Intelligent control method and system for carbon fiber composite material forming process

Abstract

This application discloses an intelligent control method and system for the molding process of carbon fiber composite materials, relating to the field of composite material molding control technology. The method includes the following steps: constructing a digital model of the component for process simulation; acquiring and inferring the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component, and updating the internal state information to the digital model; predicting the evolution of the internal state of the component within a preset time period based on the updated digital model, and identifying potential process risks of the component; determining control commands for adjusting the process parameters of curing based on the potential process risks; performing another simulation in the updated digital model with the control commands as input to verify the control commands; and sending the verified control commands to the autoclave for execution. This application can improve the molding quality, uniformity, and long-term reliability of composite material components with complex shapes.

Description

Technical Field

[0001] This application relates to the field of composite material molding control technology, and in particular to an intelligent control method and system for the molding process of carbon fiber composite materials. Background Technology

[0002] In advanced manufacturing, especially for the production of high-performance materials such as carbon fiber composites, autoclave curing is a crucial step determining the quality and reliability of the final product. Existing intelligent control systems face significant technical challenges when dealing with components with complex shapes and significant thickness variations. Due to the complex geometry of the components, the thickness distribution is uneven, resulting in significant differences in heat transfer rates across different regions and creating large temperature gradients within the components. Thin-walled areas tend to heat up too quickly, while the temperature rise in the thick-walled core lags behind, causing an uncoordinated resin curing process.

[0003] Furthermore, this uneven temperature field directly affects the viscosity evolution of the resin during the curing process, leading to impeded pressure transmission. The limitation of existing control systems lies in their reliance on limited external sensor feedback, lacking a holistic understanding and predictive capability regarding the temperature field, resin state (such as viscosity), and pressure transmission path distribution within the entire component. When the system continues heating to compensate for the delayed curing of the thick-walled region, the already cured thin-walled region is prone to thermal damage; conversely, if heating is prematurely terminated to protect the thin-walled region, the thick-walled region will not cure sufficiently. Simultaneously, premature gelation of the resin in the thin-walled region hinders pressure transmission to the thick-walled region, preventing the internal fiber layer from being adequately compacted. This ultimately results in hidden defects such as pores and delamination within the component, seriously threatening product reliability and proving difficult to detect through conventional testing methods. Summary of the Invention

[0004] This application proposes an intelligent control method and system for the molding process of carbon fiber composite materials, aiming to solve the technical problems of hidden defects such as overheating damage, incomplete curing, porosity, and delamination caused by uneven thickness of components due to complex component geometry during the autoclave curing process of carbon fiber composite materials.

[0005] In a first aspect, this application provides a method for curing and molding components with complex geometries in an autoclave, the method comprising the following steps: Acquire and construct a digital model of the component for process simulation based on the component's geometric structure and material properties; Based on the internal environmental parameters of the autoclave and the surface state parameters of the component, the internal state information of the component is inferred, including the internal temperature distribution, resin state and degree of curing, and the internal state information is updated to the digital model. Based on the updated digital model, the internal state evolution of the component is predicted within a future preset time period to obtain a first simulation result. Based on the first simulation result, potential process risks of the component are identified. Based on the potential process risks, control instructions for adjusting the process parameters of the curing and molding process are determined. In the updated digital model, the control command is used as input to perform another simulation to obtain a second simulation result. The control command is then verified based on the second simulation result, and the verified control command is sent to the autoclave for execution.

[0006] In some embodiments of this application, the step of acquiring and constructing a digital model of the component for process simulation based on the component's geometric structure information and material property information includes: The three-dimensional geometric data of the component is obtained using computer-aided design software, which serves as the geometric structure information of the component. Based on a pre-set material database, the thermophysical parameters and resin curing kinetic parameters of the carbon fiber prepreg used in the component are obtained as material property information of the component; wherein, the thermophysical parameters include specific heat capacity and thermal conductivity, and the resin curing kinetic parameters include Arrhenius equation parameters used to describe the relationship between curing reaction rate and temperature; During the curing process, a preset disturbance is applied to the autoclave, and the autoclave's response to the disturbance is obtained to correct the material property information, thereby obtaining the corrected material property information. Based on the geometric structure information and the corrected material property information, a digital model is constructed to simulate the internal heat conduction and resin curing reaction process of the component.

[0007] In some embodiments of this application, the step of applying a preset disturbance to the autoclave during the curing process and obtaining the autoclave's response to the disturbance to correct the material property information, thereby obtaining the corrected material property information, includes: During the curing process, a temperature or pressure pulse of preset amplitude and duration is applied to the autoclave; Monitor the dynamic response of the autoclave and components to the temperature or pressure pulse; The dynamic response is matched with a pre-stored theoretical response feature library based on different combinations of material parameters to identify the actual values ​​of the resin curing kinetic parameters of the current batch of materials. By using the actual values ​​of the resin curing kinetic parameters of the current batch of materials, the resin curing kinetic parameters in the material property information are corrected to obtain the corrected material property information.

[0008] In some embodiments of this application, the step of obtaining and inferring the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component, wherein the internal state information includes the internal temperature distribution, resin state, and degree of curing, and updating the internal state information to the digital model includes: Based on a first type of sensor installed inside the autoclave, the internal temperature and pressure of the autoclave are acquired as internal environmental parameters. Based on a second type of sensor, which is arranged at multiple preset points on the surface of the mold used for the component, the temperature at the multiple preset points is obtained as a surface state parameter of the component. Based on the internal environment parameters and the surface state parameters, the heat conduction equation and the resin curing reaction kinetic equation, which include the resin curing exothermic term, are solved by a preset physical mechanism inference engine. The first round of inference is performed to infer the internal state information of the component and the temperature prediction values ​​of multiple preset points in real time. The internal state information of the component includes the internal temperature distribution, resin state and degree of curing of the component. The measured temperature values ​​of multiple preset points are obtained. Based on the temperature deviation between the predicted temperature value and the measured temperature value, the digital model is calibrated, and a second round of inference is performed to obtain the calibrated internal state information. Based on the calibrated internal state information, the calibrated digital model is adjusted and updated to obtain the updated digital model.

[0009] In some embodiments of this application, the steps of obtaining measured temperature values ​​at multiple preset points, calibrating the digital model based on the temperature deviation between the predicted temperature value and the measured temperature value, and performing a second round of inference to obtain calibrated internal state information include: Continuously monitor the temperature deviation between the measured temperature value at the preset point and the predicted temperature value at the corresponding preset point calculated by the digital model; When the temperature deviation is detected to continuously exceed the preset threshold at any preset point, the material property parameters of the local area in the digital model associated with the preset point where the temperature deviation occurs are adjusted to obtain the adjusted digital model. Based on the adjusted digital model, the physical mechanism inference engine is rerun to perform the second round of inference and obtain the calibrated internal state information.

