Multi-stage cooling control method and system for autoclave based on flow direction switching

Abstract

The invention discloses an autoclave multi-stage cooling control method and system based on flow direction switching, and belongs to the technical field of autoclave temperature control, and the method comprises the steps: obtaining the data of an autoclave internal temperature monitoring point and the three-dimensional geometrical characteristics of a composite material component, and obtaining the three-dimensional geometrical characteristics of the composite material component through a space mapping algorithm; dividing a plurality of dynamic virtual micro-regions corresponding to heat dissipation parts of the component in the autoclave; by comparing the average thermal state distribution of each micro-area with the target cooling track, generating a micro-area regulation and control instruction set containing regulation and control priority and heat dissipation demand intensity; and responding to the instruction set, scheduling a plurality of distributed fluid guide units with independent rotating angles and opening degrees in the autoclave, and forming directional cooling flow fields aiming at different parts through local pneumatic coupling self-organization. According to the method, a multi-stage cooperative control strategy based on flow direction switching is adopted, dynamic virtual micro-area division is combined, and fine directional regulation and control of the cooling process of the composite material component in the autoclave can be achieved.

Description

Technical Field

[0001] This invention relates to the field of autoclave temperature control technology, and in particular to a multi-stage cooling control method and system for autoclaves based on flow direction switching. Background Technology

[0002] As a key piece of equipment in the manufacture of high-performance composite material components, the autoclave's cooling control stage has a significant impact on the structural strength and dimensional stability of the finished product. Precise control of the cooling rate and spatial temperature difference ensures uniform shrinkage of the matrix material during the glass transition. With the increasing complexity of components in the aerospace field, the uniformity and response speed of the flow field inside the autoclave have become core indicators for evaluating process quality.

[0003] In related technologies, Chinese invention patent CN117075661A discloses a temperature and pressure feedback control system for an autoclave combined with fiber optic monitoring. The system includes: an autoclave for autoclave forming; a fiber optic monitoring system for acquiring real-time temperature and pressure data of the autoclave; and a main control system for connecting the autoclave and the fiber optic monitoring system and implementing software control. The main control system includes: an interface subsystem for displaying real-time temperature and pressure data of the autoclave; a subsystem for developing a temperature and pressure feedback control algorithm suitable for the autoclave; and a control subsystem for controlling the temperature and pressure feedback control of the autoclave. The output of the fiber optic monitoring system is connected to the main control system, and the output of the main control system is connected to the temperature control element and pressure control element of the autoclave. The input of the interface subsystem is connected to the autoclave, and the output of the control subsystem is connected to the autoclave. The control subsystem controls the temperature and pressure of the autoclave based on the feedback control algorithm of the algorithm subsystem to perform autoclave forming.

[0004] However, the aforementioned existing technical solutions have the following technical drawbacks. Existing solutions may primarily rely on macroscopic temperature and pressure feedback from the tank body for uniform control, failing to provide discretized and differentiated cooling flow field control based on the three-dimensional geometric characteristics and real-time thermal state of different parts of the composite material component. This can easily lead to internal temperature gradients and residual stress in complex structural components due to uneven heat dissipation during cooling. Traditional autoclaves may employ a large-volume macroscopic convection cooling method with a fixed or coarsely adjusted airflow path. This method struggles to eliminate stagnation zones formed by hot air in complex structural components, failing to achieve targeted cooling of locally overheated areas and thus suppressing spatial temperature gradients during cooling. Existing technologies may primarily rely on current monitoring data for feedback control, representing a delayed adjustment. They lack the ability to predict future temperature field evolution trends using real-time data, therefore failing to proactively intervene and compensate for flow field fluctuations before the risk of localized temperature gradient exceeding limits occurs, making it difficult to achieve precise gradient control and stress release throughout the entire process. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a multi-stage cooling control method and system for autoclaves based on flow direction switching. By employing a multi-stage collaborative control strategy based on flow direction switching and combining it with dynamic virtual micro-region partitioning, it is possible to achieve precise directional control of the cooling process of composite material components in the autoclave.

[0006] The above objectives can be achieved through the following approach: A multi-stage cooling control method and system for autoclaves based on flow direction switching includes acquiring temperature monitoring point data and three-dimensional geometric features of composite material components inside the autoclave; dividing the autoclave into multiple dynamic virtual micro-regions corresponding to the heat dissipation parts of the components through a spatial mapping algorithm; generating a micro-region control instruction set containing control priority and heat dissipation demand intensity by comparing the average thermal distribution of each micro-region with the target cooling trajectory; responding to the instruction set, scheduling multiple distributed fluid guiding units with independent rotation angles and openings inside the autoclave to form directional cooling flow fields for different parts through local aerodynamic coupling self-organization.

[0007] Optionally, the process of dividing the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component includes: identifying the cross-sectional thickness span and surface curvature changes in the three-dimensional geometric features, resolving the physical entity of the composite material component into multiple component sub-regions containing thick-walled regions, thin-walled regions, and connecting regions; based on the spatial coordinates of the component sub-regions, retrieving airflow space nodes located within the range of the outward normal direction, establishing a mapping relationship between the physical entity and the airflow space, and generating initial virtual micro-regions; obtaining the temperature difference and thermal conductivity ratio between adjacent initial virtual micro-regions, calculating a collaborative evaluation index reflecting the thermodynamic synchronicity between regions; comparing the collaborative evaluation index with a preset fusion threshold, and outputting dynamic virtual micro-regions whose boundaries dynamically evolve with the temperature gradient through spatial topological reorganization of adjacent initial virtual micro-regions.

[0008] Optionally, generating a micro-area control instruction set that includes control priority and heat dissipation demand intensity includes: extracting the desired temperature value corresponding to the current moment in the target cooling trajectory; calculating the difference between the real-time average temperature of the dynamic virtual micro-area and the desired temperature value to obtain a first temperature difference parameter; retrieving the instantaneous temperature extreme values ​​of all monitoring points within the dynamic virtual micro-area to obtain a second temperature difference parameter characterizing the gradient within the micro-area; obtaining the material density and specific heat capacity parameters associated with the three-dimensional geometric features; performing normalized weighted calculations in combination with the first temperature difference parameter and the second temperature difference parameter to generate a comprehensive control score; and assigning corresponding execution order fields according to the descending order of the comprehensive control score, thus forming a micro-area control instruction set.

[0009] Optionally, the formation of a directional cooling flow field through local aerodynamic coupling self-organization includes: receiving the micro-area control command set, converting the control commands for the macro-area into flow field sub-tasks for specific guiding nodes; distributing the flow field sub-tasks to the corresponding distributed fluid guiding units, obtaining real-time jet pressure and guide vane attitude angles fed back by adjacent units, and performing local communication and anti-interference coordination; adjusting the controllable guide vane rotation angle and micro-airflow valve adjustment amount inside each unit according to the feedback results of the local communication and anti-interference coordination; and utilizing the convergence effect of the air jets discharged from each distributed fluid guiding unit in the tank space to construct a directional cooling flow field that meets the heat dissipation directionality requirements.

