Method for preparing machine learning assisted functional polyimide composite of Janus structure

Abstract

The application belongs to the technical field of polymer material manufacturing, and discloses a preparation method of a functional polyimide composite material with a Janus structure assisted by machine learning, which comprises the following steps: S1, database establishment; S2, data preprocessing; S3, basic structure design and preparation process; S4, machine learning model construction and training; S5, model testing and evaluation; S6, feature importance analysis; S7, composite material optimization design; and S8, experimental verification. By using a machine learning algorithm model, the application analyzes the influence of various factors on the performance of the functional polyimide composite material, evaluates the degree of the influence of key features on the performance, and obtains the composite material and its preparation process. The material has a functional layered structure, the surface layer serves as an impedance matching layer, the middle layer serves as an electromagnetic wave absorbing layer, and the base layer serves as a reflection layer. After hot pressing, the structure is completely encapsulated, the overall wave absorption efficiency and heat management performance of the material are improved, and the manufacturing requirements of various high-end electronic devices can be met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of high polymer material manufacturing, in particular to a machine learning assisted preparation method of functional polyimide composite material with Janus structure. BACKGROUND

[0002] Wave-absorbing materials have recently become the focus of key technology fields, particularly in improving the clarity of wireless signal transmission, reducing electromagnetic noise interference, and optimizing the electromagnetic compatibility of electronic devices. With the continuous advancement of technology, these materials are rapidly developing towards being lighter, thinner, and more efficient. Research and development of advanced wave-absorbing materials play a crucial role in ensuring the safety and reliability of information communication and in future technological innovation and application.

[0003] Current research on wave-absorbing materials often focuses on building efficient conductive networks. However, relying solely on highly conductive networks to manufacture electromagnetic shielding materials is not foolproof. For example, although most metals (such as copper, silver, gold) and their composite materials (including metal fabrics, metal foams, and metal laminates) effectively reflect electromagnetic radiation due to their ultra-high conductivity, they can also cause environmental secondary pollution. In contrast, carbon-based materials (such as graphene, carbon nanotubes, and carbon black) have become the dominant materials for electromagnetic wave absorption due to their good dielectric constant. However, using these materials alone often fails to meet the demand for wide frequency band and high strength absorption performance. In this context, the importance of studying thermal conductivity cannot be ignored. Effective thermal management is crucial to ensure the stability and prolong the life of electronic devices during high-performance operation. In particular, in dense electronic components and microchips, excellent thermal conductive materials can prevent heat accumulation, thereby avoiding performance degradation and device damage caused by overheating. Therefore, materials that simultaneously possess wave-absorbing and good thermal conductivity are particularly important in modern electronic and communication devices, as such materials not only absorb and shield unnecessary electromagnetic interference but also effectively manage heat flow, enhancing the overall performance and safety of the devices. However, due to the high complexity and uncertainty in the design of components, microstructure, and preparation process of materials with wave-absorbing and good thermal conductivity, which involves a wide range of factors such as materials, ratios, structures, temperature, pressure, time, and many other interrelated and complex changes, traditional design and experimental methods, as well as conventional AI technology, cannot achieve the prediction and optimization of functional polyimide composite materials with specific wave-absorbing and good thermal conductivity optimal combinations. SUMMARY

[0004] The present application is to solve the problems of the prior art, overcome the above-mentioned deficiencies existing in the prior art, and provide a preparation method of a machine learning assisted Janus structure functional polyimide composite material. The machine learning technology is used to construct a model capable of predicting the functional polyimide composite material. The model optimizes the design of the composite material and the preparation process based on the functional layered Janus structure by evaluating the degree of influence of key features on performance, and precisely controls the material properties and interaction of each layer to significantly improve the absorption efficiency of electromagnetic waves and enhance the thermal conductivity of the material, thereby improving the development efficiency and reducing the test cost. The present application provides a new technical concept for manufacturing a more efficient wave-absorbing and heat-conducting integrated material with specific composite performance combination requirements, to meet the needs of high-end and precision electronic equipment with dual requirements of electromagnetic compatibility and thermal management.

[0005] To achieve the above-mentioned purposes, the technical solution adopted by the present application is:

[0006] A preparation method of a machine learning assisted Janus structure functional polyimide composite material, comprising the following steps:

[0007] S1, establishing a database

[0008] Collecting information of known functional polyimide composite materials, including classification features, numerical features and preparation processes, and establishing a database;

[0009] S2. Data preprocessing

[0010] Convert the classification features into binary subcategory features (“0” and “1”), preprocess the data, and divide it into a training set and a test set according to the proportion;

[0011] S3: Designing a basic Janus structure and a preparation process

[0012] Designing a basic structure of the polyimide composite material to be prepared including a Janus surface layer, an intermediate layer and a base layer, and then designing the basic components, ratio and preparation process of the composite material. The process is to prepare each layer of the composite material and then hot press;

[0013] S4. Constructing and training a machine learning model

[0014] Selecting a machine learning algorithm, using the training set to construct a machine learning prediction model, taking the feature information of the composite material as an input parameter, evaluating the accuracy of the machine learning prediction model with an evaluation index, and selecting the best machine learning prediction model through cross-validation;

[0015] S5. Model testing and evaluation

[0016] The accuracy of the machine learning model is tested using the test set, and relevant evaluation indicators are used for performance evaluation. The importance of the features is analyzed and verified through the Pearson correlation coefficient to ensure the accuracy and reliability of the model;

[0017] S6. Feature importance analysis

[0018] The recursive feature elimination method is combined with the machine learning prediction model to effectively identify key feature variables. Based on the analysis results of the Pearson correlation coefficient, the key features are further adjusted and optimized, and the model parameters and feature selection are optimized to accurately adjust the model and ensure the highest prediction accuracy and improve the value of practical application. The model prediction results and key features are obtained;

[0019] S7. Optimization design of composite materials

[0020] Machine learning is used to assist in the optimization of composite material selection and structure design. Based on the model prediction results and key features obtained in the above steps, Janus structure polyimide composite materials with wave absorption and heat conduction performance and preparation processes are designed. According to the key features identified by the machine learning model, the Janus structure, component and ratio of each layer of the optimized composite material, and the preparation process are determined.

[0021] S8. Experimental verification

[0022] Janus structure composite material samples are prepared, and control samples with reverse features are prepared at the same time. Then, the wave absorption performance and heat conduction performance of the composite material samples are verified by comprehensive performance testing. Among them, the one that meets the design requirements is selected as the optimized functional polyimide composite material.