[0010] In some embodiments of this application, the step of predicting the internal state evolution of the component within a predetermined time period based on an updated digital model to obtain a first simulation result, and identifying potential process risks of the component based on the first simulation result, includes: In the updated digital model, the evolution of the temperature field, resin viscosity field, and degree of curing field inside the component is predicted over a future preset time period at a speed faster than the real-time curing and molding speed of the component, to obtain a first simulation result; the first simulation result includes predicted data of temperature, viscosity, and degree of curing distribution information at each time point within the future preset time period. The first simulation result is compared with a preset quality threshold to identify potential process risks of the component; the potential process risks include overheating damage risk, incomplete curing risk, and pressure transmission obstruction risk.

[0011] In some embodiments of this application, the step of comparing the first simulation result with a preset quality threshold to identify potential process risks of the component includes: When the first simulation results show that the predicted temperature of a thin-walled region in the component with a thickness less than a preset first thickness threshold exceeds a preset resin thermal damage temperature threshold, it is identified as a risk of overheating damage. When the first simulation result shows that the predicted degree of curing of the thick-walled region in the component with a thickness greater than a preset second thickness threshold is lower than a preset target degree of curing threshold, it is identified as a risk of incomplete curing; the second thickness threshold is greater than the first thickness threshold. When the first simulation results show that the predicted resin viscosity of the thin-walled region of the component rises above the preset gel viscosity threshold, the predicted resin viscosity of the thick-walled region is lower than the gel viscosity threshold, and the predicted curing degree of the thick-walled region is lower than the preset compaction window curing degree threshold, then it is identified as a risk of pressure transmission obstruction.

[0012] In some embodiments of this application, the step of determining control instructions for adjusting the process parameters of the curing molding based on the potential process risks includes: In response to the identified risk of overheating damage, control commands are determined to reduce the heating power of the autoclave or set the temperature. In response to the identified risk of incomplete curing, control commands are determined to increase the heating power of the autoclave, set the temperature, or extend the holding time of the corresponding temperature stage. When a risk of pressure transmission obstruction is identified, a control command is determined to adjust the pressure application rate or pressure value of the autoclave.

[0013] In some embodiments of this application, the step of performing a second simulation in the updated digital model with the control command as input to obtain a second simulation result, verifying the control command based on the second simulation result, and sending the verified control command to the autoclave for execution includes: Before the control command is sent to the autoclave for execution, a new prediction of the internal state evolution of the component is performed in the digital model with the control command as input, to obtain a second simulation result; Determine whether the second simulation result meets the preset verification conditions, the verification conditions including: the potential process risk is eliminated and no new process deviation is introduced; If so, the control command is confirmed as verified and sent to the controller of the autoclave for execution; If not, based on the potential process risks reflected in the second simulation results, the parameters of the control command are iteratively adjusted using a preset optimization algorithm, the control command is redefined, and a new round of verification is performed.

[0014] Secondly, this application also provides an intelligent control system for the molding process of carbon fiber composite materials, applied to the curing and molding of parts with complex geometries in an autoclave, the system comprising: The model building module is used to acquire and construct a digital model of the component for process simulation based on the component's geometric structure information and material property information; The model update module is used to acquire and infer the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component. The internal state information includes the internal temperature distribution, resin state and degree of curing, and the internal state information is updated to the digital model. The risk identification module is used to predict the internal state evolution of the component within a preset time period in the future based on the updated digital model, obtain a first simulation result, and identify potential process risks of the component based on the first simulation result. The instruction determination module is used to determine control instructions for adjusting the process parameters of the curing molding based on the potential process risks. The instruction sending module is used to simulate again in the updated digital model with the control instruction as input to obtain a second simulation result, verify the control instruction based on the second simulation result, and send the verified control instruction to the autoclave for execution.

[0015] The technical solution according to the embodiments of this application has at least the following beneficial effects: By implementing forward-looking simulation prediction and active intervention, this application can avoid overheating damage in thin-walled areas, incomplete curing in thick-walled areas, and hidden defects such as porosity and delamination caused by the lag and localization of traditional control methods. This significantly improves the molding quality, uniformity, and long-term reliability of composite material parts with complex shapes.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0018] Figure 1 This is a flowchart illustrating an intelligent control method for the molding process of carbon fiber composite materials, provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the architecture of an intelligent control system for the molding process of carbon fiber composite materials, provided in an embodiment of this application. Detailed Implementation

[0020] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0021] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] like Figure 1As shown, this application discloses an intelligent control method for the molding process of carbon fiber composite materials, applied to the curing and molding of parts with complex geometries in an autoclave. The method includes the following steps: S110, acquire and construct a digital model of the component for process simulation based on the geometric structure information and material property information of the component; S120: Obtain and infer the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component. The internal state information includes the internal temperature distribution, resin state and degree of curing. Update the internal state information to the digital model. S130, based on the updated digital model, predict the internal state evolution of the component within a future preset time period to obtain a first simulation result, and based on the first simulation result, identify the potential process risks of the component; S140, Based on the potential process risks, determine control instructions for adjusting the process parameters of the curing and molding process; S150, In the updated digital model, the control command is used as input to perform another simulation to obtain a second simulation result. The control command is verified based on the second simulation result, and the verified control command is sent to the autoclave for execution.

[0023] The "digital model" refers to a mathematical description and computer simulation of the physical behavior (such as heat conduction, resin flow, and curing reaction) of carbon fiber composite components during the curing process. It reflects key information such as internal temperature distribution, resin viscosity changes, and the degree of curing. "Internal state information" is the core output of the digital model, including the internal temperature distribution, resin state (such as viscosity), and degree of curing. This information is crucial for assessing curing quality. "Potential process risks" refer to problems that may occur during the curing process that could lead to product defects, such as overheating damage, incomplete curing, or obstructed pressure transmission.

[0024] In practical implementation, the first step is to construct a digital model of the component for process simulation. For example, geometric information can be obtained by manually inputting the component's dimensions and shape parameters, or by scanning the component's 3D data. Material property information can be obtained from technical manuals provided by material suppliers or through laboratory testing; for example, the specific heat capacity, thermal conductivity, and resin curing kinetic parameters of carbon fiber prepreg can be obtained. Based on this information, a finite element model or finite difference model can be constructed as a digital model to simulate the heat conduction and resin curing reaction process inside the component.

[0025] Subsequently, the internal state information of the component is inferred, and the digital model is updated. For example, the internal environmental parameters of the autoclave can be acquired in real time using temperature and pressure sensors installed inside the autoclave. The surface state parameters of the component can be acquired by attaching thermocouples or infrared sensors to the component surface. Based on this real-time data, combined with the digital model, information such as the internal temperature distribution, resin viscosity, and degree of curing of the component is inferred. This inferred internal state information is then used to update the digital model, enabling it to more accurately reflect the current curing state.

[0026] Next, potential process risks to the component are identified. For example, in the updated digital model, simulations can be run at a faster rate than the actual curing process to predict the evolution of the temperature field, resin viscosity field, and degree of curing field inside the component over the next few minutes or hours. The first simulation results will include predicted data on the temperature, viscosity, and degree of curing distributions at various time points within a preset future time period. By comparing these predicted data with preset quality thresholds, potential process risks can be identified.