[0010] Optionally, after the self-organized directional cooling flow field is formed, the method further includes: collecting the current configuration parameters of the directional cooling flow field and obtaining the thermophysical property parameters reflecting the thermal conductivity of the material, and deducing and generating thermal state prediction parameters describing the temperature field evolution over a future time period; comparing the predicted local temperature difference values ​​in the thermal state prediction parameters with the gradient safety threshold set internally by the system, and identifying target micro-regions where the predicted gradient has a risk of exceeding the limit; generating a preventive flow field compensation command for the target micro-region, and inserting the command into the current cycle's micro-region control command set to trigger real-time pre-adjustment of the directional cooling flow field.

[0011] Optionally, the generation of preventative flow field compensation instructions includes: parsing the three-dimensional temperature distribution cloud map in the thermal state prediction parameters, identifying grid node groups with temperatures higher than the average reference as overheated sub-regions, and identifying grid node groups with temperatures lower than the average reference as underheated sub-regions; calculating the geometric center straight-line distance and predicted temperature difference span between the overheated sub-regions and the underheated sub-regions, and evaluating the adjustment intensity score required to achieve thermal balance; calculating the execution increment for heat transfer based on the adjustment intensity score, and driving the corresponding distributed fluid guiding units located on the heat transfer axis of the overheated sub-regions and the underheated sub-regions to perform flow direction bias.

[0012] Optionally, the method further includes: during the cooling operation, collecting actual temperature data of each of the dynamic virtual micro-regions within a continuous monitoring period; calculating the decrease in the actual temperature data to obtain the actual temperature change rate reflecting the actual heat exchange intensity; comparing the actual temperature change rate with the reference change rate extracted from the target cooling trajectory to calculate the rate deviation parameter; and using the rate deviation parameter to correct the intensity ratio coefficient in the micro-region control command set.

[0013] Optionally, the method further includes: after starting the cooling process, dividing the cooling process into a first stage for breaking thermal stratification, a second stage for gradient control, and a third stage for stress release based on the initial temperature field distribution; setting corresponding stage-specific target temperature ranges and spatial temperature difference limits for each stage; obtaining the status bit signal of the current process stage, and loading corresponding control strategy weights accordingly to drive cyclic execution under different stage constraints.

[0014] Optionally, the step of acquiring real-time temperature data of each monitoring point inside the autoclave and three-dimensional model data of the composite material component, and extracting three-dimensional geometric features from the three-dimensional model data, includes: acquiring real-time temperature data of monitoring points throughout the autoclave, and performing noise reduction and interpolation completion processing on the acquired data to generate a spatiotemporally continuous temperature field dataset; performing lightweight topology reconstruction and geometric feature extraction on the composite material component, eliminating redundant surfaces and invalid control points, and generating three-dimensional geometric features; and registering the temperature field dataset with the three-dimensional geometric features in a spatial coordinate system to extract the three-dimensional geometric features of the temperature attributes.

[0015] Based on the same inventive concept, this invention also provides a multi-stage cooling control system for autoclaves based on flow direction switching. The system includes: a data acquisition and three-dimensional geometric feature extraction module, used to acquire real-time temperature data of each monitoring point inside the autoclave and three-dimensional model data of the composite material component, and extract three-dimensional geometric features from the three-dimensional model data; a dynamic virtual micro-region division module, used to divide the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component based on the real-time temperature data and the three-dimensional geometric features through a preset spatial mapping algorithm; a control instruction set generation module, used to acquire the average thermal distribution of the dynamic virtual micro-regions and the target cooling trajectory matching the process, and generate a micro-region control instruction set containing control priority and heat dissipation demand intensity through comparison calculation; and a flow direction switching and control execution module, used to respond to the micro-region control instruction set, schedule multiple distributed fluid guiding units with independent rotation angles and openings inside the autoclave, and form a directional cooling flow field through local aerodynamic coupling self-organization.

[0016] Compared with the prior art, the present invention has the following advantages: This invention utilizes a spatial mapping algorithm based on three-dimensional geometric features to divide dynamic virtual micro-regions, achieving a discretized and refined characterization of the non-uniform spatial thermodynamic state within an autoclave. This approach transforms the differentiated physical properties of components, such as thick-walled and thin-walled regions, into quantifiable flow field control variables, fundamentally solving the problem of uneven heat dissipation caused by geometrical differences in large and complex structural components during cooling, and ensuring the synchronicity of thermal response within the micro-region.

[0017] This invention introduces a distributed fluid scheduling mechanism based on a micro-area control instruction set, constructing a directional cooling flow field through multiple independently directional fluid guiding units. Compared to the large-volume macroscopic convection used in traditional autoclaves, this local aerodynamic coupling control method can dynamically switch airflow paths according to the heat dissipation requirements of different parts, eliminating the stagnation effect of hot air in shielded areas and suppressing the local spatial temperature gradient generated during the cooling process.

[0018] This invention deconstructs the cooling process into phased target management through multi-level cooling control and real-time pre-adjustment strategies. The system can use predictive algorithms to identify and compensate for gradient overshoot risks in advance, transforming temperature control from passive, lagging adjustment to active, feedforward control. This end-to-end gradient monitoring and deviation correction mechanism releases internal residual stress generated during component cooling and shrinkage, preventing interlaminar cracking and geometric temperature-induced distortion in composite materials, and improving the stability of molding quality.

[0019] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic flowchart of a multi-stage cooling control method for autoclaves based on flow direction switching, according to an embodiment of the present invention.

[0022] Figure 2 This is an evolution trajectory diagram of the dynamic virtual micro-region collaborative evaluation index according to an embodiment of the present invention.

[0023] Figure 3 This is a thermal state prediction cloud map according to an embodiment of the present invention.

[0024] Figure 4 This is a schematic diagram of the structure of a multi-stage cooling control system for autoclaves based on flow direction switching, according to an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Reference Figure 1 One embodiment of the present invention proposes a multi-stage cooling control method for autoclaves based on flow direction switching. By adopting a multi-stage collaborative control strategy based on flow direction switching and combining dynamic virtual micro-region division, it is possible to achieve fine-grained directional control of the cooling process of composite material components in autoclaves.

[0027] The method described in this embodiment specifically includes: S1. Obtain real-time temperature data and three-dimensional model data of composite material components at various monitoring points inside the autoclave, and extract the three-dimensional geometric features from the three-dimensional model data; Optionally, acquiring real-time temperature data at various monitoring points inside the autoclave and three-dimensional model data of the composite material components, and extracting three-dimensional geometric features from the three-dimensional model data includes: Real-time temperature data from all monitoring points inside the tank is acquired, and the collected data is denoised and interpolated to generate a spatiotemporally continuous temperature field dataset. The composite material component is subjected to lightweight topology reconstruction and geometric feature extraction, redundant surfaces and invalid control points are removed, and three-dimensional geometric features are generated. The temperature field dataset is registered with the three-dimensional geometric features in a spatial coordinate system to extract the three-dimensional geometric features of the temperature attributes.