[0023] A functional polyimide composite material obtained by the aforementioned machine learning assisted design method, comprising: a surface layer, an intermediate layer and a base layer stacked in turn, wherein the surface layer uses a porous graphene / polyimide felt as an impedance matching layer to regulate the surface impedance characteristics of the material, so that a large amount of incident electromagnetic waves enter the material interior without being reflected; the intermediate layer uses a four-iron modified boron nitride coating as an electromagnetic wave absorption layer, four-iron causes eddy current loss, converts electromagnetic energy into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes the rapid dissipation of heat energy, thereby improving the overall wave absorption efficiency and heat management performance; the base layer introduces a carbonized graphene / carbon fiber felt layer as a reflection layer to fully reflect the unabsorbed electromagnetic waves into the absorption layer for multiple absorption; the wave absorption layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; finally, through hot pressing, the polyimide resin acts as an adhesive to encapsulate the designed Janus structure completely, obtaining a functional polyimide composite material with excellent electromagnetic wave management and efficient heat management.

[0024] The application provides a preparation method of a machine learning assisted Janus structure functional polyimide composite material.

[0025] 1. The functional polyimide composite material and design method provided by the application can cope with the challenge of difficult effective prediction and optimization design caused by complex changes of various elements of the material, an accurate prediction model is constructed and trained based on machine learning technology, especially an AI model such as an XGBoost algorithm and preprocessed data, machine learning understands and analyzes how various material parameter characteristics affect the final material performance, so that the composition and microstructure of the material are optimized, and finally an integrated wave-absorbing and heat-conducting material with specific performance combination requirements is intelligently and efficiently designed; this intelligent design process not only improves the development efficiency, but also reduces the cost of experiments, and provides a new technical concept for manufacturing an integrated wave-absorbing and heat-conducting material with better performance parameter combination and higher efficiency.

[0026] 2. The composite material and method provided by the application adopt machine learning technology to construct a model capable of predicting the functional polyimide composite material, analyze the importance of material characteristics through the accurate machine learning model obtained by pre-training, the model optimizes the design of the composite material and the preparation process based on the functional layered Janus structure by evaluating the degree of influence of key characteristics on performance, and significantly improves the absorption efficiency of electromagnetic waves and enhances the heat conduction performance of the material by accurately controlling the material characteristics and interaction of each layer; this intelligent material design process significantly shortens the research and development cycle, effectively improves the material research and development efficiency, shortens the research and development cycle, significantly saves the scientific research cost of design and experiment, and provides a new technical concept for developing a series of composite materials with specific performance combination for efficient electromagnetic wave management and heat management.

[0027] 3. The composite material and method provided by the application, the multi-layer microstructure of the composite material obtained by the model optimization design can achieve efficient electromagnetic wave absorption and shielding through the special design of different layers; the porous structure design of the surface layer reduces the reflection of electromagnetic waves and allows electromagnetic waves to enter the inside of the material, the absorption layer in the middle converts electromagnetic energy into heat energy through eddy current loss, and the conductive layer of the base layer enhances the electromagnetic shielding ability, and reflects the electromagnetic waves that are not absorbed back to the middle layer for reabsorption, which significantly improves the overall absorption and shielding efficiency of the composite material, and provides a new technical concept for industrializing the preparation of an integrated wave-absorbing and heat-conducting material with specific composite performance combination requirements and higher efficiency, and the composite material can be widely used in the manufacturing of high-end and precise electronic equipment (especially communication equipment) with electromagnetic compatibility and heat management dual requirements, and can ensure the safety and reliability of information communication.

[0028] 4The composite material and method provided by the application, by pouring polyimide resin in the wave-absorbing coating and the conductive layer, and firmly connecting different layers through the hot pressing technology, the overall structural stability and durability of the material are enhanced. This structural design not only ensures the effective blocking of electromagnetic waves, but also guarantees the mechanical strength and environmental adaptability of the material in long-term use.

[0029] 5The composite material and method provided by the application, by using a machine learning model to predict the optimal combination of material components, microstructure, and hot pressing process parameters, the time and resources required by the traditional trial-and-error method can be significantly reduced. Through the optimization design predicted by the model, the technical scheme after test verification can directly enter the production process, avoiding unnecessary material waste and time delay, thereby improving production efficiency and reducing cost. The processing technology of the polyimide composite obtained by the optimal process parameters is more perfect, the mechanical properties of the polyimide composite are better, and the service life is longer.

[0030] 6The composite material and method provided by the application have a highly customized model optimization, by using multiple machine learning models to fit and predict the relationship between the components, microstructure, and hot pressing process parameters of each layer of the polyimide composite and the tensile strength and corresponding thermal stability, the model with the optimal prediction performance can be identified. This not only improves the accuracy of the model, but also ensures the precise adjustment of the process parameters by selecting the most suitable model, maximizing the performance of the material.

[0031] 7Compared with the traditional physical experiment method, the use of machine learning model reduces the demand for chemicals and potential environmental pollution in experiments and tests; the application of this technology conforms to the development concept of green environmental protection and supports the sustainable development goal. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 is the prediction result graph of the machine learning prediction model of the embodiment of the application in the test set;

[0033] Figure 2 is the Pearson correlation coefficient heat map between the feature variables of the embodiment of the application;

[0034] Figure 3 is the flowchart of the preparation method of the functional polyimide composite material provided by the application;

[0035] Figure 4 is the reflection coefficient R of the embodiment 2 and the comparative example 1 of the application;

[0036] Figure 5 is the SER value of the different impedance matching layers of the application. DETAILED DESCRIPTION

[0037] In order to accurately set forth the object, technical scheme and advantages of the present application, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0038] The above scheme will be further described below in combination with the drawings and specific embodiments, and the equipment, model in each embodiment all adopt the prior art, and each material is a commercially available product.

[0039] Basic embodiment

[0040] The preparation method of the machine learning assisted Janus structure functional polyimide composite material provided in the embodiment comprises the following steps:

[0041] S1, establishing a database

[0042] Collect the information of known functional polyimide composite materials, including classification features, numerical features and preparation processes, and establish a database; specifically, a database is established by collecting a variety of known functional polyimide composite material related information, and the information and database include: the structure of the impedance matching layer (including porous structure and film structure), the filler of the impedance matching layer (carbon material and magnetic material), the wave absorber (including Fe3O4 and other wave absorbing materials), the stacking order of the layer (such as impedance matching layer-wave absorbing layer-conductive layer, etc.), the forming process (including freeze-drying, hot pressing and coating), the material structure, the filler of the conductive layer (such as carbon material and magnetic material), the thickness and the reflection coefficient R;