[0027] Control instructions are determined to adjust the process parameters for the curing and molding process. For example, if an overheating damage risk is identified, control instructions can be determined to reduce the autoclave heating power or set the temperature. If an incomplete curing risk is identified, control instructions can be determined to increase the autoclave heating power, set the temperature, or extend the holding time at the corresponding temperature stage. If a risk of pressure transmission obstruction is identified, control instructions can be determined to adjust the autoclave pressure application rate or pressure value. These control instructions aim to eliminate or mitigate identified potential process risks by adjusting the autoclave's operating parameters.

[0028] Finally, the control command is verified, and the verified control command is sent to the autoclave for execution. For example, before sending the control command to the autoclave, a new internal state evolution prediction is performed in the digital model with the control command as input, resulting in a second simulation result. Then, it is determined whether the second simulation result meets the preset verification conditions. If the verification conditions are met, the control command is determined to have passed verification and sent to the autoclave's controller for execution. If the verification conditions are not met, the control command is redefined and a new round of verification is performed until the control command passes verification.

[0029] The technical solution of this application, firstly, achieves accurate inference of the internal temperature distribution, resin state, and curing degree of components by constructing and updating a digital model in real time, overcoming the limitations of traditional methods that rely solely on surface sensor data. Secondly, based on the updated digital model, this method can predict the evolution of the internal state within a preset time period, thereby identifying process risks before they actually occur. This predictive capability enables the control system to proactively intervene rather than passively respond. Furthermore, this method introduces a "pre-verification" mechanism for control commands. Before sending control commands to the autoclave for execution, simulation verification is performed in the digital model to ensure that the commands can effectively eliminate potential risks and will not introduce new process deviations. This step effectively avoids process accidents and material waste caused by improper control commands, significantly reducing trial-and-error costs. Compared to existing control strategies based on experience or simple feedback, this method can more precisely manage heat and pressure transfer during the curing process, ensuring the curing uniformity and integrity of various regions within complex geometric components, thereby effectively solving the hidden defect problem existing in traditional methods and significantly improving the quality and reliability of carbon fiber composite components.

[0030] In summary, by achieving forward-looking simulation prediction and active intervention, this application can avoid overheating damage in thin-walled areas, incomplete curing in thick-walled areas, and hidden defects such as porosity and delamination caused by the lag and localization of traditional control methods. This significantly improves the molding quality, uniformity, and long-term reliability of composite material parts with complex shapes.

[0031] In some embodiments of this application, the step of acquiring and constructing a digital model of the component for process simulation based on the component's geometric structure information and material property information preferably includes: The three-dimensional geometric data of the component is obtained using computer-aided design software, which serves as the geometric structure information of the component. Based on a pre-set material database, the thermophysical parameters and resin curing kinetic parameters of the carbon fiber prepreg used in the component are obtained as material property information of the component; wherein, the thermophysical parameters include specific heat capacity and thermal conductivity, and the resin curing kinetic parameters include Arrhenius equation parameters used to describe the relationship between curing reaction rate and temperature; During the curing process, a preset disturbance is applied to the autoclave, and the autoclave's response to the disturbance is obtained to correct the material property information, thereby obtaining the corrected material property information. Based on the geometric structure information and the corrected material property information, a digital model is constructed to simulate the internal heat conduction and resin curing reaction process of the component.

[0032] The geometric information refers to the spatial geometric features of the component, such as its physical dimensions, shape, and internal structure. This information can be directly obtained through computer-aided design (CAD) software, for example, by extracting three-dimensional geometric data from CAD model files, ensuring the accuracy of the digital model in the spatial dimension.

[0033] The material property information refers to the physical and chemical behavior parameters of carbon fiber prepregs during the curing process. Among these, thermophysical parameters, such as specific heat capacity and thermal conductivity, describe the material's ability to absorb and transfer heat; resin curing kinetic parameters, such as Arrhenius equation parameters, describe how the rate of resin curing reaction changes with temperature. These parameters are typically obtained from a pre-designed material database containing standard property data for different types of carbon fiber prepregs.

[0034] During the curing process, a preset disturbance is applied to the autoclave. By monitoring the autoclave's response to these disturbances, real-time data on the actual behavior of the current batch of material can be obtained. Based on this actual response data, the initial material property information obtained from the material database can be corrected to obtain corrected material property information that better reflects the actual situation.

[0035] Therefore, based on geometric structure information and corrected material property information, a digital model can be constructed to simulate the internal heat conduction and resin curing reaction process of the component, which can more accurately reflect the internal state evolution of the component during the actual curing process.

[0036] Through the above technical solution, this application overcomes the problem of inaccurate material property information that may result from relying solely on a preset material database. By applying disturbance to the autoclave during the curing process and acquiring its response, real-time correction of material property information is achieved, particularly the identification of resin curing kinetic parameters. Consequently, the constructed digital model more realistically reflects the actual behavior of the current batch of materials, significantly improving the accuracy of process simulation. This high-precision digital model provides a more reliable foundation for subsequent internal state inference, risk identification, and control command determination, thereby enhancing the effectiveness and robustness of the entire intelligent control method, helping to reduce scrap rates and optimize production efficiency.

[0037] In a specific embodiment of this application, the step of applying a preset disturbance to the autoclave during the curing process and obtaining the autoclave's response to the disturbance to correct the material property information, thereby obtaining the corrected material property information, preferentially includes: During the curing process, a temperature or pressure pulse of preset amplitude and duration is applied to the autoclave; Monitor the dynamic response of the autoclave and components to the temperature or pressure pulse; The dynamic response is matched with a pre-stored theoretical response feature library based on different combinations of material parameters to identify the actual values ​​of the resin curing kinetic parameters of the current batch of materials. By using the actual values ​​of the resin curing kinetic parameters of the current batch of materials, the resin curing kinetic parameters in the material property information are corrected to obtain the corrected material property information.

[0038] Specifically, a temperature or pressure pulse of preset amplitude and duration is applied to the autoclave, intentionally introducing a preset disturbance through the autoclave's control system. This "preset amplitude and duration" aims to ensure that the applied disturbance sufficiently stimulates the material's response without adversely affecting the curing quality of the component. For example, a temperature pulse lasting several minutes with an amplitude of a few degrees Celsius, or a pressure pulse lasting several seconds with an amplitude of tens of kilopascals, can be applied.

[0039] Monitoring the dynamic response of the autoclave and components to the temperature or pressure pulses refers to using various sensors arranged inside the autoclave and on the surface of the components or mold to collect detailed data in real time on the changes of environmental parameters (such as temperature and pressure) inside the autoclave and surface parameters (such as surface temperature) of the components over time under the action of temperature or pressure pulses. These dynamic response data reflect the thermophysical and curing kinetic characteristics of the current batch of material under the actual curing environment.