[0028] Specifically, real-time temperature values ​​of all monitoring points deployed throughout the autoclave are first collected. Then, a filtering algorithm is used to remove noise data generated by equipment, environment and other factors during the collection process. At the same time, a spatial interpolation algorithm is used to fill in the blank areas or temporarily missing temperature data of the measuring points. Finally, the data are integrated to form a spatiotemporal continuous temperature field dataset that reflects the dynamic changes of temperature in the time dimension and covers all locations in the autoclave in the spatial dimension. This provides accurate and complete basic temperature data for subsequent registration with the three-dimensional geometric features of composite material components and division of dynamic virtual micro-regions.

[0029] Data from missing measurement points is recovered using an interpolation completion algorithm, thereby constructing a temperature field dataset that is continuous in both time and space. Simultaneously, a lightweight topology reconstruction is performed on the 3D model of the composite material component. By identifying and eliminating redundant surface information and invalid control points, 3D geometric features that can accurately characterize the component's morphology are extracted.

[0030] The generated temperature field dataset is registered with the three-dimensional geometric features in a spatial coordinate system, so that each geometric feature point has a corresponding real-time temperature attribute, achieving deep fusion of temperature attributes and geometric features. In order to achieve refined characterization of different heat dissipation parts of the component, the system uses a preset spatial mapping algorithm to identify the cross-sectional thickness span and surface curvature changes in the three-dimensional geometric features, and resolves the physical entity of the component into multiple sub-regions including thick-walled regions, thin-walled regions, and connection regions. The spatial mapping algorithm employs a hybrid architecture of physical modeling and data-driven approaches. It presents a four-layer modular closed-loop structure encompassing geometric feature analysis, spatial node matching, thermodynamic synchronicity assessment, and dynamic topology reorganization and feedback. Input parameters include the 3D geometric features of the component, in-tank temperature monitoring, algorithm preset controls, and basic material thermophysical properties. After processing by each module, it first outputs preliminary and intermediate results such as component physical sub-region division and initial mapping relationships. The final output is the core result: a dynamic virtual micro-region whose boundary evolves dynamically with the temperature gradient. This algorithm does not require deep learning training with large-scale labeled data. It adopts a lightweight training mode of offline initialization calibration and online iterative optimization. First, it constructs a dataset based on historical process data, initially sets and calibrates core parameters, and then solidifies the basic model. Then, during actual process operation, it forms a closed-loop optimization by collecting cooling data, evaluating the division effect, and dynamically correcting parameters, continuously improving the accuracy of mapping and micro-region division to meet the real-time and diverse process requirements of autoclave cooling. Furthermore, it establishes a mapping relationship between physical entities and the in-tank airflow space, generating initial virtual micro-regions. To ensure the dynamic nature of the regulation, the thermodynamic synchronicity between regions is assessed by calculating the ratio of the temperature difference between adjacent micro-regions to the thermal conductivity of the material.

[0031] For example, taking the processing of an aerospace composite panel component with reinforcing ribs as an example, the reinforcing rib area of ​​this component is thicker and is analyzed by the system as a thick-walled region, while the panel skin area is thinner and is analyzed as a thin-walled region. At a certain moment during the cooling process, the system, through spatial coordinate system registration, finds that the real-time average temperature of the thick-walled region is higher, while the real-time average temperature of the thin-walled region is lower. The system performs comprehensive calculations based on the real-time temperature difference between the two regions, a preset reference temperature, and the thermal conductivity of the material itself, to derive an evaluation index reflecting the synchronous trend of the thermodynamic evolution of the two regions. If the index shows that the adjacent regions have a high degree of synchronicity in thermal evolution, the system will initiate a spatial topology reorganization procedure to merge the two original initial virtual micro-regions into a unified dynamic virtual micro-region for synchronous control. As cooling continues, if the temperature difference in the thick-walled region expands due to obstructed heat dissipation paths, causing the synchronicity index to exceed a preset threshold, the system will immediately trigger boundary evolution, splitting the merged region into independent dynamic virtual micro-regions and generating a new set of control instructions. At this time, the flow direction switching module will schedule specific distributed fluid guiding units to increase the deflection angle of the cooling airflow to the thick-walled area where the temperature drops slowly. By strengthening local heat dissipation, the temperature difference is smoothed out, thereby ensuring that the entire component cools down uniformly along the target trajectory, effectively avoiding micro-cracks or shrinkage deformation in the product.

[0032] S2. Based on the real-time temperature data and the three-dimensional geometric features, a preset spatial mapping algorithm is used to divide the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component. Optionally, the division of the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component includes: Identify the cross-sectional thickness span and surface curvature changes in the three-dimensional geometric features, and analyze the physical entity of the composite material component into multiple component sub-regions containing thick-walled regions, thin-walled regions, and connection regions; Based on the spatial coordinates of the component sub-region, airflow space nodes located within the range of the outer normal direction are retrieved, a mapping relationship between physical entities and airflow space is established, and an initial virtual micro-region is generated; The temperature difference and thermal conductivity ratio between adjacent initial virtual micro-regions are obtained, and a collaborative evaluation index reflecting the thermodynamic synchronicity between regions is calculated. By comparing the collaborative evaluation index with the preset fusion threshold, and by reorganizing the spatial topology of adjacent initial virtual micro-regions, a dynamic virtual micro-region whose boundary evolves dynamically with the temperature gradient is output.

[0033] Specifically, the evolution trajectory of dynamic virtual micro-region collaborative evaluation indicators is as follows: Figure 2As shown, by retrieving the digital data of the established three-dimensional model of the composite component, the overall physical morphology of the component is analyzed and identified. Two key geometric features are extracted from the model: first, the cross-sectional thickness span of different parts of the component, that is, the range and specific value of the cross-sectional thickness at each location, to distinguish between thick-walled, thin-walled, and other regions; second, the curvature change of the component surface, to clarify the degree of curvature, surface type, and location of curvature abrupt change, which serve as the core geometric basis for subsequent division of component sub-regions and matching cooling requirements.

[0034] Based on the spatial coordinates of each component's sub-region, airflow space nodes within that range are retrieved along the outward normal direction of its surface. Taking each sub-region of the composite component after analytical division as the object, the three-dimensional spatial coordinates of each sub-region are first determined. Then, using the surface of the sub-region as a reference, the outward normal direction perpendicular to the surface and pointing towards the airflow space inside the autoclave is determined. Subsequently, within the effective distance range pre-set by the process, airflow space nodes that fall within this direction and range and are pre-planned and deployed inside the autoclave are retrieved and screened. This establishes a precise spatial correspondence between the component's physical sub-regions and the airflow monitoring nodes inside the autoclave, laying a spatial matching foundation for subsequent directional cooling of the component based on airflow control.

[0035] To achieve intelligent adjustment of micro-region boundaries, the temperature difference between adjacent initial virtual micro-regions and the corresponding material thermal conductivity ratios are acquired in real time. A synergistic evaluation index reflecting the thermodynamic synchronicity between regions is then calculated using a formula. The formula for this index is: , Representative collaborative evaluation indicators, and These represent the real-time average temperature values ​​of two adjacent micro-regions. The system's preset reference temperature value was obtained by statistical fitting of the average peak temperature of batch components during the curing process, collected by thermocouples built into the autoclave. and These represent the thermal conductivity of the materials corresponding to the two regions. The temperature data is collected and transmitted in real time by monitoring points distributed within the autoclave, while the thermal conductivity is obtained from a pre-set database of thermophysical property parameters based on the component material.