[0043] S2. Data preprocessing

[0044] The classification features are converted into binary sub-class features ("0" and "1"), the data is preprocessed, and the training set and the test set are divided according to the proportion; specifically, the classification features in step S1 are converted into binary sub-class features ("0" and "1"), wherein "0" indicates that the sample does not belong to the sub-class, and "1" indicates that the sample belongs to the sub-class; the conversion results are as follows: the structure of the impedance matching layer: porous structure (1), film structure (0); the filler of the impedance matching layer: carbon material (1) and magnetic material (0); wave absorber: Fe3O4 (1), other wave absorbing materials (0); the stacking order of the layer: impedance matching layer-wave absorbing layer-conductive layer (1), other stacking order (0); forming process: freeze-drying, hot pressing and coating; material structure: asymmetric (1), symmetric (0); filler of the conductive layer: carbon material (1), magnetic material (0), etc.;

[0045] S3: Designing basic Janus structure and preparation process

[0046] The design includes the basic structure of the polyimide composite material to be prepared, which is formed by a Janus structure of a surface layer, an intermediate layer and a base layer, and the basic components, ratio and preparation process of the composite material are designed. The process is: preparing each layer of the composite material, and then hot pressing;

[0047] S4. Construct and train machine learning model

[0048] Select a machine learning algorithm, use the training set to build a machine learning prediction model, use the feature information of the composite material as the input parameter, use the evaluation index to evaluate the accuracy of the machine learning prediction model, and use cross-validation to screen the best machine learning prediction model; The machine learning algorithm includes one or more of the following: extreme gradient boosting (XGBoost) method, ridge regression (Ridge Regresso) method, gradient boosting decision tree (GBDT) method and random forest regression method;

[0049] S5. Model testing and evaluation

[0050] The accuracy of the machine learning model is tested by using the test set, and the performance is evaluated by using the related evaluation index. The importance of the characteristics is analyzed and verified by the Pearson correlation coefficient, so as to ensure the accuracy and reliability of the model;

[0051] In machine learning, first, the ridge regression (Ridge Regressor) linear regression algorithm is used, and the correlation coefficient R and the root mean square error (RMSE) are used as the evaluation index of the model accuracy. On the training set, the regression coefficient between the reflectance R predicted by the optimal model obtained by the algorithm and the actual value is 0.74, and the root mean square error is 0.19. Apply this model to the test set, and the regression coefficient between the predicted value and the actual value is 0.67, and the root mean square error is 0.18, indicating that the model has a certain prediction ability on the test set, but there is room for improvement.

[0052] Subsequently, the extreme gradient boosting (XGBoost) algorithm is introduced. The application results of this method on the training set show that the regression coefficient between the reflectance R predicted by the model and the actual value is improved to 0.84, and the root mean square error is reduced to 0.11. The regression coefficient applied to the test set is 0.89, and the root mean square error is 0.12. These results show that the XGBoost model is superior to the ridge regression model in terms of accuracy and stability, and shows higher prediction accuracy, and is more suitable for predicting the reflectance R of the composite material, and assisting the design and development of the composite material.

[0053] S6. Feature importance analysis

[0054] The key feature variables are effectively identified by using recursive feature elimination method combined with machine learning prediction model, and the key features are further adjusted and the model parameters and feature selection are optimized based on the analysis results of Pearson correlation coefficient, so as to accurately adjust the model and ensure the highest prediction accuracy and improve the value of practical application, and the model prediction results and key features are obtained;

[0055] S7. Optimization design of composite material

[0056] The machine learning is used to assist the optimization of composite material selection and structure design, the Janus structure polyimide composite material and preparation process with wave absorption and heat conduction performance are designed based on the model prediction results and key features obtained in the above steps, and the key features identified by the machine learning model are used to determine the Janus structure, component and ratio of each layer of the optimized composite material and the preparation process;

[0057] In the optimization design of the functional layered Janus structure composite material, the dielectric properties and the difference in electrical conductivity of the surface layer, the intermediate layer and the base layer are adjusted to make each surface show different electromagnetic wave reflection characteristics, including the following steps:

[0058] S7-1 The porous graphene / polyimide felt is used as an impedance matching layer in the surface layer to adjust the surface impedance characteristics of the material, so that a large amount of incident electromagnetic waves enter the material interior without being reflected;

[0059] In the surface layer, the polyimide felt is not filled with resin to maintain its porous structure, so as to allow electromagnetic waves to enter; the intermediate layer and the base layer are filled with polyimide resin to prevent electromagnetic waves from penetrating;

[0060] S7-2 The four-iron oxide modified boron nitride coating is used as an electromagnetic wave absorbing layer in the intermediate layer, the four-iron oxide causes eddy current loss to convert electromagnetic energy into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes the rapid dissipation of heat energy, thereby improving the overall wave absorption efficiency and thermal management performance;

[0061] S7-3 The carbonized graphene / carbon fiber felt layer is introduced as a reflection layer in the base layer to fully reflect the electromagnetic waves that are not absorbed into the absorbing layer for multiple absorption;

[0062] S7-4 The wave absorbing layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; and through hot pressing, the polyimide resin is used as an adhesive to encapsulate the designed Janus structure completely, so as to obtain an optimized structure of the polyimide composite material with excellent electromagnetic wave management and efficient thermal management.

[0063] The preparation method of the optimized design includes the following steps:

[0064] A1. Preparation of the substrate layer as a reflective layer: uniformly coat the graphene solution on the surface of the polyimide (PI) felt, stack to prepare a multi-layer polyimide felt / graphene structure, place the multi-layer graphene / PI felt structure in a specially designed mold, apply vertical downward pressure to ensure tight bonding between the layers, and perform high-temperature carbonization treatment;

[0065] A2. Preparation of the intermediate layer as an absorbing layer: use ultrasonic-assisted liquid phase exfoliation method to exfoliate and modify boron nitride (BN) powder to prepare functionalized boron nitride nanosheets (f-BNNSs), uniformly mix f-BNNSs and iron oxide (Fe3O4) and coat on the surface of the carbonized PI felt.

[0066] A3. Preparation of the surface layer as an impedance matching layer: immerse the PI felt in a low concentration graphene solution and ultrasonically treat for half an hour to ensure uniform attachment of graphene particles on the surface of the felt;

[0067] A4. Heat pressing of each layer: prepare a PI solution by dissolving PI powder in an organic solvent, immerse the conductive layer coated with wave absorbers in the PI solution, and evaporate the solvent by programmed heating; use polyimide resin as an adhesive to connect the impedance matching layer and the cured absorbing layer-conductive layer by heat pressing, and the PI resin is further cured and forms a seamless connection during the heat pressing process, finally obtaining an optimized preparation method of functional polyimide composites.

[0068] S8. Experimental verification

[0069] Prepare Janus structure composite samples and simultaneously prepare control samples with reverse characteristics, then test the comprehensive performance of these samples to verify the wave absorption and thermal conductivity performance of the composite samples, and select those that meet the design requirements as the optimized functional polyimide composites.