[0040] The dynamic response is matched with a pre-stored theoretical response feature library based on different material parameter combinations to identify the actual values ​​of the resin curing kinetic parameters for the current batch of materials. Specifically, the theoretical response feature library is a pre-established database containing theoretical dynamic response curves or features corresponding to different combinations of resin curing kinetic parameters under the same perturbation conditions, obtained through numerical simulation or experimental calibration. By comparing and matching the actually monitored dynamic response with the theoretical responses in this feature library, the actual values ​​of the resin curing kinetic parameters for the current batch of carbon fiber prepreg can be inferred.

[0041] By using the actual values ​​of the resin curing kinetic parameters of the current batch of materials, the resin curing kinetic parameters in the material property information are corrected to obtain the corrected material property information. This means that once the actual resin curing kinetic parameters are identified, they are used to update the corresponding material property information in the digital model, thereby enabling the digital model to more accurately reflect the true material behavior of the component currently curing.

[0042] This application utilizes the unique response characteristics of materials to external stimuli. By comparing these characteristics with a pre-set theoretical response feature library, the actual values ​​of the resin curing kinetic parameters for the current batch of materials can be accurately identified. As a result, the material parameters in the digital model are updated in real time, ensuring a high degree of consistency between the model and the actual physical process, effectively overcoming the impact of batch-to-batch material differences on simulation accuracy.

[0043] In a further embodiment of this application, the step of acquiring and inferring the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component, wherein the internal state information includes the internal temperature distribution, resin state, and degree of curing, and updating the internal state information to the digital model preferably includes: Based on a first type of sensor installed inside the autoclave, the internal temperature and pressure of the autoclave are obtained as internal environmental parameters. Based on the second type of sensor, the second type of sensor is arranged at multiple preset points on the surface of the mold used for the component to obtain the temperature at multiple preset points as the surface state parameters of the component; Based on internal environmental parameters and surface state parameters, the heat conduction equation and resin curing reaction kinetic equation, which include the resin curing exothermic term, are solved by a preset physical mechanism inference engine. The first round of inference is performed to infer the internal state information of the component and the temperature prediction values ​​of multiple preset points in real time. The internal state information of the component includes the internal temperature distribution, resin state and degree of curing. The measured temperature values ​​at multiple preset points are obtained. Based on the temperature deviation between the predicted temperature value and the measured temperature value, the digital model is calibrated, and a second round of inference is performed to obtain the calibrated internal state information. Based on the calibrated internal state information, the calibrated digital model is adjusted and updated to obtain the updated digital model.

[0044] The first type of sensor can be understood as a device used to monitor the macroscopic environmental conditions inside the autoclave, such as thermocouples and pressure sensors. Its purpose is to acquire key environmental parameters such as temperature and pressure inside the autoclave, which are the external conditions driving the composite material curing process. In practical applications, this type of sensor can be deployed in different areas of the autoclave to obtain more comprehensive internal environmental data.

[0045] The second type of sensor refers to devices used to acquire local temperature information of the component surface, such as surface thermocouple arrays or infrared temperature sensors. Their purpose is to provide real-time temperature data at the interface between the component and the mold during the curing process. These preset points are typically selected in areas with complex component geometry, drastic changes in thermal conductivity, or those prone to process defects, to ensure accurate monitoring of critical areas.

[0046] The physics mechanism inference engine can be understood as a computational system based on physical laws and mathematical models. Its core lies in solving a set of partial differential equations describing heat conduction and resin curing reactions. Specifically, the heat conduction equation describes the temperature variation within the component over time and space, taking into account the exothermic effect during resin curing; the resin curing reaction kinetic equation describes the changes in resin curing degree and viscosity with temperature and time. Through numerical simulation, the engine uses known internal environmental parameters and component surface state parameters to infer the temperature distribution, resin state (such as viscosity), and degree of curing within the component—parameters that are difficult to measure directly. The first round of inference aims to provide a preliminary estimate of the internal state and a prediction of the surface temperature based on the current input data.

[0047] Temperature deviation refers to the difference between the predicted temperature value at a preset point obtained by the physical mechanism inference engine and the actual measured temperature value at the preset point obtained by the second type of sensor. This deviation reflects the inconsistency between the digital model and the actual physical process under the current operating conditions, which may be caused by factors such as slight deviations in material parameters, local inhomogeneities in the autoclave environment, or model simplification. Calibrating the digital model involves adjusting certain parameters in the digital model (such as the thermophysical parameters of the material or curing kinetic parameters) based on this temperature deviation to make the model's predictions closer to the actual measured values. The second round of inference involves running the physical mechanism inference engine again after the digital model has been calibrated to obtain more accurate and reliable calibrated internal state information.

[0048] Adjusting and updating the calibrated digital model based on calibrated internal state information means feeding back the calibrated, more accurate internal state information to the digital model, enabling it to reflect the current, actual solidified state. This allows the digital model to continuously synchronize with the actual process, ensuring its accuracy in subsequent predictions.

[0049] This solution effectively compensates for the discrepancy between the model and the actual process by integrating internal environmental parameters of the autoclave and measured temperature data at multiple points on the component surface, combined with a physical mechanism inference engine for two rounds of inference and model calibration. As a result, the obtained information on internal temperature distribution, resin state, and degree of curing more accurately reflects the actual curing process of the component, providing a solid and reliable data foundation for subsequent process risk identification and control command determination. This high-precision real-time internal condition monitoring can effectively reduce the risk of process defects such as overheating damage, incomplete curing, or pressure transmission obstruction caused by inaccurate information, thereby improving the molding quality and production efficiency of complex geometric carbon fiber composite components.

[0050] The following is a specific example to illustrate this.

[0051] In the case of autoclaving a carbon fiber composite component with complex curvature and thickness variations, the system first installs multiple thermocouples and pressure sensors at different locations inside the autoclave as first-type sensors to monitor the internal temperature and pressure in real time. Simultaneously, multiple surface thermocouples are positioned on the surface of the mold used for the component, particularly in thin-walled, thick-walled, and areas with significant curvature changes, as second-type sensors to acquire the real-time temperature at these critical points. For example, if the engine predicts a surface temperature of 150°C for a thin-walled area, but the second-type sensor measures 155°C, there is a 5°C temperature discrepancy. Based on this discrepancy, the system fine-tunes and calibrates the material thermal conductivity or resin exothermic parameters related to that area in the digital model. After calibration, the engine immediately performs a second round of inference to obtain more accurate, calibrated information about the component's internal state. For example, after calibration, the predicted internal temperature of the thin-walled area might be adjusted from 160°C to 164°C, more accurately reflecting the actual situation. Based on this calibrated internal state information, the digital model can be updated in real time, ensuring that subsequent process risk assessments and control command generation can be based on the component state as close as possible to reality.

[0052] In a further embodiment of this application, the steps of obtaining measured temperature values ​​at multiple preset points, calibrating the digital model based on the temperature deviation between the predicted temperature value and the measured temperature value, and performing a second round of inference to obtain calibrated internal state information preferably include: Continuously monitor the temperature deviation between the measured temperature value at the preset point and the predicted temperature value at the corresponding preset point calculated by the digital model; When the temperature deviation is detected to continuously exceed the preset threshold at any preset point, the material property parameters of the local area in the digital model associated with the preset point where the temperature deviation occurs are adjusted to obtain the adjusted digital model. Based on the adjusted digital model, the physical mechanism inference engine is rerun to perform the second round of inference and obtain the calibrated internal state information.