[0036] After the calculation is completed, the collaborative evaluation indicators will be... The result is compared with a preset fusion threshold. If the index reflects a high degree of synchronization in the thermal evolution of adjacent regions, a spatial topology reorganization procedure is initiated. This dynamically merges or evolves the boundaries of adjacent micro-regions, outputting the final dynamic virtual micro-region. Based on thermodynamic principles, this method achieves a finely discretized characterization of the non-uniform space within the autoclave. It can adjust the control granularity in real time according to the thickness and shape differences of different parts of the component, ensuring the on-demand allocation of cooling resources in space. This fundamentally suppresses the temperature gradient caused by local thermal hysteresis during the cooling process of complex components, effectively preventing micro-cracks in the product.

[0037] For example, taking the processing of an aerospace composite panel component with reinforcing ribs as an example, the reinforcing rib area of ​​the component is thicker and is analyzed as a thick-walled region, while the panel skin area is thinner and is analyzed as a thin-walled region. At a certain moment during the cooling process, the sensor measures the real-time average temperature of the thick-walled region. The real-time average temperature of the thin-walled region is 155 degrees Celsius. The system's reference temperature is 150 degrees Celsius. The temperature is set to 180 degrees Celsius. Because the wall panels are made of homogeneous materials, the thermal conductivity ratio is... The result equals 1. The collaborative evaluation index is calculated. If the system's preset fusion threshold is 1.05, due to the current collaborative evaluation indicators... If the temperature is below this threshold, the system determines that the two regions have high thermodynamic synchronicity at the current cooling rate. Therefore, it merges the two initial virtual micro-regions into a unified dynamic virtual micro-region for synchronous control through spatial topology reorganization. As cooling continues, if the temperature of the thick-walled region decreases due to obstructed heat dissipation paths... Maintained at 130 degrees Celsius, while the temperature in the thin-walled region... The temperature has dropped to 110 degrees Celsius, at which point the temperature difference has widened to 20 degrees Celsius. The indicators were calculated again. Since this value exceeds the fusion threshold of 1.05, the system immediately triggers boundary evolution, splitting the original merged region into two independent dynamic virtual micro-regions and generating a new micro-region control instruction set. At this time, the flow direction switching module will schedule specific fluid guiding units according to the instructions to increase the offset angle of the cooling airflow to the thick-walled region, thereby reducing the temperature difference by enhancing local heat dissipation. This ensures that the entire component cools down uniformly according to the target cooling trajectory, preventing shrinkage deformation of the wall panel caused by uneven cooling.

[0038] S3. Obtain the average thermal distribution of the dynamic virtual micro-region and the target cooling trajectory that matches the process, and generate a micro-region control instruction set containing control priority and heat dissipation demand intensity through comparison calculation; Optionally, generating a micro-area control instruction set that includes control priority and heat dissipation demand intensity includes: Extract the desired temperature value corresponding to the current moment from the target cooling trajectory, calculate the difference between the real-time average temperature of the dynamic virtual micro-region and the desired temperature value, and obtain the first temperature difference parameter; The instantaneous temperature extreme values ​​of all monitoring points within the dynamic virtual micro-region are retrieved to obtain the second temperature difference parameter characterizing the gradient within the micro-region; The material density and specific heat capacity parameters associated with the three-dimensional geometric features are obtained, and a normalized weighted calculation is performed by combining the first temperature difference parameter and the second temperature difference parameter to generate a comprehensive control score. Based on the descending order of the comprehensive control scores, corresponding execution order fields are assigned and aggregated into a micro-area control instruction set.

[0039] Specifically, the desired temperature value corresponding to the current moment is extracted from the target cooling trajectory, which refers to the ideal cooling curve required by the process. The difference between the real-time average temperature of the dynamic virtual micro-region and this desired temperature value is calculated to obtain the first temperature difference parameter.

[0040] To further understand the temperature uniformity within the micro-region, the system retrieves the instantaneous temperature extremes of all monitoring points within the dynamic virtual micro-region, i.e., the difference between the highest and lowest temperatures, to obtain a second temperature difference parameter characterizing the thermal gradient within the micro-region.

[0041] The density and specific heat capacity parameters of the composite material, associated with its three-dimensional geometric features, are obtained. Specific heat capacity refers to the amount of heat required to change the temperature per unit mass of material. These parameters are then used to perform a normalized weighted calculation to generate a comprehensive control score. The formula for calculating this score is as follows: , In this formula, Represents the overall regulatory score. This represents the material density obtained by retrieving data from a database linked to a 3D model. This represents the specific heat capacity of materials obtained in the same manner. and The preset weighting coefficients, The first temperature difference parameter obtained from the aforementioned calculation, This is the second temperature difference parameter characterizing the gradient within the micro-region.

[0042] First, calculate the comprehensive control score of all dynamic virtual micro-regions. Then, sort the dynamic virtual micro-regions in descending order of their scores. Based on this sorting result, assign the corresponding cooling control execution order field information to each micro-region. Finally, integrate and summarize all the control information, including the control priority, heat dissipation demand intensity, and corresponding execution order of each micro-region, to form a complete micro-region control instruction set.

[0043] For example, suppose there are two dynamic virtual micro-regions during the autoclave cooling process: region A is a thick-walled region, and region B is a thin-walled region. The target cooling trajectory at the current process moment requires a desired temperature of 100 degrees Celsius. Monitoring shows that the average temperature of region A is 110 degrees Celsius, with the highest temperature at its internal monitoring point being 113 degrees Celsius and the lowest being 108 degrees Celsius. Then, the first temperature difference parameter for region A... The second temperature difference parameter is 10 degrees Celsius. The temperature is 5 degrees Celsius. The average temperature of region B is 105 degrees Celsius, with a maximum internal temperature of 106 degrees Celsius and a minimum internal temperature of 104 degrees Celsius. What is the first temperature difference parameter for region B? The second temperature difference parameter is 5 degrees Celsius. The temperature is 2 degrees Celsius. Set the material density. It is 1600 kg per cubic meter, with a specific heat capacity of 1600 kg per cubic meter. The weighting factor is 1200 joules per kilogram Kelvin. and All values ​​are set to 1. Substituting these values ​​into the formula, the comprehensive control score for region A is calculated. Joules per cubic meter. The comprehensive control score for Region B. Joules per cubic meter. The system identifies region A as having a higher score than region B. Therefore, in the generated micro-area control instruction set, the execution order field of region A is set first, prioritizing the scheduling of distributed fluid guiding units to enhance directional cooling for it. Through this priority allocation, the system can ensure that thick-walled regions with large heat loads receive more timely flow field compensation, maintaining the synchronicity of the overall cooling process.

[0044] S4. In response to the micro-area control command set, schedule multiple distributed fluid guiding units with independent rotation angles and openings in the autoclave to form a directional cooling flow field through local aerodynamic coupling self-organization.