[0070] The application discloses a functional polyimide composite material, which is obtained by using the machine learning aided design method, and comprises a surface layer, an intermediate layer and a base layer which are sequentially stacked, wherein the surface layer uses a porous graphene / polyimide felt as an impedance matching layer to regulate the impedance characteristics of the material surface, so that a large amount of incident electromagnetic waves enter the material interior and are not reflected; the intermediate layer uses a ferroferric oxide modified boron nitride coating as an electromagnetic wave absorbing layer, the ferroferric oxide causes eddy current loss, electromagnetic energy is converted into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes rapid dissipation of the heat energy, so that the overall wave absorbing efficiency and heat management performance are improved; the base layer introduces a carbonized graphene / carbon fiber felt layer as a reflection layer to fully reflect the electromagnetic waves which are not absorbed into the absorbing layer for multiple absorption; the wave absorbing layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; finally, heat pressing is performed, so that the polyimide resin serves as an adhesive to completely encapsulate the designed Janus structure, and the functional polyimide composite material with excellent electromagnetic wave management and high-efficiency heat management is obtained; the number of layers of the multilayer polyimide felt / graphene structure is 3-x layers of graphene / PI felt arranged alternately (x is greater than or equal to 5). In the electromagnetic shielding composite material, the thickness of the surface layer and the base layer is 500 microns to 3000 microns.

[0071] The application provides a functional polyimide composite material based on machine learning aided Janus structure design and a preparation method thereof. The application focuses on using an extreme gradient boosting (XGBoost) algorithm to analyze the influence of various factors on the performance of the functional polyimide composite material, and to evaluate the degree of influence of key features on the performance. The Janus structure composite material based on functional layering is designed and optimized. The surface layer uses a porous graphene / polyimide felt as an impedance matching layer to regulate the impedance characteristics of the material surface, so that a large amount of incident electromagnetic waves enter the material interior and are not reflected. The intermediate layer uses a ferroferric oxide modified boron nitride coating as an electromagnetic wave absorbing layer. The ferroferric oxide causes eddy current loss, electromagnetic energy is converted into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes rapid dissipation of the heat energy, so that the overall wave absorbing efficiency and heat management performance are improved. The base layer introduces a carbonized graphene / carbon fiber felt layer as a reflection layer to fully reflect the electromagnetic waves which are not absorbed into the absorbing layer for multiple absorption. The wave absorbing layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material. Heat pressing is performed, so that the polyimide resin serves as an adhesive to completely encapsulate the designed Janus structure, and the functional polyimide composite material with excellent electromagnetic wave management and high-efficiency heat management is obtained.

[0072] Embodiment 1

[0073] The preparation method of the machine learning assisted Janus structure functional polyimide composite material provided by the embodiment of the application is a specific application based on the above basic embodiment. A Janus structure based functional polyimide composite material and a preparation method thereof are designed to enable the composite material to finally have a specific performance combination. A composite material with different structures and electromagnetic response characteristics is used as a functional module to realize assembly according to the requirements of the target performance combination. The graphene / polyimide felt with a porous structure is used to adjust the surface impedance of the material and allow electromagnetic wave incidence. The ferroferric oxide / functionality boron nitride nanosheet coating is used as a microwave absorption layer to convert electromagnetic energy into heat energy through eddy current loss. The incorporation of the thermally conductive boron nitride helps to dissipate heat energy and improve the overall thermal conductivity. The carbonized graphene / carbon fiber felt is used as a conductive layer to ensure electromagnetic shielding performance and reflect electromagnetic waves that are not absorbed by the top layer and the middle layer for reabsorption. The wave-absorbing coating and the conductive layer are filled with polyimide resin to improve the stability of the material and prevent electromagnetic wave penetration. Finally, the different layers are connected using a polyimide resin as an adhesive through a hot pressing technology to obtain a functional polyimide composite material with a specific performance combination.

[0074] The preparation method of the machine learning assisted Janus structure functional polyimide composite material with a specific performance combination further uses the following technical solutions based on the basic embodiment.

[0075] S1. Design the Janus structure of the functional material with a specific performance combination: based on theory and preliminary data, propose a basic scheme of a three-layer polyimide composite material with integrated wave-absorbing performance and thermal conductivity performance and a Janus structure, including material components, proportions, layer structures, and basic preparation processes, and then use machine learning to assist in optimizing material selection, structure design, and preparation processes.

[0076] S2. Establish a database: according to the basic scheme of the three-layer polyimide composite material, collect known composite material related information to establish a database, which includes classification features and numerical features.

[0077] S3. Data preprocessing: convert the classification features into binary subcategory features (“0” and “1”), preprocess the data, and divide the data into a training set and a test set according to the proportion.

[0078] S4. Build and verify the machine learning model: select a machine learning algorithm, use the training set to build a machine learning prediction model, use the feature information of the composite material as input parameters, use evaluation indicators to evaluate the accuracy of the machine learning prediction model, and use cross-validation to select the best machine learning prediction model. The machine learning algorithm selects the XGBoost algorithm, such as Figure 1The regression coefficient between the predicted reflectivity R and the actual value obtained by the best model on the training set is 0.84, and the root mean square error is about 0.11. Using the trained prediction model to predict the reflectivity R in the test set, the regression coefficient between the model prediction value and the actual value reaches 0.89, and the root mean square error is about 0.12. Therefore, the XGBoost model has high accuracy and can be used for prediction of reflectivity R to assist the design and development of composite materials.

[0079] S5. Model testing and evaluation: The accuracy of the machine learning model is tested using the test set, and the performance is evaluated using relevant evaluation indicators. The importance of the features is analyzed and verified by the Pearson correlation coefficient to ensure the accuracy and reliability of the model;

[0080] S6. Feature importance analysis: The key feature variables are effectively identified by using recursive feature elimination method combined with machine learning prediction model. Based on the analysis results of the Pearson correlation coefficient, the key features are further adjusted and optimized model parameters and feature selection to accurately adjust the model and ensure the highest prediction accuracy and improve the value of practical application.

[0081] S7. Design of functional materials with Janus structure: Based on the model prediction results and key features obtained in the above steps, polyimide composite materials with optimized Janus structure with wave absorption and thermal conductivity performance are designed;

[0082] It includes: regulating the dielectric properties of the top layer, middle layer and bottom layer, and the difference in electrical conductivity makes each surface exhibit different electromagnetic wave reflection characteristics; the surface layer uses polyimide felt without filling resin, which maintains its inherent porous structure, allowing electromagnetic waves to enter; the middle layer and the base layer are filled with polyimide resin to prevent electromagnetic waves from penetrating.

