commands and real-time feedback, thereby driving the physical actuators to achieve precise, coordinated, and stable process parameter adjustments.

[0066] Among the above-mentioned optional methods, a control signal calculation formula that includes basic response gain, cross-domain coupling compensation and differential response term is further adopted. The coupling influence between actuators and the rate of change of deviation are introduced, which solves the problem of insufficient accuracy of multi-actuator collaborative control and improves the robustness and adaptability of control signal generation.

[0067] In an alternative embodiment, the quality traceability and visualization module 150 is specifically used for: Align and associate the real-time manufacturing data, the evolution data of the dynamic digital twin, and the process parameter adjustment instructions according to the production time sequence.

[0068] Based on the aligned data, a structured quality data tree is established with the production batch as the root node, multiple processes as intermediate nodes, and the output of the dynamic digital twin in the macro-scale product performance prediction model as the leaf nodes.

[0069] Aligned data refers to real-time manufacturing data, evolution data, and instruction data that have undergone preprocessing and timestamp alignment to ensure complete synchronization of the time series. For example, the environmental humidity at time t, the curing degree predicted by the digital twin at that time, and the latest adjustment instruction issued before time t all have identical time stamps. A production batch refers to a unique identifier for a group of products produced using the same batch of raw materials, within the same time period, and according to the same process specifications. For example, a structural component production task numbered "Batch-20251027-01". A structured quality data tree refers to a data model that organizes quality traceability data in a tree-like hierarchical structure. For example, with the production batch as the root node, its child nodes are each process node, and each process node carries all time-series data and adjustment instruction records for that process.

[0070] The raw material data, process data, process intervention data, and final product performance data stored in the structured quality data tree are associated and encapsulated through a unified traceability identifier.

[0071] Among them, process data refers to data collected and generated in each production process that reflects the status of the processing; for example, the positioning accuracy log of the laying head in the laying process, and the temperature-pressure time series curve of the autoclave in the curing process. Process intervention data refers to instructions and operation records initiated automatically or manually by the system to change the original process; for example, the "reduce heating rate" instruction issued by the simulation analysis and decision module at time t and its execution result feedback. Final product performance data refers to the performance index values ​​of the finished product obtained through actual measurement or prediction based on the final state of the digital twin; for example, the porosity distribution map obtained by ultrasonic testing of a component, or the bending stiffness of the component predicted by the digital twin. Traceability identification refers to a code assigned to each production batch, semi-finished product, or finished product to uniquely identify and associate all its related data in the traceability chain; for example, a QR code affixed to each component, whose code is associated with the production batch number to which the component belongs and all related data in the structured quality data tree.

[0072] Based on the traceability identifier, retrieve and reconstruct associated data to generate and visualize the full lifecycle quality traceability chain covering raw materials to finished products.

[0073] Among the above-mentioned optional methods, multi-source data are further aligned and associated according to the production time sequence to establish a structured quality data tree with production batch as the root node, process as the intermediate node, and performance output as the leaf node. The entire life cycle quality traceability chain is encapsulated and visualized through unified traceability identification, which solves the problems of scattered quality data and broken traceability paths, and improves the transparency of quality information and traceability efficiency.

[0074] In an alternative embodiment, the quality traceability and visualization module 150 is further configured to: Based on the structured quality data tree, performance consistency analysis is performed on multiple finished products from the same raw material batch, and the analysis results are associated, labeled, and visualized with the corresponding process data and process intervention data.

[0075] Among them, performance consistency analysis refers to the process of statistically analyzing the performance data of multiple finished products in the same production batch to evaluate the uniformity and stability of the performance of the batch of products; for example, analyzing the tensile strength of all components in a batch and calculating the average, standard deviation and range to determine production stability.

[0076] Among the above-mentioned optional methods, performance consistency analysis of finished products in the same batch is further performed based on structured quality data trees, and the analysis results are associated with process data and process intervention data and visualized. This solves the problem of difficulty in attributing quality fluctuations within a batch and improves the accuracy of root cause location of quality anomalies and the targeted nature of improvement.

[0077] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0078] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and do not imply a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0079] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A digital twin control system for the production process of carbon fiber composite materials, characterized in that, The system includes: The data acquisition module is used to generate real-time manufacturing data by collecting process parameters and environmental parameters of multiple processes in the production of carbon fiber composite materials in real time. The digital twin modeling module is used to construct a multi-scale digital twin model based on material properties and process mechanisms, and to use the real-time manufacturing data to drive the multi-scale digital twin model to perform dynamic digital mapping, generating a dynamic digital twin that maps the evolution of the microstructure and the formation of the macroscopic properties of the physical material. The simulation analysis and decision-making module is used to perform process simulation and performance evolution simulation based on the dynamic digital twin, compare and analyze the simulation results with the predefined process specifications in real time to obtain process deviation data, and generate process parameter adjustment instructions based on the process deviation data. A precision execution control module is used to convert the process parameter adjustment command into a control signal for a physical entity actuator, and the control signal is used to drive the actuator to adjust the process parameter. The quality traceability and visualization module is used to integrate the real-time manufacturing data, the evolution data of the dynamic digital twin, and the process parameter adjustment instructions to generate and display a quality traceability chain covering the entire life cycle from raw materials to finished products.

2. The digital twin control system for the carbon fiber composite material production process according to claim 1, characterized in that, The data acquisition module is specifically used for: The process parameters of the multiple processes in the production process of carbon fiber composite materials, including prepreg laying, autoclave curing and resin transfer molding, are collected in real time, and the environmental parameters of the multiple processes are collected in real time simultaneously. The process parameters and environmental parameters are preprocessed and timestamped, and the processed process parameters and environmental parameters are then structurally integrated to generate the real-time manufacturing data.