[0045] Optionally, the self-organization of the directional cooling flow field through local aerodynamic coupling includes: Receive the micro-area control instruction set and transform the control instructions for the macro-area into flow field sub-tasks for specific guiding nodes; The flow field subtask is sent to the corresponding distributed fluid guiding unit to obtain the real-time jet pressure and guide vane attitude angle fed back by adjacent units, and to perform local communication and anti-interference coordination. Based on the feedback results of the local communication and anti-interference coordination, adjust the rotation angle of the controllable guide vanes and the adjustment amount of the micro-airflow valves inside each unit; By utilizing the convergence effect of the air jets discharged from each distributed fluid guiding unit in the tank space, a directional cooling flow field that meets the heat dissipation directionality requirements is constructed.

[0046] Specifically, after receiving the micro-area control instruction set, it is necessary to transform the control instructions for the macro-area into flow field sub-tasks for specific guiding nodes. The so-called distributed fluid guiding unit refers to a modular terminal deployed inside the autoclave, possessing independent power execution capabilities, and capable of autonomously adjusting the physical attitude of its airflow jet according to instructions.

[0047] The flow field subtasks are distributed to the corresponding distributed fluid guidance units. Each unit acquires its own jet pressure and guide vane attitude angle in real time through integrated sensors, and retrieves operating parameters fed back by adjacent units using the in-tank communication network, thereby performing local communication and anti-interference coordination. This local aerodynamic coupling self-organizing process aims to avoid the collision of multiple airflows in space through momentum balance between nodes.

[0048] To accurately and quantitatively perform the cooling operation, a pneumatic feedback correction formula is used to calculate the micro-flow valve adjustment amount. This formula is expressed as: , In this formula, The adjustment amount of the micro-flow valve, which represents the final output, directly determines the actual exhaust flow rate of the nozzle. It is the base flow preset value assigned to this node by the flow field subtask. It is a preset dimensionless coupling coefficient used to adjust the sensitivity of local pressure deviation to flow correction. This represents the average jet pressure of adjacent guide units obtained through communication. This represents the real-time jet pressure directly measured by the current unit sensor.

[0049] Each distributed fluid guiding unit, according to the control command, precisely adjusts the direction, flow rate and pressure of the jet before ejecting an air jet. These jets converge and superimpose with each other in the internal space of the autoclave to form a convergence effect, allowing the airflow to flow precisely in the direction of heat dissipation according to the heat dissipation needs of each area of ​​the component, adapting to the differentiated heat dissipation requirements of different parts, and finally constructing a directional cooling flow field that meets the heat dissipation directionality.

[0050] For example, taking the cooling of a composite wing leading edge component with a deeply grooved structure as an example, this grooved area corresponds to four distributed fluid guiding units. When performing a cooling task, the system issues a task to one of the core guiding units based on the current heat dissipation demand intensity, with a preset base flow rate value. During execution, the unit's sensors provide real-time feedback on the jet pressure. The average jet pressure of surrounding adjacent units obtained through a local communication network. If the system's coupling coefficient is set... Substituting the values ​​into the formula for analysis, the final determined micro-airflow valve adjustment amount was obtained. Meanwhile, the controllable guide vanes of this unit rotate at 45 degrees, and through momentum convergence with the jets generated by adjacent units at the center of the groove, a powerful directional cooling cyclone is formed targeting the bottom of the groove. This precise flow correction and angle coordination provides additional heat transfer compensation to the groove area, which was originally slow to cool due to airflow obstruction, successfully controlling the temperature deviation in this area within the allowable range of the process.

[0051] Optionally, after the self-organization forms the directional cooling flow field, it further includes: The current configuration parameters of the directional cooling flow field are collected and the thermal property parameters reflecting the thermal conductivity of the material are obtained. Thermal state prediction parameters describing the evolution of the temperature field in the future time period are generated by deduction. By comparing the predicted local temperature difference in the thermal state prediction parameters with the gradient safety threshold set internally by the system, target micro-regions where the predicted gradient is at risk of exceeding the limit are identified. A preventative flow field compensation command is generated for the target micro-region, and this command is inserted into the current cycle's micro-region control command set to trigger real-time pre-adjustment of the directional cooling flow field.

[0052] Specifically, after constructing the directional cooling flow field using distributed fluid guiding units, the current configuration parameters of the directional cooling flow field are further collected, including the jet angle and real-time flow rate of each guiding unit. Thermal property parameters reflecting the material's thermal conductivity, such as thermal conductivity, are also obtained from a pre-set database. Based on these real-time and inherent parameters, a thermodynamic extrapolation algorithm generates thermal state prediction parameters describing the temperature field evolution over a future time period. These parameters quantify the temperature distribution trend of each dynamic virtual micro-region after a preset time step. The thermal state prediction cloud map is shown below. Figure 3 As shown, the thermodynamic deduction algorithm employs a hybrid architecture of physical modeling and data-driven approach. It forms a closed loop with four main modules: preprocessing, core deduction, calibration and optimization, and feedback storage. Inputs include static fundamental parameters such as tank geometry and material thermodynamic parameters, as well as dynamic real-time data such as distributed temperature sensing and cooling system operating parameters, and feedback calibration data. Through core calculations such as spatial discretization, thermal equilibrium iteration, and temporal evolution, the algorithm outputs a three-dimensional thermodynamic digital image centered on the three-dimensional temperature distribution field within the tank and the temporal evolution of the thermal field. Simultaneously, it outputs auxiliary data such as deduction residuals and hardware control parameter suggestions. Training is divided into two stages: offline initialization training and online iterative training. First, the basic parameters of the physical model are calibrated and the mesh generation is optimized based on historical process data. Then, deduction deviations are calculated through real-time measured feedback, dynamically correcting core parameters such as the heat transfer coefficient. Combined with historical deviation analysis, the deduction logic is optimized to continuously reduce deduction errors and improve algorithm accuracy and adaptability.

[0053] The predicted local temperature difference in the thermal state prediction parameters is compared with the gradient safety threshold set internally by the system. If the predicted temperature difference exceeds the threshold, a target micro-region with a risk of exceeding the predicted gradient limit is identified. To eliminate this potential risk, a preventative flow field compensation command is generated for the target micro-region. This command enhances the heat transfer capability of the overheated parts by modifying the bias parameters of the guiding unit.

[0054] The flow field compensation command is inserted into the current cycle's micro-area control command set, triggering real-time pre-adjustment of the directional cooling flow field, thereby achieving a shift from passive response to proactive prevention. During the prediction process, a thermal gradient prediction and evaluation formula is used to assist decision-making; the formula is expressed as: , in, The parameter representing the thermal state prediction is a dimensionless value used to characterize the risk level of future gradient exceeding limits. It refers to the predicted value of local temperature difference at future moments derived from the prediction model, which is obtained by linear extrapolation of historical temperature change rates. It refers to the allowable extreme value of the gradient set at a corresponding moment in the process flow. It is the thermal response hysteresis time constant of the material, which is determined by the specific heat capacity and thickness of the material. This is the current real-time cooling rate being monitored.