[0083] S8. Experimental verification: According to the key features identified by the machine learning model, Janus structure composite material samples are prepared, and control samples with reverse characteristics are prepared at the same time. By testing the comprehensive performance of these samples, the wave absorption performance and thermal conductivity performance are verified.

[0084] The preparation method of the functional polyimide composite material with Janus structure after optimization includes the following steps:

[0085] A1. Preparation of reflection layer: uniformly coat graphene solution on the surface of polyimide (PI) felt, stack to prepare multi-layer polyimide felt / graphene structure, place the multi-layer graphene / PI felt structure in a specially designed mold, apply vertical downward pressure to ensure tight bonding between layers, and perform high-temperature carbonization treatment;

[0086] The number of layers of the multi-layer polyimide felt / graphene structure is 3-x layers of graphene / PI felt arranged alternately (x≥5), preferably 5-7 layers. With the increase of the number of alternating layers, the internal defects of the material also increase, reducing the effective absorption of energy.

[0087] A2. Preparation of the absorption layer: functionalized boron nitride nanosheets (f-BNNSs) are prepared by ultrasonic-assisted liquid phase exfoliation of boron nitride (BN) powder. f-BNNSs and Fe3O4 are uniformly mixed and coated on the surface of the carbonized PI felt.

[0088] A3. Preparation of the impedance matching layer: the PI felt is immersed in a low-concentration graphene solution and ultrasonically treated for half an hour to ensure uniform attachment of graphene particles on the surface of the felt material. The thickness of the surface layer and the base layer is 500 μm - 3000 μm. A thinner wave-absorbing material cannot provide sufficient dielectric loss and magnetic loss, and has lower mechanical strength. A thicker wave-absorbing material will increase the weight of the overall structure.

[0089] A4. Hot pressing to obtain a composite material: PI powder is dissolved in an organic solvent to prepare a PI solution. The coated conductive layer is immersed in the PI solution, and the solvent is evaporated by programmed heating. The impedance matching layer and the cured absorption layer-conductive layer are connected by hot pressing with polyimide resin as the adhesive. During the hot pressing process, the PI resin is further cured and forms a seamless connection.

[0090] The hot pressing process is applied at a pressure of 0-20 MPa, a temperature of 150℃-200℃, and a hot pressing time of 20min-60min.

[0091] The machine learning assisted Janus structure functional polyimide composite material and the preparation method thereof provided by the embodiment of the application, by using the extreme gradient boosting (XGBoost) algorithm, the influence of various factors on the performance of the functional polyimide composite material is analyzed, and the degree of influence of key characteristics on the performance is evaluated, a Janus structure based on functional layering is designed, the surface impedance characteristics of the material are regulated by using the porous graphene / polyimide felt as an impedance matching layer on the surface layer, so that a large amount of incident electromagnetic waves enter the material interior without being reflected; the middle layer adopts the ferroferric oxide modified boron nitride coating as an electromagnetic wave absorption layer, the ferroferric oxide causes eddy current loss, converts electromagnetic energy into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes the rapid dissipation of heat energy, thereby improving the overall wave absorption efficiency and heat management performance; the substrate layer introduces the carbonized graphene / carbon fiber felt layer as a reflection layer, fully reflects the electromagnetic waves that are not absorbed into the absorption layer for multiple absorption; the wave absorption layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; and through hot pressing, the polyimide resin serves as an adhesive to encapsulate the designed Janus structure completely, so as to obtain a polyimide composite material with excellent electromagnetic wave management and high-efficiency heat management.

[0092] The vertical thermal conductivity of the composite material sample prepared in the embodiment is 0.47 Wm - 1 K -1 ; and the reflection coefficient R thereof obtained by testing is 0.49.

[0093] Embodiment 2

[0094] Referring to Figure 3 , the flowchart of the functional polyimide-based composite material preparation process provided by the embodiment of the application is shown.

[0095] The machine learning assisted Janus structure functional polyimide composite material preparation method provided by the embodiment of the application is a specific application based on the above basic embodiment, which is basically the same as embodiment 1, and the difference lies in that:

[0096] Firstly, based on the theory and preliminary data, a basic structure and preparation method of Janus structure polyimide composite material integrating wave-absorbing performance and heat-conducting performance are conceived and proposed, and then machine learning is used to assist optimization of material selection, structure design and preparation process. By collecting known information about composite materials to establish a database, the database contains the following information: structure of impedance matching layer (including porous structure and film structure), filler of impedance matching layer (carbon material and magnetic material), wave-absorbing agent (including Fe3O4 and other wave-absorbing materials), stacking order of layers (such as impedance matching layer-wave absorbing layer-conductive layer, etc.), forming process (including freeze-drying, hot pressing and coating), material structure, filler of conductive layer (such as carbon material and magnetic material), thickness and reflection coefficient R; some of these features are categorical features and some are numerical features, the categorical features are converted into binary sub-category features (“0” and “1”), where “0” means that the sample does not belong to this sub-category, and “1” means that the sample belongs to this sub-category. The conversion results are as follows: structure of impedance matching layer: porous structure (1), film structure (0); filler of impedance matching layer: carbon material (1) and magnetic material (0); wave-absorbing agent: Fe3O4 (1), other wave-absorbing materials (0); stacking order of layers: impedance matching layer-wave absorbing layer-conductive layer (1), other stacking orders (0); forming process: freeze-drying, hot pressing and coating; material structure: asymmetric (1), symmetric (0); filler of conductive layer: carbon material (1), magnetic material (0). The data set is shuffled randomly by pre-processing, and then divided into training set and test set in the ratio of 8:2, the training set is used to train the machine learning model, and the test set is used to evaluate the generalization ability of the trained model.

[0097] The XGBoost algorithm is selected in this embodiment, and the correlation coefficient R and the root mean square error RMSE are used as evaluation indexes to judge the precision of the model. The regression coefficient between the reflection coefficient R predicted by the best model obtained on the training set and the actual value is 0.84, and the root mean square error is about 0.11. The trained prediction model is used to predict the reflection coefficient R in the test set, and the regression coefficient between the model prediction value and the actual value reaches 0.89, and the root mean square error is about 0.12. Therefore, the XGBoost model has high accuracy and can be used for prediction of the reflection coefficient R to assist the design and development of composite materials.

[0098] The calculation results of the Pearson correlation coefficients between the 11 feature variables are shown in Table 1. Figure 2 The results show that the absolute value of the Pearson correlation coefficient between any two feature variables is less than 0.7, so there is no multicollinearity phenomenon between the feature variables. Since a lower reflection coefficient R indicates better wave-absorbing performance, a higher negative correlation between a feature and the reflection coefficient R indicates that the feature is more critical to the wave-absorbing performance.