3. The digital twin control system for the carbon fiber composite material production process according to claim 1, characterized in that, The digital twin modeling module is specifically used for: Based on the aforementioned material properties of carbon fiber composites, a microscale state evolution model is established to characterize the internal fiber arrangement, pore formation, and resin curing degree evolution of the material. Based on the aforementioned process mechanism, a mesoscale process physics model is established to characterize the prepreg laying trajectory, the temperature and pressure field distribution inside the autoclave, and the resin flow and filling behavior. The process physics data output by the mesoscale process physics model is used as boundary conditions and loads input into the microscale state evolution model. Based on the microstructure state data output by the microscale state evolution model, a macroscale product performance prediction model for predicting the mechanical properties and defect distribution of the finished product is established. The multi-scale digital twin model is constructed by defining the data interface and calling relationship between the microscale state evolution model, the mesoscale process physics model, and the macroscale product performance prediction model.

4. The digital twin control system for the carbon fiber composite material production process according to claim 3, characterized in that, The digital twin modeling module is specifically used for: The process parameters in the real-time manufacturing data are input into the mesoscale process physics model in the multi-scale digital twin model to drive the mesoscale process physics model to evolve in real time. The process physics data output by the evolved mesoscale process physics model is used as real-time boundary conditions and loads input into the microscale state evolution model in the multi-scale digital twin model to drive the microscale state evolution model to evolve in real time. The microstructure state data output in real time from the microscale state evolution model is input into the macroscale product performance prediction model in the multiscale digital twin model to drive the macroscale product performance prediction model to perform real-time calculations. The evolution data of the mesoscale process physics field model, the evolution data of the microscale state evolution model, and the calculation results of the macroscale product performance prediction model are integrated in real time to generate the dynamic digital twin that maps the microstructure evolution and macroscopic performance formation process of the physical entity material.

5. The digital twin control system for the carbon fiber composite material production process according to claim 4, characterized in that, The simulation analysis and decision-making module is specifically used for: Based on the dynamic digital twin, the process of carbon fiber composite materials is simulated and the performance evolution of the finished product is simulated to obtain the simulation results that include multiple scale states. Extract a set of key performance indicators corresponding to the predefined process specifications from the simulation results; The value of each indicator in the set of key performance indicators is compared with the corresponding target value in the predefined process specification in real time, and the comprehensive process deviation data of the set of key performance indicators is calculated based on the comparison results. Based on the comprehensive process deviation data and combined with the influence weights of process parameters on the key performance indicators in the mesoscale process physics model, the process parameter adjustment instructions are generated.

6. The digital twin control system for the carbon fiber composite material production process according to claim 5, characterized in that, The formula for calculating the comprehensive process deviation data is as follows: in, This represents the comprehensive process deviation data at time t. The total number of process parameters to be controlled. This represents the deviation between the actual value and the target setpoint of the j-th process parameter at time t. This represents the reference value for the j-th process parameter. This represents the influence coefficient of the deviation of the j-th process parameter. The total number of indicators in the set of key performance indicators. This represents the value of the i-th key performance indicator in the simulation results at time t. This represents the target value of the i-th critical performance indicator in the predefined process specification. This represents the weighting coefficient for the independent deviation of the i-th key performance indicator. The interaction coefficient represents the coupling deviation between the i-th and k-th key performance indicators. This is the scale coupling adjustment factor.

7. The digital twin control system for the carbon fiber composite material production process according to claim 6, characterized in that, The precise execution control module is specifically used for: The process parameter adjustment command is parsed to obtain the target process parameter to be adjusted and the corresponding target set value; Map the target process parameters to the corresponding physical entity actuator; Based on the target setpoint, the current state feedback of the physical entity actuator, and the pre-established dynamic response characteristics of the actuator, a control signal is calculated and generated to drive the actuator to adjust the process parameters. The control signal is sent to the corresponding physical entity actuator.

8. The digital twin control system for the carbon fiber composite material production process according to claim 7, characterized in that, The formula for calculating the control signal is: in, Indicates in The control signal value sent to the nth physical entity actuator at any given time. This represents the deviation between the target setpoint and the current actual value of the nth target process parameter at time t. This represents the basic response gain coefficient of the nth actuator with respect to its own parameter deviation. This represents the cross-domain coupling compensation coefficient between the action of the m-th actuator and the control signal of the n-th actuator. This represents the coupling effect of the state of the m-th actuator at time t on the n-th control signal. This represents the differential response coefficient of the nth actuator to changes in the control command. Indicates the deviation amount The rate of change per unit time.

9. The digital twin control system for the carbon fiber composite material production process according to claim 8, characterized in that, The quality traceability and visualization module is specifically used for: Align and associate the real-time manufacturing data, the evolution data of the dynamic digital twin, and the process parameter adjustment instructions according to the production time sequence; Based on the aligned data, a structured quality data tree is established with the production batch as the root node, multiple processes as intermediate nodes, and the output of the dynamic digital twin in the macro-scale product performance prediction model as the leaf nodes. The raw material data, process data, process intervention data and final product performance data stored in the structured quality data tree are associated and encapsulated through a unified traceability identifier; Based on the traceability identifier, retrieve and reconstruct associated data to generate and visualize the full lifecycle quality traceability chain covering raw materials to finished products.

10. The digital twin control system for the carbon fiber composite material production process according to claim 9, characterized in that, The quality traceability and visualization module is also used for: Based on the structured quality data tree, performance consistency analysis is performed on multiple finished products from the same raw material batch, and the analysis results are associated, labeled, and visualized with the corresponding process data and process intervention data.

Images