[0055] For example, taking the fabrication of a large-size composite material wing stringer component as an example, at a certain moment in the cooling section of this component, the system detects that the actual temperature difference in its thickened area meets the current requirements. However, by collecting the configuration parameters of the directional cooling flow field and performing thermal state extrapolation, it is found that the predicted local temperature difference in this area will be higher in the next 180 seconds. It will reach 12 degrees Celsius, while the system's set gradient safety threshold The temperature is 10 degrees Celsius. The real-time cooling rate measured at this point... The hysteresis time constant of the thermal response at this location is known to be 0.02 degrees Celsius per second. The time is 15 seconds. The thermal state prediction parameters are calculated. Since the value is greater than 1, the system determines that there is a clear risk of the predicted gradient exceeding the limit in this area, and then identifies it as the target micro-region. To address this risk, the system generates a flow field compensation command, increasing the flow direction offset of the corresponding guide unit in this area by 20%, and immediately inserts it into the micro-region control command set. After the actuator responds, by increasing the convection intensity of the cooling medium in advance, it successfully mitigates the upward trend of the temperature difference before the predicted exceedance time arrives, ensuring that the actual temperature difference remains within the safe threshold.

[0056] Optionally, the command for generating preventative flow field compensation includes: The three-dimensional temperature distribution cloud map in the thermal state prediction parameters is analyzed, and the grid node groups with temperatures higher than the average reference are identified as overheated sub-regions, and the grid node groups with temperatures lower than the average reference are identified as underheated sub-regions. Calculate the straight-line distance between the geometric centers of the superheated sub-region and the underheated sub-region and the predicted temperature difference span, and evaluate the adjustment intensity score required to achieve thermal equilibrium. The execution increment for heat transfer is calculated based on the adjustment intensity score, and the corresponding distributed fluid guiding units located on the heat transfer axis of the superheated sub-region and the underheated sub-region are driven to perform flow direction bias.

[0057] Specifically, when performing preventative flow field compensation, the three-dimensional temperature distribution cloud map in the thermal state prediction parameters is analyzed, and the grid node groups with temperatures higher than the current global average benchmark are identified as overheated sub-regions through spatial clustering algorithms, while the grid node groups with temperatures lower than the average benchmark are identified as underheated sub-regions. The spatial clustering algorithm adapted for identifying high-heat areas in autoclaves adopts a four-layer modular closed-loop structure: data preprocessing, core clustering, post-processing, and feedback optimization. The core is based on density clustering. The input includes three types of structured data: the three-dimensional spatial coordinate tensor of the thermal field grid points, the temperature / temperature gradient feature array of each grid point, and the algorithm control parameter dictionary such as neighborhood radius and minimum number of points. After a series of calculations such as coordinate normalization, noise filtering, density calculation, neighborhood search, cluster partitioning, and cluster merging, the output includes high-heat area clustering results, clustering evaluation indicators, and high-heat area priority ranking. The training adopts an unsupervised learning mode. First, it traverses parameter combinations based on historical thermal field data and completes offline initialization training to determine the initial parameters with the goal of maximizing the silhouette coefficient. Then, it compares the clustering results with the high-heat areas verified by actual processes to calculate the positioning deviation. The control parameters are adjusted by gradient descent, false clusters are removed, and adjacent small clusters are merged to achieve online iterative optimization and continuously improve the accuracy of high-heat area identification.

[0058] A local thermal equilibrium coordinate system is established. The linear distance between the geometric centers of the superheated and underheated sub-regions is calculated using geometric analytical methods, and the predicted temperature difference span between them is extracted. To eliminate this potential non-uniformity, the regulation intensity score required to achieve thermal equilibrium is evaluated, and the execution increment for heat transfer is calculated based on this score. Execution Increment The calculation formula is: , In this formula, The jet intensity of the distributed fluid guiding unit is incremented. It is a preset proportionality coefficient used to convert the temperature gradient into a flow rate increment. The predicted temperature difference span between the superheated and underheated sub-regions is obtained by analyzing the three-dimensional temperature distribution cloud map. The straight-line distance between the geometric centers of the two sub-regions is calculated using spatial coordinate vectors.

[0059] First, the adjustment intensity score for achieving thermal balance is evaluated by combining key data such as the geometric distance between the overheated and underheated sub-regions and the predicted temperature difference. Then, based on the score, the heat transfer execution increment, such as the jet intensity and angle adjustment range that the fluid guiding unit needs to increase, is calculated. Subsequently, the distributed fluid guiding units arranged on the heat transfer axis of the two sub-regions are precisely driven. By adjusting the rotation angle of their guide vanes and the jet direction, the flow direction is offset, allowing the cooling medium to flow directionally along the heat transfer axis. This forces the excess heat in the overheated sub-region to be guided to the underheated sub-region. Through active heat spatial allocation, the thermal field gradient is smoothed in advance, avoiding component cooling defects caused by excessive temperature difference.

[0060] For example, suppose that during the cooling of a large composite wing panel in an autoclave, the thermal state prediction program identifies the intersection of stiffeners as an overheated sub-region, while the adjacent open area of ​​the skin is identified as an underheated sub-region. The system calculates the straight-line distance between the geometric centers of these two sub-regions. Meters, and predicts the span of temperature difference over a future period. Degrees Celsius. The system's preset scaling factor is known. Square meters per second. Substitute the values ​​into the formula for a calculation example analysis, and perform incremental calculations. At this moment, the system immediately locates the third distributed fluid guide unit on the heat transfer axis of these two regions, drives its guide vanes to deflect towards the superheated sub-region side, and increases the jet intensity by 6 units. This operation generates a diagonal cross-regional convective airflow, accelerating the heat accumulated at the reinforcing ribs towards the underheated area of ​​the skin, successfully achieving spatial heat mitigation through flow direction deflection before the temperature difference truly widens.

[0061] Optionally, the method further includes: During the cooling operation, the actual temperature data of each of the aforementioned dynamic virtual micro-regions is collected within a continuous monitoring period; The actual temperature change rate, which reflects the actual heat transfer intensity, is obtained by calculating the decrease in the actual temperature data. The actual temperature change rate is compared with the reference change rate extracted from the target cooling trajectory to calculate the rate deviation parameter; The intensity ratio coefficient in the micro-region control command set is corrected using the rate deviation parameter.

[0062] Specifically, during the cooling operation of the autoclave, for each previously divided dynamic virtual micro-zone, a continuous monitoring cycle is completed according to the fixed frequency set by the process. In each cycle, the actual temperature data of all monitoring points inside each micro-zone is continuously collected to form a complete temperature data sequence of each micro-zone in the continuous time dimension. This operation is the basis for subsequent evaluation of the actual cooling effect of each micro-zone and the calculation of the rate, ensuring that the data can accurately reflect the real cooling process and temperature change status of each micro-zone.

[0063] Based on the complete temperature data collected by each dynamic virtual micro-region during the continuous monitoring cycle, the total temperature drop of each micro-region within that cycle is first calculated, which is the difference between the average temperature of the micro-region at the beginning of the cycle and the average temperature of the micro-region at the end of the cycle. Then, this temperature drop is ratioed to the duration of the monitoring cycle to obtain the actual temperature change rate of each micro-region. This rate is a core indicator reflecting the actual heat exchange intensity of each micro-region. The higher the rate value, the stronger the cooling and heat exchange effect of the micro-region, providing a quantitative basis for judging whether the heat dissipation status of the micro-region meets the process expectations.