[0099] Six key feature variables were identified by recursively eliminating nested XGBoost models: impedance matching layer structure, impedance matching layer filler, wave-absorbing agent, layer stacking order, forming process, and thickness. According to the Pearson correlation coefficient heat map and the six key feature variables, the following functional materials were designed: impedance matching layer structure: porous structure; impedance matching layer filler: carbon material; wave-absorbing agent selection: Fe3O4; layer stacking order: impedance matching layer-wave-absorbing layer-conductive layer; forming process: coating and hot pressing; thickness: 3-6 mm. Experimental verification was performed on the designed materials, and the optimized preparation method is as follows:

[0100] (1) Soak the polyimide (PI) felt in the prepared acetone-ethanol mixed solution, and ultrasonic treatment for 2h to remove the surface attachments of the fibers, and rinse with deionized water until no foam is generated. Cut two pieces of PI felt into a size of 20mm x 60mm. Prepare a graphene concentration of 100mg / ml and a carboxylated cellulose nanofiber concentration of 10mg / ml according to the mass ratio of graphene to carboxylated cellulose nanofiber of 10:1, and ultrasonic for 0.5 hours. Uniformly coat the mixed solution on the surface of the PI felt. To ensure the uniformity and good adhesion of the coating, the coated PI felt sample is placed in an oven at 75°C for drying for 12 hours to obtain a single-layer graphene / PI felt preform. Place a layer of PI felt on the surface of the single-layer graphene / PI felt preform and fix it in a specific compression mold. Apply vertical pressure to compress the multi-layer PI felt structure together, and perform carbonization treatment at 1400°C in an argon environment. The temperature is raised to 800°C at a rate of 10°C / min, to 1100°C at a rate of 8°C / min, and to 1400°C at a rate of 5°C / min, and held for 2h. The temperature is lowered to 1000°C at a rate of 5°C / min, and to 800°C at a rate of 8°C / min, and the sample is cooled with the furnace to obtain a three-layer structure of the conductive layer.

[0101] (2) Use ultrasonic-assisted liquid phase exfoliation method to exfoliate and modify h-BN powder to prepare f-BNNSs. Briefly, first add 2g of h-BN powder to a mixture of deionized water and isopropyl alcohol, and ultrasonic treatment for 6h followed by centrifugation to obtain f-BNNSs. Add Fe3O4 at a ratio of 1:1, and then uniformly coat it on the surface of the conductive layer to obtain a wave-absorbing coating.

[0102] (3) Take dry PI powder, add polar organic solvent N,N-dimethylacetamide (DMAC), and prepare a PI solution with a mass fraction of 8wt%, and ultrasonic treatment for 0.5h to ensure complete and uniform dissolution of the PI powder. Soak the conductive layer / wave-absorbing layer in the solution, and move it into the oven under vacuum for 12h to remove bubbles, and then program the temperature of the vacuum oven to evaporate the solvent.

[0103] (4) The PI felt is soaked in a graphene solution with a concentration of 20 mg / ml, and ultrasonic treatment is performed for 0.5 h to ensure that the graphene is uniformly dispersed in the felt, and the impedance matching layer is obtained by drying.

[0104] (5) The aforementioned impedance matching layer and the conductive layer / absorbing layer are compounded by a hot pressing process using polyimide resin as an adhesive to form a structurally complete multifunctional composite material. The hot pressing process is performed at a pressure of 5 MPa for a duration of 30 min, and the hot pressing temperature is set to 165°C.

[0105] The composite material sample prepared in this example is tested for its thermal conductivity on a thermal conductivity tester, and the vertical thermal conductivity of the composite material is calculated to be 0.55 Wm -1 K -1 ; and its reflection coefficient R obtained by testing is 0.47.

[0106] Comparative Example 1

[0107] According to the same preparation procedure as in Example 2, only the filler of the impedance matching layer in the key characteristic variable is changed to no graphene, and the composite material is prepared as follows:

[0108] The machine learning selection of key characteristic variables is unchanged, and the experimental verification step is as follows:

[0109] (1) The polyimide (PI) felt is soaked in the prepared acetone-ethanol mixed solution, and ultrasonic treatment is performed for 2 h to remove the surface attachments of the fibers, and then the PI felt is washed with deionized water until no foam is generated. Two pieces of PI felt are cut into a size of 20 mm x 60 mm. A graphene concentration of 100 mg / ml and a carboxylated cellulose nanofiber concentration of 10 mg / ml are prepared according to a mass ratio of graphene to carboxylated cellulose nanofiber of 10:1, and ultrasonic treatment is performed for 0.5 h. The mixed solution is uniformly coated on the surface of the PI felt. In order to ensure the uniformity and good adhesion of the coating, the coated PI felt sample is placed in an oven at 75°C for drying for 12 hours to obtain a single-layer graphene / PI felt preform. A layer of PI felt is placed on the surface of the single-layer graphene / PI felt preform, and fixed in a specific compression mold. The structure of the multi-layer PI felt is compressed together by applying vertical pressure, and carbonization treatment is performed in an argon environment at 1400°C, with a heating rate of 10°C / min to 800°C; a heating rate of 8°C / min to 1100°C; a heating rate of 5°C / min to 1400°C, and a holding time of 2 h; a cooling rate of 5°C / min to 1000°C; and a cooling rate of 8°C / min to 800°C. The sample is cooled with the furnace to obtain a three-layer conductive layer.

[0110] (2) The h-BN powder was exfoliated and modified to prepare f-BNNSs by using ultrasonic-assisted liquid exfoliation method. Briefly, 2 g of h-BN powder was added to a mixed solvent of deionized water and isopropyl alcohol, and after ultrasonic treatment for 6 h, centrifugation was performed to obtain f-BNNSs. Fe3O4 was added at a ratio of 1:1, and then uniformly coated on the surface of the conductive layer to obtain the wave-absorbing layer.

[0111] (3) Dry PI powder was taken and added to a polar organic solvent N,N-dimethylacetamide (DMAC) to prepare a PI solution with a mass fraction of 8wt%, and ultrasonic treatment was performed for 0.5 h to ensure complete and uniform dissolution of the PI powder. The conductive layer / wave-absorbing layer was immersed in the solution and moved into an oven under vacuum for 12 h to remove air bubbles, and then the vacuum oven program was raised in temperature to evaporate the solvent.

[0112] (4) The PI felt was used as an impedance matching layer.

[0113] (5) The impedance matching layer was combined with the conductive layer / wave-absorbing layer by using polyimide resin as an adhesive through a hot pressing process to form a structurally complete multifunctional composite material. The hot pressing process was carried out at a pressure of 5 MPa for a duration of 30 min, and the hot pressing temperature was set to 165°C.

[0114] Figure 4 The addition of carbon materials in the filler of the impedance matching layer was shown to improve the reflection coefficient R.