[0053] Continuous monitoring refers to the system continuously acquiring and comparing the measured temperature values ​​at each preset point with the predicted temperature values ​​calculated by the digital model based on the current process conditions, either at a preset sampling frequency or in real-time. The purpose is to promptly detect and track discrepancies between predictions and actual values. Specifically, when a temperature deviation at any preset point consistently exceeds a preset threshold, it indicates a potential systematic error in the digital model for that local area, rather than an occasional fluctuation. The preset threshold can be set based on process requirements and experience; for example, if the temperature deviation exceeds ±X degrees Celsius for N consecutive sampling cycles, it is considered to have consistently exceeded the preset threshold. In practical applications, adjusting the material property parameters of the local area associated with the preset point exhibiting the temperature deviation in the digital model involves using reverse engineering or optimization algorithms to deduce the actual or corrected values ​​of the material property parameters (such as thermal conductivity, specific heat capacity, resin curing kinetic parameters, etc.) within that local area based on the continuous temperature deviation. The aim is to enable the digital model to better reflect the true physical characteristics of that local area, thereby improving the model's local accuracy. Furthermore, based on the adjusted digital model, the physical mechanism inference engine is rerun to perform a second round of inference. The purpose is to use the digital model after local material property correction to infer the internal state information again, so as to obtain more accurate and reliable information such as the internal temperature distribution, resin state and curing degree of the component.

[0054] This application's solution effectively addresses the problem that traditional calibration methods may not adequately handle local or persistent model inaccuracies by introducing continuous monitoring of temperature deviations and a mechanism for adjusting local material property parameters based on these deviations. When a persistent deviation occurs between the predictions and actual measurements in a specific local area of ​​the digital model, it typically indicates a discrepancy between the material property parameters set in the digital model and the actual situation. By identifying this persistent deviation and specifically adjusting the material property parameters of that local area in the digital model—for example, if a local area is continuously overheated, it may mean that the actual thermal conductivity of that area is lower than the model's set value, or the heat release during curing is higher than the model's set value—the system will adjust these parameters accordingly. It is precisely this refined adjustment of local parameters that enables the digital model to more accurately reflect the true physical behavior of components under complex geometries and actual process conditions, thereby significantly improving the accuracy and reliability of internal state inference.

[0055] It is worth mentioning that the scheme of predicting the internal state evolution of a component within a predetermined time period based on an updated digital model and identifying potential process risks based on the first simulation results. Specifically, this step preferably includes: In the updated digital model, the evolution of the temperature field, resin viscosity field, and degree of curing field inside the component is predicted over a future preset time period at a speed faster than the real-time curing and molding of the component, to obtain the first simulation result; the first simulation result includes predicted data of temperature, viscosity, and degree of curing distribution information at each time point within the future preset time period. The first simulation result is compared with a preset quality threshold to identify potential process risks of the component; the potential process risks include overheating damage risk, incomplete curing risk, and pressure transmission obstruction risk.

[0056] "Predicting at a speed faster than the real-time curing and molding of the component" means that the digital model can run in accelerated mode, simulating the evolution trend of the component's internal state before the actual physical curing process occurs. This accelerated simulation capability enables the system to be forward-looking, buying time for subsequent control decisions. Among them, the "temperature field, resin viscosity field, and degree of curing field" are the most critical internal state parameters describing the curing process of carbon fiber composites. The temperature field reflects the heat distribution inside the component; the resin viscosity field describes the changes in resin flowability and compaction ability; and the degree of curing field indicates the progress of the resin crosslinking reaction. The predicted data of these fields provides a comprehensive view of the component's internal state. The "future preset time period" refers to a configurable time window, such as 5 minutes, 10 minutes, or longer, within which the system makes predictions to identify potential problems in advance. The "first simulation results" not only include the predicted data of these fields but also specify detailed distribution information of these temperatures, viscosities, and degrees of curing at various time points within the future preset time period.

[0057] "Preset quality thresholds" refer to standards set based on material properties, component design requirements, and process experience to determine whether a component's quality is up to standard or whether there are any risks. These thresholds can be upper temperature limits, lower limits for degree of curing, viscosity ranges, etc. "Potential process risks" specifically include "overheating damage risk," "incomplete curing risk," and "pressure transmission obstruction risk." Overheating damage risk typically occurs when the heating rate is too rapid or the holding temperature is too high, potentially leading to resin degradation. Incomplete curing risk may result from insufficient heating or holding time, causing the component's mechanical properties to fail to meet standards. Pressure transmission obstruction risk may occur when the resin viscosity rises too early, preventing effective compaction of the pores within the fiber prepreg and affecting the component's density.

[0058] This application leverages the rapid predictive capabilities of digital models to transform process control from a reactive response to a proactive prevention approach. Specifically, by identifying risks such as overheating damage, incomplete curing, and pressure transmission obstruction, it provides operators or automated control systems with clear early warning information and adjustment directions. This effectively avoids component scrapping due to process deviations, significantly improving product yield and production efficiency. This proactive risk identification mechanism greatly enhances the robustness and intelligence of the molding process.

[0059] In a preferred embodiment of this application, the step of comparing the first simulation result with a preset quality threshold to identify potential process risks of the component includes: When the first simulation results show that the predicted temperature of a thin-walled region in the component with a thickness less than a preset first thickness threshold exceeds a preset resin thermal damage temperature threshold, it is identified as a risk of overheating damage. When the first simulation result shows that the predicted degree of curing of the thick-walled region in the component with a thickness greater than a preset second thickness threshold is lower than a preset target degree of curing threshold, it is identified as a risk of incomplete curing; the second thickness threshold is greater than the first thickness threshold. When the first simulation results show that the predicted resin viscosity of the thin-walled region of the component rises above the preset gel viscosity threshold, the predicted resin viscosity of the thick-walled region is lower than the gel viscosity threshold, and the predicted curing degree of the thick-walled region is lower than the preset compaction window curing degree threshold, then it is identified as a risk of pressure transmission obstruction.

[0060] During the curing and molding process of carbon fiber composite components, the complexity of their geometry leads to significant differences in thermal response and curing behavior in different regions. For example, thin-walled regions, due to their small heat capacity and rapid heat dissipation, are more prone to temperature overshoot, thus posing a risk of overheating damage. Thick-walled regions, on the other hand, have longer heat transfer paths and slower heat dissipation, which may result in incomplete curing. Therefore, this application introduces a first thickness threshold and a second thickness threshold to distinguish between thin-walled and thick-walled regions of the component. The second thickness threshold is set to be greater than the first thickness threshold to ensure that the risk assessment of different regions has a clear geometric basis.

[0061] The resin thermal damage temperature threshold refers to the critical temperature at which the resin material may degrade or deteriorate in performance at high temperatures. Once the predicted temperature exceeds this threshold, it indicates a risk of overheating damage. The target degree of curing threshold refers to the minimum degree of curing required for the component to achieve the expected mechanical properties and structural integrity. If the predicted degree of curing in thick-walled areas is lower than this threshold, it indicates a risk of incomplete curing.