[0064] From a pre-set target cooling trajectory that matches the process, the ideal cooling rate, i.e., the reference rate of change, should be achieved by each dynamic virtual micro-region within the corresponding monitoring period. Then, the actual calculated temperature change rate of each micro-region is precisely compared with this reference rate of change. The rate deviation parameter is finally calculated by calculating the difference or ratio between the two. This parameter quantifies the degree of deviation between the actual cooling rate of each micro-region and the ideal cooling rate of the process, clearly reflecting the gap between the actual cooling process and the target requirements. It provides a clear numerical reference for subsequent correction of micro-region control commands and adjustment of cooling flow field configuration.

[0065] To eliminate thermal hysteresis or overcooling during the cooling process, the system uses the rate deviation parameter to correct the intensity proportionality coefficient in the micro-area control command set. The correction formula for the intensity proportionality coefficient is as follows: , in, This represents the modified intensity scaling factor, used to adjust the jet intensity increment of the distributed fluid guide unit; This represents the original strength scaling factor before correction; This represents the reference rate of change extracted from the target cooling trajectory, reflecting the ideal cooling rate required by the process. This represents the actual rate of temperature change calculated from real-time monitoring data.

[0066] For example, taking the cooling of a composite S-shaped air intake component with variable thickness characteristics as an example, the component is divided into multiple dynamic virtual micro-regions. During a certain minute of the cooling phase, the system monitors the actual temperature of a thick-walled region decreasing from 150 degrees Celsius to 147 degrees Celsius, and calculates the actual temperature change rate. The rate is 3 degrees Celsius per minute. At this point, by querying the target cooling trajectory, the required reference rate of change for that period can be determined. The temperature is 4 degrees Celsius per minute. The initial intensity scaling factor for this micro-region is known. The value is 1.0. Substituting this into the formula, the corrected strength proportionality coefficient is obtained. The system then responds to the micro-area control command set, scheduling the distributed fluid guide unit directed to that area and increasing its micro-airflow valve adjustment and jet intensity by a gain of 1.25. Through this precise rate correction, the thick-walled area, which was originally lagging in cooling, receives stronger directional cooling flow field support, allowing its cooling rate to quickly catch up with the target trajectory and effectively avoiding geometric deformation of the component caused by local heat accumulation.

[0067] Optionally, the method further includes: After the cooling process is started, the cooling process is divided into a first stage for breaking thermal stratification, a second stage for gradient control, and a third stage for stress release, based on the initial temperature field distribution. Set corresponding target temperature ranges and spatial temperature difference limits for each stage; Obtain the status bit signal of the current process stage, load the corresponding control strategy weights accordingly, and drive the cyclic execution under different stage constraints.

[0068] Specifically, after initiating the autoclave cooling process, based on the initial temperature field distribution, the entire cooling process is divided into a first stage for breaking thermal stratification, a second stage for gradient control, and a third stage for stress release. The first stage focuses on eliminating the temperature difference between the upper and lower layers caused by static thermal convection within the autoclave, establishing a uniform initial thermal environment through high-intensity flow field disturbance. The second stage addresses the thermal hysteresis characteristics of the composite material during cooling, focusing on controlling the cooling gradient of different parts of the component through a directional flow field. The third stage, as the material approaches its glass transition temperature, releases internal residual stress by slowing the cooling rate.

[0069] In the multi-stage cooling control of autoclaves based on flow direction switching, three cooling stages are set: breaking thermal stratification, gradient control, and stress release. Each stage has a dedicated target temperature range and spatial temperature difference limit, which serve as the temperature control benchmark and uniform temperature constraint for each stage. The indicators for each stage are adjusted according to the process focus. Based on this, the system matches the control strategy and regulates the flow field to achieve refined cooling and ensure the quality of component molding.

[0070] Based on this signal, the corresponding control strategy weights are automatically loaded, driving each module to execute cyclically under different stage constraints. To accurately determine the control strength at each stage in complex flow field environments, a stage control gain calculation formula is introduced: , In this formula, This represents the phased control gain, used to correct the heat dissipation intensity weight of each micro-region. The characteristic coefficients representing the preset stages are obtained by experimentally measuring the fluid coupling sensitivity at different cooling stages. This represents the current real-time average temperature. This represents the space temperature difference limit set at the current stage. This represents the target temperature range corresponding to the current stage.

[0071] For example, taking the cooling of a large composite wing sparsity component as an example, in the initial stage of cooling startup, due to the rising of internal hot air, the initial temperature difference between the top and bottom of the tank reaches 15 degrees Celsius. After the system recognizes this state, it automatically enters the first stage for breaking thermal stratification. At this time, the status signal acquired by the system drives the control strategy to tilt towards global disturbances, setting a temperature difference limit. The target temperature range is 5 degrees Celsius. The temperature is 50 degrees Celsius. This is based on the currently monitored real-time average temperature. The temperature was 175 degrees Celsius, and the characteristic coefficients for this stage were experimentally measured. The control gain is calculated to be 1.2. At this point, the flow direction switching module will, based on this high gain value, schedule the distributed fluid guiding units across the entire domain to perform maximum angle switching and jet output, quickly eliminating thermal stratification. Once the overall temperature difference converges to within 5 degrees Celsius, the system automatically switches to the second stage state, where the temperature difference limit is... The temperature was updated to 3 degrees Celsius, and the control strategy weights were adjusted accordingly to focus on directional cooling of the thick-walled and thin-walled regions of the component. Through this phased weighting and real-time gain adjustment, the system can ensure that the component remains within the optimal flow field throughout the entire cooling cycle, avoiding the failure problem of a single control logic under complex process spans.

[0072] Based on the same inventive concept, such as Figure 4 As shown, the present invention also provides a multi-stage cooling control system for autoclaves based on flow direction switching, the system comprising: The data acquisition and 3D geometric feature extraction module is used to acquire real-time temperature data of each monitoring point inside the autoclave and 3D model data of composite material components, and extract the 3D geometric features from the 3D model data. The dynamic virtual micro-area division module is used to divide the internal space of the autoclave corresponding to different heat dissipation parts of the component into multiple dynamic virtual micro-areas based on the real-time temperature data and the three-dimensional geometric features, using a preset spatial mapping algorithm. The control instruction set generation module is used to obtain the average thermal distribution of the dynamic virtual micro-region and the target cooling trajectory that matches the process, and generate a micro-region control instruction set containing control priority and heat dissipation demand intensity through comparison calculation; The flow direction switching and control execution module is used to respond to the micro-area regulation instruction set, schedule multiple distributed fluid guiding units with independent rotation angles and openings in the autoclave, and form a directional cooling flow field through local aerodynamic coupling self-organization.