[0115] Comparative Example 2

[0116] According to the same preparation procedure as in Example 2, only the key feature variable of the impedance matching layer structure was changed to a non-porous structure and a composite material was prepared, as follows:

[0117] The machine learning selection of key feature variables step was unchanged, and the experimental verification step was as follows:

[0118] (1) The PI felt was immersed in the prepared acetone-ethanol mixed solution, and ultrasonic treatment was performed for 2 h to remove the surface attachments of the fibers, and then washed with deionized water until no foam was generated. Two pieces of PI felt were cut into a size of 20 mm x 60 mm. According to the mass ratio of graphene to carboxylated cellulose nanofiber of 10:1, the graphene concentration was 100 mg / ml, and the carboxylated cellulose nanofiber concentration was 10 mg / ml, and ultrasonic treatment was performed for 0.5 h. The mixed solution was uniformly coated on the surface of the PI felt. In order to ensure the uniformity and good adhesion of the coating, the coated PI felt sample was placed in an oven at 75°C for drying for 12 h to obtain a single-layer graphene / PI felt preform. Two single-layer graphene / PI felt preforms and one layer of PI felt were stacked and fixed in a special compression mold, and the multi-layer PI felt structure was compressed together by applying vertical pressure, and carbonization treatment was performed at 1400°C in an argon environment, and the temperature was raised to 800°C at a rate of 10°C / min; the temperature was raised to 1100°C at a rate of 8°C / min; the temperature was raised to 1400°C at a rate of 5°C / min, and the temperature was kept for 2 h; the temperature was lowered to 1000°C at a rate of 5°C / min; the temperature was lowered to 800°C at a rate of 8°C / min, and the sample was cooled with the furnace to obtain a five-layer conductive layer.

[0119] (2) The h-BN powder was peeled off and modified by ultrasonic-assisted liquid phase peeling method to prepare f-BNNSs. Briefly, first, 2 g of h-BN powder was added to a mixed solvent of deionized water and isopropyl alcohol, and ultrasonic treatment was performed for 6 h, and then centrifugation was performed to obtain f-BNNSs. Fe3O4 was added in a ratio of 1:1, and then uniformly coated on the surface of the conductive layer to obtain a wave-absorbing coating.

[0120] (3) The PI felt was immersed in a graphene solution with a concentration of 20 mg / ml, and ultrasonic treatment was performed for 0.5 h to ensure that the graphene was uniformly dispersed in the felt, and then dried to obtain an impedance matching layer.

[0121] (4) The dried polyimide (PI) powder was added to a polar organic solvent N,N-dimethylacetamide (DMAC) to prepare a PI solution with a mass fraction of 8wt%, and ultrasonic treatment was performed for 0.5 h to ensure that the PI powder was completely and uniformly dissolved. The conductive layer / wave-absorbing layer / impedance matching layer was immersed in the solution, and then moved into an oven for 12 h in a vacuum state to remove bubbles, and then the vacuum oven was programmed to evaporate the solvent.

[0122] By Figure 5 It can be seen that the SER value obtained by this method is higher, which is 9.8 dB, and compared with the PI felt without filling resin used in the impedance matching layer, the porous structure is retained due to the absence of filling resin, which promotes the penetration and internal multiple reflection of electromagnetic waves, and improves the absorption capacity of electromagnetic waves, so the SER value is reduced from 8 dB to 2.3 dB.

[0123] The preparation method of the machine learning assisted Janus structure functional polyimide composite material provided by each embodiment of the present application focuses on using machine learning technology to construct a model capable of predicting the functional polyimide composite material. The model optimizes the design of the composite material and the preparation process based on the functional layered Janus structure by evaluating the degree of influence of key features on performance, precisely controlling the material properties and interactions of each layer, significantly improving the absorption efficiency of electromagnetic waves, and enhancing the thermal conductivity of the material, thereby improving the development efficiency, reducing the test cost, and providing a new technical concept for manufacturing more efficient wave-absorbing and heat-conducting integrated materials with specific composite performance combination requirements.

[0124] The above embodiments of the present application use multiple machine learning models for the optimization design of the functional polyimide composite material and the preparation process by combining components, formulations, and preparation processes. The trained model is used to predict the change relationship between the wave-absorbing performance and the thermal conductivity performance combination effect of the functional polyimide composite material, obtain the optimal structure, components, ratio, and preparation process parameters of the composite material, solve the technical problems that cannot be solved by the prior art, and solve the problem of high complexity and uncertainty in the design of components, microstructure, and preparation process of materials with wave-absorbing and good thermal conductivity performance. The linkage relationship between the components, ratio, structure, temperature, pressure, time, and many other complex and changing factors of the composite material can be predicted, and the prediction and optimization of the functional polyimide composite material with specific wave-absorbing and good thermal conductivity performance combination can be achieved.

[0125] The above describes the embodiments of the present application in combination with the drawings, but the present application is not limited to the above embodiments, and can be changed in various ways according to the purpose of the present application. Any change, modification, replacement, combination, or simplification made in accordance with the spirit and principles of the present application should be an equivalent replacement method. As long as it meets the purpose of the present application and does not deviate from the technical principles and concepts of the present application, it belongs to the protection scope of the present application.

Claims

1. A method for preparing a machine learning assisted Janus structured functional polyimide composite material, characterized by, The method comprises the following steps: S1, establishing a database Collect information of known functional polyimide composites, including classification features, numerical features and preparation processes, and establish a database; S2. Data preprocessing Convert the classification features into binary subcategory features "0" and "1", preprocess the data, and divide it into training set and test set according to the proportion; S3: design basic Janus structure and preparation process Design the basic structure of the polyimide composite to be prepared including Janus surface layer, intermediate layer and substrate layer, and then design the basic components, ratio and preparation process of the composite, the process is: prepare each layer of the composite, and then hot press; S4. Build and train machine learning model Select a machine learning algorithm, use the training set to build a machine learning prediction model, use the feature information of the composite as input parameters, evaluate the accuracy of the machine learning prediction model with evaluation indicators, and use cross validation to select the best machine learning prediction model; S5. Model testing and evaluation Test the accuracy of the machine learning model using the test set, and evaluate the performance using relevant evaluation indicators; analyze and verify the importance of the features through Pearson correlation coefficient analysis to ensure the accuracy and reliability of the model; S6. Feature importance analysis Use recursive feature elimination method combined with machine learning prediction model to effectively identify key feature variables, further adjust and optimize model parameters and feature selection based on the analysis results of Pearson correlation coefficient, to accurately adjust the model and ensure the highest prediction accuracy and improve the value of practical application, and obtain the model prediction results and key features; S7. Optimization design of composite material Use machine learning to assist in optimizing composite material selection and structure design, based on the model prediction results and key features obtained in the above steps, design Janus structure polyimide composite with wave absorption and thermal conductivity performance and preparation process, and determine the Janus structure, layer components and ratio of the optimized composite material and preparation process according to the key features identified by the machine learning model; S8. Experimental verification Prepare Janus structure composite samples and synchronous preparation of control samples with reverse characteristics, then test the comprehensive performance of the samples to verify the wave absorption performance and thermal conductivity performance of the composite samples, and select the one that meets the design requirements as the optimized functional polyimide composite.