[0062] The gel viscosity threshold refers to the critical viscosity at which the resin transitions from a liquid to a gel state. Once the resin reaches the gel state, its fluidity decreases sharply. The compaction window curing degree threshold defines the range of curing degree within which the resin can still effectively transmit pressure and achieve good compaction during the curing process. When the predicted resin viscosity in the thin-walled region rises above the gel viscosity threshold, while the predicted resin viscosity in the thick-walled region remains below the gel viscosity threshold, and the predicted curing degree in the thick-walled region is below the compaction window curing degree threshold, this indicates that the thin-walled region may have prematurely gelled, hindering the effective transmission of pressure to the thick-walled region. This results in insufficient compaction of the thick-walled region, creating a risk of impaired pressure transmission.

[0063] This application introduces refined judgment criteria based on component geometry (thin-walled and thick-walled regions) and key material state parameters (temperature, viscosity, and degree of curing). This allows for more precise diagnosis of the specific types and potential areas of risks such as overheating damage, incomplete curing, and pressure transmission obstruction. This refined risk identification avoids vague risk warnings, enabling more targeted subsequent control commands. Consequently, it not only effectively prevents component defects, improves product quality and production efficiency, but also reduces unnecessary process adjustments and optimizes resource consumption, thus providing a more reliable guarantee for the intelligent manufacturing of complex geometric carbon fiber composite components.

[0064] The following is a specific example to illustrate this.

[0065] Suppose a carbon fiber composite component with a complex geometry is being cured in an autoclave. The component comprises a thin-walled flange structure with a thickness of 2 mm and a thick-walled web structure with a thickness of 18 mm.

[0066] During the curing process, the digital model continuously predicts the evolution of the internal state of the component.

[0067] When the first simulation results showed that the predicted temperature of the flange structure (identified as a thin-walled region with a thickness of less than 3 mm) reached 195 degrees Celsius, while the preset resin heat damage temperature threshold was 185 degrees Celsius, the system immediately identified that the flange structure was at risk of overheating damage.

[0068] When the predicted curing degree of the web structure (identified as a thick-walled region with a thickness greater than 10 mm) is 88%, while the preset target curing degree threshold is 95%, the system identifies that the web structure has a risk of incomplete curing.

[0069] When the predicted resin viscosity of the flange structure rises to 800 Pa·s (higher than the preset gel viscosity threshold of 500 Pa·s), while the predicted resin viscosity of the web structure remains at 300 Pa·s, and the predicted degree of curing of the web structure is 35% (lower than the preset curing degree threshold of 45% for the compaction window), the system identifies that there is a risk of pressure transmission obstruction in this component, indicating that the flange structure may have gelled prematurely, hindering the effective compaction of the web structure.

[0070] In a more preferred embodiment of this application, the step of determining the control instructions for adjusting the process parameters of the curing molding based on the potential process risks includes: In response to the identified risk of overheating damage, control commands are determined to reduce the heating power of the autoclave or set the temperature. In response to the identified risk of incomplete curing, control commands are determined to increase the heating power of the autoclave, set the temperature, or extend the holding time of the corresponding temperature stage. When a risk of pressure transmission obstruction is identified, a control command is determined to adjust the pressure application rate or pressure value of the autoclave.

[0071] When the system identifies a risk of overheating damage, it typically means that the predicted internal temperature of a localized area of ​​the component, particularly thin-walled areas, has exceeded or is about to exceed the resin's thermal damage temperature threshold. To prevent resin degradation or internal defects in the component, immediate measures are needed to reduce heat input. Therefore, the control command is determined to reduce the autoclave's heating power or set temperature. Reducing the heating power directly decreases the heat transferred from the autoclave to the component, while lowering the set temperature lowers the autoclave's temperature control target, thereby slowing down or reversing the upward trend in internal component temperature.

[0072] When the system identifies a risk of incomplete curing, this typically occurs in thick-walled areas of the component where the predicted degree of curing is below the preset target curing threshold. This means the resin in these areas has not yet fully cross-linked, potentially leading to substandard mechanical properties in the final component. To promote further resin curing, increased heat input is required. Therefore, control commands are determined to increase the autoclave's heating power, set the temperature, or extend the holding time at the corresponding temperature stage. Increasing the heating power or setting the temperature accelerates the resin's curing reaction rate, while extending the holding time provides the resin with a longer reaction time, ensuring it reaches the target degree of curing.

[0073] When the system identifies a risk of pressure transmission obstruction, it indicates uneven resin viscosity distribution within the component, potentially leading to insufficient compaction or excessive porosity. Specifically, the resin in thin-walled areas may have gelled, while the resin in thick-walled areas remains at a lower viscosity but is not sufficiently cured to withstand compaction. To ensure effective pressure transmission and component compaction, fine-tuning of pressure parameters is required. Therefore, the control command is determined to adjust the pressure application rate or pressure value of the autoclave. Adjusting the pressure application rate controls the rate of pressure rise to accommodate variations in resin viscosity in different areas, preventing excessive pressure on prematurely gelled areas or insufficient pressure on late-gelled areas. Adjusting the pressure value directly alters the compaction intensity applied to the component, optimizing resin flow and pore drainage.

[0074] This application's solution establishes a direct mapping relationship between potential process risks and specific process parameter adjustment instructions, achieving precise and intelligent control of the carbon fiber composite molding process. Because the system can determine the most appropriate control instruction for each identified risk type, the autoclave can make targeted parameter adjustments based on the real-time internal state and future evolution trends of the components. This refined control strategy avoids over- or under-adjustments that may occur in traditional control methods, thereby improving control efficiency and accuracy.

[0075] In some embodiments of this application, the step of performing a second simulation in the updated digital model with control commands as input to obtain a second simulation result, verifying the control commands based on the second simulation result, and sending the verified control commands to the autoclave for execution preferably includes: Before sending control commands to the autoclave for execution, a new prediction of the internal state evolution of the component is performed in the digital model with the control commands as input, to obtain a second simulation result; Determine whether the second simulation result meets the preset verification conditions, the verification conditions including: the potential process risk is eliminated and no new process deviation is introduced; If so, the control command is confirmed as verified and sent to the autoclave's controller for execution; If not, based on the potential process risks reflected in the second simulation results, the parameters of the control commands are iteratively adjusted through a preset optimization algorithm, the control commands are redefined, and a new round of verification is conducted.

[0076] Before sending the control command to the autoclave for execution, the system performs a forward-looking simulation using the updated digital model as input. This simulation aims to predict the evolution of key parameters such as internal temperature distribution, resin state, and degree of curing within a preset time period after the application of the control command, thus obtaining a second simulation result. This step ensures that any actual control action is based on thorough prediction and evaluation.

[0077] Determining whether the second simulation results meet the preset verification conditions is the core step. These verification conditions are set with a dual standard: first, previously identified potential process risks must be effectively eliminated; second, no new process deviations must be introduced. These verification conditions aim to ensure the comprehensive effectiveness and safety of control instructions.