[0073] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0074] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A multi-stage cooling control method for autoclaves based on flow direction switching, characterized in that, The method includes: Real-time temperature data and three-dimensional model data of composite material components are obtained from each monitoring point inside the autoclave, and three-dimensional geometric features are extracted from the three-dimensional model data. Based on the real-time temperature data and the three-dimensional geometric features, a preset spatial mapping algorithm is used to divide the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component. The average thermal distribution of the dynamic virtual micro-region and the target cooling trajectory that matches the process are obtained. A micro-region control instruction set containing control priority and heat dissipation demand intensity is generated through comparison calculation. In response to the micro-area control command set, multiple distributed fluid guiding units with independent rotation angles and openings within the autoclave are scheduled to form a directional cooling flow field through local aerodynamic coupling self-organization.

2. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The division of the internal space of the autoclave into multiple dynamic virtual micro-regions corresponding to different heat dissipation parts of the component includes: Identify the cross-sectional thickness span and surface curvature changes in the three-dimensional geometric features, and analyze the physical entity of the composite material component into multiple component sub-regions containing thick-walled regions, thin-walled regions, and connection regions; Based on the spatial coordinates of the component sub-region, airflow space nodes located within the range of the outer normal direction are retrieved, a mapping relationship between physical entities and airflow space is established, and an initial virtual micro-region is generated; The temperature difference and thermal conductivity ratio between adjacent initial virtual micro-regions are obtained, and a collaborative evaluation index reflecting the thermodynamic synchronicity between regions is calculated. By comparing the collaborative evaluation index with the preset fusion threshold, and by reorganizing the spatial topology of adjacent initial virtual micro-regions, a dynamic virtual micro-region whose boundary evolves dynamically with the temperature gradient is output.

3. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The generation of the micro-area control instruction set, which includes control priority and heat dissipation demand intensity, includes: Extract the desired temperature value corresponding to the current moment from the target cooling trajectory, calculate the difference between the real-time average temperature of the dynamic virtual micro-region and the desired temperature value, and obtain the first temperature difference parameter; The instantaneous temperature extreme values ​​of all monitoring points within the dynamic virtual micro-region are retrieved to obtain the second temperature difference parameter characterizing the gradient within the micro-region; The material density and specific heat capacity parameters associated with the three-dimensional geometric features are obtained, and a normalized weighted calculation is performed by combining the first temperature difference parameter and the second temperature difference parameter to generate a comprehensive control score. Based on the descending order of the comprehensive control scores, corresponding execution order fields are assigned and aggregated into a micro-area control instruction set.

4. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The formation of a directional cooling flow field through local aerodynamic coupling self-organization includes: Receive the micro-area control instruction set and transform the control instructions for the macro-area into flow field sub-tasks for specific guiding nodes; The flow field subtask is sent to the corresponding distributed fluid guiding unit to obtain the real-time jet pressure and guide vane attitude angle fed back by adjacent units, and to perform local communication and anti-interference coordination. Based on the feedback results of the local communication and anti-interference coordination, adjust the rotation angle of the controllable guide vanes and the adjustment amount of the micro-airflow valves inside each unit; By utilizing the convergence effect of the air jets discharged from each distributed fluid guiding unit in the tank space, a directional cooling flow field that meets the heat dissipation directionality requirements is constructed.

5. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The self-organized formation of the directional cooling flow field also includes: The current configuration parameters of the directional cooling flow field are collected and the thermal property parameters reflecting the thermal conductivity of the material are obtained. Thermal state prediction parameters describing the evolution of the temperature field in the future time period are generated by deduction. By comparing the predicted local temperature difference in the thermal state prediction parameters with the gradient safety threshold set internally by the system, target micro-regions where the predicted gradient is at risk of exceeding the limit are identified. A preventative flow field compensation command is generated for the target micro-region, and this command is inserted into the current cycle's micro-region control command set to trigger real-time pre-adjustment of the directional cooling flow field.

6. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 5, characterized in that, The flow field compensation command with preventative properties includes: The three-dimensional temperature distribution cloud map in the thermal state prediction parameters is analyzed, and the grid node groups with temperatures higher than the average reference are identified as overheated sub-regions, and the grid node groups with temperatures lower than the average reference are identified as underheated sub-regions. Calculate the straight-line distance between the geometric centers of the superheated sub-region and the underheated sub-region and the predicted temperature difference span, and evaluate the adjustment intensity score required to achieve thermal equilibrium. The execution increment for heat transfer is calculated based on the adjustment intensity score, and the corresponding distributed fluid guiding unit located on the heat transfer axis of the superheated sub-region and the underheated sub-region is driven to perform flow direction bias.

7. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The method further includes: During the cooling operation, the actual temperature data of each of the aforementioned dynamic virtual micro-regions is collected within a continuous monitoring period; The actual temperature change rate, which reflects the actual heat transfer intensity, is obtained by calculating the decrease in the actual temperature data. The actual temperature change rate is compared with the reference change rate extracted from the target cooling trajectory to calculate the rate deviation parameter; The intensity ratio coefficient in the micro-region control command set is corrected using the rate deviation parameter.

8. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The method further includes: After the cooling process is started, the cooling process is divided into a first stage for breaking thermal stratification, a second stage for gradient control, and a third stage for stress release, based on the initial temperature field distribution. Set corresponding target temperature ranges and spatial temperature difference limits for each stage; Obtain the status bit signal of the current process stage, load the corresponding control strategy weights accordingly, and drive the cyclic execution under different stage constraints.

9. The multi-stage cooling control method for autoclaves based on flow direction switching according to claim 1, characterized in that, The process of acquiring real-time temperature data from various monitoring points inside the autoclave and three-dimensional model data of the composite material components, and extracting the three-dimensional geometric features from the three-dimensional model data, includes: Real-time temperature data from all monitoring points inside the tank is acquired, and the collected data is denoised and interpolated to generate a spatiotemporally continuous temperature field dataset. The composite material component is subjected to lightweight topology reconstruction and geometric feature extraction, redundant surfaces and invalid control points are removed, and three-dimensional geometric features are generated. The temperature field dataset is registered with the three-dimensional geometric features in a spatial coordinate system to extract the three-dimensional geometric features of the temperature attributes.

10. A multi-stage cooling control system for autoclaves based on flow direction switching, applied to the multi-stage cooling control method for autoclaves based on flow direction switching as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition and 3D geometric feature extraction module is used to acquire real-time temperature data of each monitoring point inside the autoclave and 3D model data of composite material components, and extract the 3D geometric features from the 3D model data. The dynamic virtual micro-area division module is used to divide the internal space of the autoclave corresponding to different heat dissipation parts of the component into multiple dynamic virtual micro-areas based on the real-time temperature data and the three-dimensional geometric features, using a preset spatial mapping algorithm. The control instruction set generation module is used to obtain the average thermal distribution of the dynamic virtual micro-region and the target cooling trajectory that matches the process, and generate a micro-region control instruction set containing control priority and heat dissipation demand intensity through comparison calculation; The flow direction switching and control execution module is used to respond to the micro-area regulation instruction set, schedule multiple distributed fluid guiding units with independent rotation angles and openings in the autoclave, and form a directional cooling flow field through local aerodynamic coupling self-organization.

Images