2. The method for preparing the machine learning-assisted Janus structure functional polyimide composite material according to claim 1, characterized in that, The step S1 is specifically to collect a plurality of known functional polyimide composite related information to establish a database, and the information and database includes: structure of impedance matching layer, filler of impedance matching layer, wave absorbing agent, stacking order of layer, forming process, material structure, filler of conductive layer, thickness and reflection coefficient R.

3. The method for preparing the machine learning-assisted Janus structure functional polyimide composite material according to claim 1, characterized in that, The data preprocessing of step S2 converts the classification features in step S1 into binary sub-class features "0" and "1", where "0" represents that the sample does not belong to the sub-class, and "1" represents that the sample belongs to the sub-class; the conversion results are as follows: the structure of the impedance matching layer: porous structure "1", film structure "0"; the filler of the impedance matching layer: carbon material "1" and magnetic material "0"; the wave-absorbing agent: Fe3O4 "1", other wave-absorbing materials "0"; the stacking order of the layer: impedance matching layer-wave absorbing layer-conductive layer "1", other stacking order "0"; forming process: freeze-drying, hot pressing and coating; material structure: asymmetric "1", symmetric "0"; the filler of the conductive layer: carbon material "1", magnetic material "0".

4. The method for preparing the machine learning-assisted Janus structure functional polyimide composite material according to claim 1, characterized in that, The machine learning algorithm in step S4 includes one or more of extreme gradient boosting (XGBoost) method, ridge regression (Ridge Regresso) method, gradient boosting decision tree (GBDT) method and random forest regression method.

5. The method for preparing the machine learning-assisted Janus structure functional polyimide composite material according to claim 1, characterized in that, The optimization design of the Janus structure composite material in step S7 is to regulate the dielectric properties and the difference in conductivity of the surface layer, the intermediate layer and the base layer to make each surface exhibit different electromagnetic wave reflection characteristics, including the following steps: S7-1 The surface layer uses a porous graphene / polyimide felt as an impedance matching layer to regulate the impedance characteristics of the material surface, so that a large amount of incident electromagnetic waves enter the material interior without being reflected; Among them, the surface layer polyimide felt is not filled with resin, keeping its porous structure, thereby allowing electromagnetic waves to enter; the intermediate layer and the base layer are filled with polyimide resin to prevent electromagnetic waves from penetrating; S7-2 The intermediate layer uses a four-iron oxide modified boron nitride coating as an electromagnetic wave absorbing layer, and the four-iron oxide causes eddy current loss, converting electromagnetic energy into heat energy, while the introduction of high-thermal-conductivity boron nitride promotes the rapid dissipation of heat energy, thereby improving the overall wave absorption efficiency and thermal management performance; S7-3 The base layer introduces a carbonized graphene / carbon fiber felt layer as a reflection layer to fully reflect the electromagnetic waves that are not absorbed into the absorbing layer for multiple absorption; S7-4 The wave-absorbing layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; and through hot pressing, the polyimide resin acts as an adhesive to encapsulate the designed Janus structure completely, obtaining an optimized structure of the polyimide composite material with excellent electromagnetic wave management and high-efficiency thermal management.

6. The method for preparing the machine learning-assisted Janus structure functional polyimide composite material according to claim 1, characterized in that, The preparation method of the optimization design in step S7 includes the following steps: A1. Preparation of the base layer as a reflection layer: uniformly coat the graphene solution on the surface of the polyimide (PI) felt, prepare a multilayer polyimide felt / graphene structure by layering, place the multilayer graphene / PI felt structure in a specially designed mold, apply a vertical downward pressure to ensure the tight combination between the layers, and perform high-temperature carbonization treatment; A2. Preparation of the intermediate layer as an absorption layer: ultrasonic-assisted liquid phase exfoliation method is used to exfoliate and modify boron nitride (BN) powder to prepare functionalized boron nitride nanosheets (f-BNNSs), and f-BNNSs and ferroferric oxide (Fe3O4) are uniformly mixed and coated on the surface of the carbonized PI felt; A3. Preparation of the surface layer as an impedance matching layer: the PI felt is immersed in a low-concentration graphene solution and ultrasonically treated for half an hour to ensure that graphene particles are uniformly attached to the surface of the felt; A4. Hot pressing of each layer: PI powder is dissolved in an organic solvent to prepare a PI solution, the coated wave-absorbing conductive layer is immersed in the PI solution, and the solvent is evaporated by programmed temperature rise; the impedance matching layer and the cured absorption layer-conductive layer are connected by hot pressing with polyimide resin as the adhesive, and the PI resin is further cured and forms a seamless connection during the hot pressing process, and finally an optimized preparation method of functional polyimide composite is obtained.

7. A functional polyimide composite, characterized by, It is obtained by the machine learning aided design method of any one of claims 1-6, which comprises a surface layer, an intermediate layer and a base layer stacked in turn, wherein the surface layer uses a porous graphene / polyimide felt as an impedance matching layer to regulate the impedance characteristics of the material surface, so that a large amount of incident electromagnetic waves enter the material interior without being reflected; the intermediate layer uses a four-iron modified boron nitride coating as an electromagnetic wave absorption layer, four-iron causes eddy current loss, converts electromagnetic energy into heat energy, and the introduction of high-thermal-conductivity boron nitride promotes the rapid dissipation of heat energy, thereby improving the overall wave-absorbing efficiency and thermal management performance; the base layer introduces a carbonized graphene / carbon fiber felt layer as a reflection layer to fully reflect the electromagnetic waves that are not absorbed into the absorption layer for multiple absorption; the wave-absorbing layer and the conductive layer are filled with polyimide resin to improve the structural stability of the material; finally, through hot pressing, the polyimide resin is used as an adhesive to encapsulate the designed Janus structure completely, and a functional polyimide composite material with excellent electromagnetic wave management and high-efficiency thermal management is obtained.

8. The functional polyimide composite according to claim 7, wherein The number of layers of the multi-layer graphene / polyimide felt structure is 3-x layers of graphene / PI felt arranged alternately, and x≥5.

9. The functional polyimide composite according to claim 7, wherein In the functional polyimide composite material, the thickness of the surface layer and the base layer is 500 μm - 3000 μm.

Images