[0078] If the second simulation result meets all the above verification conditions, the control command is considered verified and is safely sent to the autoclave controller to guide the actual curing process. This ensures that only rigorously verified commands are executed, thus avoiding potential process problems.

[0079] If the second simulation result fails to meet the preset verification conditions—that is, if potential process risks still exist or new process deviations are introduced—the system will not immediately execute the instruction. Instead, it will initiate an iterative optimization process based on the specific problems reflected in the second simulation result. The parameters of the control instruction are adjusted using preset optimization algorithms, such as genetic algorithms, particle swarm optimization, or gradient-based optimization methods. For example, if the simulation shows a continued risk of overheating, the heating rate may be further reduced; if incomplete curing is indicated, the holding time may be extended. The adjusted control instruction is then input into the digital model again for a new round of simulation and verification until an optimized instruction that meets all verification conditions is found.

[0080] The proposed solution effectively addresses potential shortcomings in the control command verification of the basic solution by conducting a forward-looking simulation in a digital model before the actual execution of control commands and evaluating the simulation results based on rigorous verification conditions. It is precisely this forward-looking simulation and verification process that ensures the stability of the molding process and product quality to the greatest extent possible when the generated control commands are actually applied to the autoclave.

[0081] The following is a specific example to illustrate this.

[0082] Suppose that during the curing and molding process of carbon fiber composite materials, the risk identification module identifies a risk of overheating damage in thin-walled areas of the component. The instruction determination module initially generates a control instruction, for example, to reduce the heating power of the autoclave by 10%. Before sending this instruction to the autoclave, the instruction sending module first performs a prospective simulation in the updated digital model, using the instruction to reduce the heating power by 10% as input.

[0083] The second simulation results show that although the risk of overheating damage in the thin-walled area was indeed eliminated, the excessive reduction in heating power caused the curing degree of the thick-walled area of ​​the component to fail to reach the preset target curing degree threshold, thus introducing a new process deviation of incomplete curing.

[0084] At this point, the system determines that the control command fails to meet the verification conditions. Based on the risk of incomplete curing reflected in the second simulation results, the system activates a preset optimization algorithm. This algorithm may suggest iterative adjustments to the control command, such as reducing the heating power by 5% while extending the holding time, or applying higher pressure at specific stages to assist curing. The adjusted control command is then input into the digital model for a new round of simulation verification. This process continues until an optimized control command is found that eliminates the risk of overheating damage, ensures sufficient curing of thick-walled areas, and does not introduce other process deviations. Finally, the verified optimized command is sent to the autoclave for execution, thereby ensuring the stability of the molding process and product quality.

[0085] like Figure 2 As shown, this application also discloses an intelligent control system 200 for carbon fiber composite material molding process, applied to the curing and molding of parts with complex geometries in an autoclave. The system includes: The model building module 210 is used to acquire and build a digital model of the component for process simulation based on the geometric structure information and material property information of the component. The model update module 220 is used to acquire and infer the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component. The internal state information includes the internal temperature distribution, resin state and degree of curing, and updates the internal state information to the digital model. The risk identification module 230 is used to predict the internal state evolution of the component within a preset time period in the future based on the updated digital model, obtain a first simulation result, and identify potential process risks of the component based on the first simulation result. The instruction determination module 240 is used to determine control instructions for adjusting the process parameters of the curing molding based on the potential process risks. The instruction sending module 250 is used to simulate again in the updated digital model with the control instruction as input to obtain a second simulation result, verify the control instruction based on the second simulation result, and send the verified control instruction to the autoclave for execution.

[0086] The model building module 210 can be implemented as a standalone software component, integrating a computer-aided design (CAD) data interface and a material database interface. In a preferred embodiment, the model building module 210 can provide a user interface that allows operators to manually input or correct geometric and material property information, or to automatically acquire the required data through integration with an external data management system.

[0087] The model update module 220 can be configured as a real-time data processing unit that communicates with a first type of sensor inside the autoclave and a second type of sensor on the surface of the component mold. For example, the model update module 220 can use a data assimilation method based on Kalman filtering or particle filtering to fuse real-time sensor data with the digital model to calibrate the internal state of the model and ensure that the model accurately reflects the actual curing process.

[0088] The risk identification module 230 can be implemented as a predictive analytics engine that receives simulation results from an updated digital model. The risk identification module 230 can run predictive simulations faster than the actual curing process, predicting the evolution of the temperature field, resin viscosity field, and degree of curing field within the component over the next few minutes or hours. The risk identification module 230 can use rule-based expert systems or machine learning models to analyze the simulation results, thereby accurately identifying potential process problems.

[0089] The instruction determination module 240 can be configured as a decision support unit. The instruction determination module 240 has a built-in library of optimization algorithms or control strategies, such as PID controllers, fuzzy logic controllers, or model predictive controllers (MPC), and automatically calculates and determines the optimal control parameters such as heating power, set temperature, holding time, pressure application rate, or pressure value based on the identified risk type.

[0090] The instruction sending module 250 can be implemented as a control instruction management unit, which pre-verifies the control instructions in the digital model before sending them to the autoclave controller.

[0091] It should be noted that the information interaction and execution process between the above modules are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0092] The foregoing has provided a detailed description of the preferred embodiments of this application. However, this application is not limited to the above-described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application. All such equivalent modifications or substitutions are included within the scope defined in this application.

Claims

1. A smart control method for the molding process of carbon fiber composite materials, applied to the curing and molding of parts with complex geometries in an autoclave, characterized in that, Includes the following steps: Acquire and construct a digital model of the component for process simulation based on the component's geometric structure and material properties; Based on the internal environmental parameters of the autoclave and the surface state parameters of the component, the internal state information of the component is inferred, including the internal temperature distribution, resin state and degree of curing, and the internal state information is updated to the digital model. Based on the updated digital model, the internal state evolution of the component is predicted within a future preset time period to obtain a first simulation result. Based on the first simulation result, potential process risks of the component are identified. Based on the potential process risks, control instructions for adjusting the process parameters of the curing and molding process are determined. In the updated digital model, the control command is used as input to perform another simulation to obtain a second simulation result, and the control command is verified based on the second simulation result. The verified control command is then sent to the autoclave for execution.

2. The intelligent control method for the molding process of carbon fiber composite materials according to claim 1, characterized in that, The step of acquiring and constructing a digital model of the component for process simulation based on the component's geometric structure information and material property information includes: The three-dimensional geometric data of the component is obtained using computer-aided design software, which serves as the geometric structure information of the component. Based on a pre-set material database, the thermophysical parameters and resin curing kinetic parameters of the carbon fiber prepreg used in the component are obtained as material property information of the component; wherein, the thermophysical parameters include specific heat capacity and thermal conductivity, and the resin curing kinetic parameters include Arrhenius equation parameters used to describe the relationship between curing reaction rate and temperature; During the curing process, a preset disturbance is applied to the autoclave, and the autoclave's response to the disturbance is obtained to correct the material property information, thereby obtaining the corrected material property information. Based on the geometric structure information and the corrected material property information, a digital model is constructed to simulate the internal heat conduction and resin curing reaction process of the component.

3. The intelligent control method for the molding process of carbon fiber composite materials according to claim 2, characterized in that, The step of applying a preset disturbance to the autoclave during the curing process and obtaining the autoclave's response to the disturbance to correct the material property information, thereby obtaining the corrected material property information, includes: During the curing process, a temperature or pressure pulse of preset amplitude and duration is applied to the autoclave; Monitor the dynamic response of the autoclave and components to the temperature or pressure pulse; The dynamic response is matched with a pre-stored theoretical response feature library based on different combinations of material parameters to identify the actual values ​​of the resin curing kinetic parameters of the current batch of materials. By using the actual values ​​of the resin curing kinetic parameters of the current batch of materials, the resin curing kinetic parameters in the material property information are corrected to obtain the corrected material property information.

4. The intelligent control method for the molding process of carbon fiber composite materials according to claim 1, characterized in that, The step of acquiring and inferring the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component, wherein the internal state information includes the internal temperature distribution, resin state, and degree of curing, and updating the internal state information to the digital model includes: Based on a first type of sensor installed inside the autoclave, the internal temperature and pressure of the autoclave are acquired as internal environmental parameters. Based on a second type of sensor, which is arranged at multiple preset points on the surface of the mold used for the component, the temperature at the multiple preset points is obtained as a surface state parameter of the component. Based on the internal environment parameters and the surface state parameters, the heat conduction equation and the resin curing reaction kinetic equation, which include the resin curing exothermic term, are solved by a preset physical mechanism inference engine. The first round of inference is performed to infer the internal state information of the component and the temperature prediction values ​​of multiple preset points in real time. The internal state information of the component includes the internal temperature distribution, resin state and degree of curing of the component. The measured temperature values ​​of multiple preset points are obtained. Based on the temperature deviation between the predicted temperature value and the measured temperature value, the digital model is calibrated, and a second round of inference is performed to obtain the calibrated internal state information. Based on the calibrated internal state information, the calibrated digital model is adjusted and updated to obtain the updated digital model.

5. The intelligent control method for the molding process of carbon fiber composite materials according to claim 4, characterized in that, The steps of obtaining measured temperature values ​​at multiple preset points, calibrating the digital model based on the temperature deviation between the predicted temperature value and the measured temperature value, and performing a second round of inference to obtain calibrated internal state information include: Continuously monitor the temperature deviation between the measured temperature value at the preset point and the predicted temperature value at the corresponding preset point calculated by the digital model; When the temperature deviation is detected to continuously exceed the preset threshold at any preset point, the material property parameters of the local area in the digital model associated with the preset point where the temperature deviation occurs are adjusted to obtain the adjusted digital model. Based on the adjusted digital model, the physical mechanism inference engine is rerun to perform the second round of inference and obtain the calibrated internal state information.

6. The intelligent control method for the molding process of carbon fiber composite materials according to claim 1, characterized in that, The step of predicting the internal state evolution of the component within a predetermined time period based on the updated digital model to obtain a first simulation result, and identifying potential process risks of the component based on the first simulation result, includes: In the updated digital model, the evolution of the temperature field, resin viscosity field, and degree of curing field inside the component is predicted over a future preset time period at a speed faster than the real-time curing and molding speed of the component, to obtain a first simulation result; the first simulation result includes predicted data of temperature, viscosity, and degree of curing distribution information at each time point within the future preset time period. The first simulation result is compared with a preset quality threshold to identify potential process risks of the component; the potential process risks include overheating damage risk, incomplete curing risk, and pressure transmission obstruction risk.

7. The intelligent control method for the molding process of carbon fiber composite materials according to claim 6, characterized in that, The step of comparing the first simulation result with a preset quality threshold to identify potential process risks of the component includes: When the first simulation results show that the predicted temperature of a thin-walled region in the component with a thickness less than a preset first thickness threshold exceeds a preset resin thermal damage temperature threshold, it is identified as a risk of overheating damage. When the first simulation result shows that the predicted degree of curing of the thick-walled region in the component with a thickness greater than a preset second thickness threshold is lower than a preset target degree of curing threshold, it is identified as a risk of incomplete curing; the second thickness threshold is greater than the first thickness threshold. When the first simulation results show that the predicted resin viscosity of the thin-walled region of the component rises above the preset gel viscosity threshold, the predicted resin viscosity of the thick-walled region is lower than the gel viscosity threshold, and the predicted curing degree of the thick-walled region is lower than the preset compaction window curing degree threshold, then it is identified as a risk of pressure transmission obstruction.

8. The intelligent control method for the molding process of carbon fiber composite materials according to claim 7, characterized in that, The step of determining the control instructions for adjusting the process parameters of the curing molding based on the potential process risks includes: In response to the identified risk of overheating damage, control commands are determined to reduce the heating power of the autoclave or set the temperature. In response to the identified risk of incomplete curing, control commands are determined to increase the heating power of the autoclave, set the temperature, or extend the holding time of the corresponding temperature stage. When a risk of pressure transmission obstruction is identified, a control command is determined to adjust the pressure application rate or pressure value of the autoclave.

9. The intelligent control method for the molding process of carbon fiber composite materials according to claim 1, characterized in that, In the updated digital model, the control command is used as input to perform another simulation to obtain a second simulation result, and the control command is verified based on the second simulation result. The steps of sending verified control commands to the autoclave for execution include: Before the control command is sent to the autoclave for execution, a new prediction of the internal state evolution of the component is performed in the digital model with the control command as input, to obtain a second simulation result; Determine whether the second simulation result meets the preset verification conditions, the verification conditions including: the potential process risk is eliminated and no new process deviation is introduced; If so, the control command is confirmed as verified and sent to the controller of the autoclave for execution; If not, based on the potential process risks reflected in the second simulation results, the parameters of the control command are iteratively adjusted using a preset optimization algorithm, the control command is redefined, and a new round of verification is performed.

10. An intelligent control system for the molding process of carbon fiber composite materials, applied to the curing and molding of parts with complex geometries in an autoclave, characterized in that, include: The model building module is used to acquire and construct a digital model of the component for process simulation based on the component's geometric structure information and material property information; The model update module is used to acquire and infer the internal state information of the component based on the internal environmental parameters of the autoclave and the surface state parameters of the component. The internal state information includes the internal temperature distribution, resin state and degree of curing, and the internal state information is updated to the digital model. The risk identification module is used to predict the internal state evolution of the component within a preset time period in the future based on the updated digital model, obtain a first simulation result, and identify potential process risks of the component based on the first simulation result. The instruction determination module is used to determine control instructions for adjusting the process parameters of the curing molding based on the potential process risks. The instruction sending module is used to simulate again in the updated digital model with the control instruction as input to obtain a second simulation result, and to verify the control instruction based on the second simulation result; The verified control command is then sent to the autoclave for execution.

Images