Wave-absorbing structure design method based on artificial intelligence assistance

Abstract

The invention discloses a wave-absorbing structure design method based on artificial intelligence assistance. The wave-absorbing structure design method comprises the steps of data collection and processing, AI model training, wave-absorbing structure design, verification and optimization and iterative optimization. The advanced artificial intelligence technology is introduced, the complex relation among the chemical composition, the physical structure parameters and the electromagnetic performance of the wave-absorbing material is automatically learned, the period from material design to performance verification is greatly shortened, the tedious trial and error process is reduced, and therefore the overall efficiency of wave-absorbing structure design is remarkably improved; the experiment frequency is reduced and the research and development period is shortened; the problems that in an existing design scheme of the wave-absorbing material, an experimental trial and error method is long in consumed time and high in cost, and all possible material combinations and structures are difficult to comprehensively explore are solved, and the problems that when a theoretical calculation method is used for processing a complex multi-layer structure or a non-linear material, the calculation amount is huge, and the requirement for calculation resources is high are solved.

Description

Technical Field

[0001] This invention relates to, but is not limited to, the interdisciplinary field of materials science and artificial intelligence, and specifically to an AI-assisted method for designing microwave absorbing structures. Background Technology

[0002] In the field of electromagnetics, absorbing materials play a crucial role, and they are widely used in aerospace low-detectability technology, radar absorbing materials, electromagnetic compatibility (EMC) design, and anti-interference of wireless communication equipment. The core function of absorbing materials is to effectively absorb, attenuate, or scatter incident electromagnetic waves, thereby reducing the reflection and transmission of electromagnetic waves and achieving the purpose of low detectability or reducing electromagnetic interference.

[0003] In traditional approaches, the design and optimization of microwave absorbing materials primarily rely on experimental trial-and-error methods and theoretical calculations. Experimental trial-and-error involves preparing samples of absorbing materials with different compositions, testing their absorption performance, and then iteratively adjusting based on the test results. While intuitive, this method is time-consuming, costly, and limited by experimental conditions and the number of samples that can be prepared, making it difficult to comprehensively explore all possible material combinations and structures. Theoretical calculations, based on electromagnetic field theory, predict the performance of absorbing materials by establishing mathematical models. These methods include transmission line theory, microwave network theory, and finite element analysis, which can accurately simulate the electromagnetic response of absorbing materials. However, when dealing with complex multilayer structures or nonlinear materials, the computational load is enormous, and the computational requirements are high. Summary of the Invention

[0004] The purpose of this invention is to provide an artificial intelligence-assisted design method for absorbing structures, which solves the problems of time-consuming, costly, and difficult-to-explore all possible material combinations and structures in the existing design schemes of absorbing materials by using experimental trial and error methods. It also solves the problems of huge computational load and high computational resource requirements when theoretical calculation methods deal with complex multilayer structures or nonlinear materials.

[0005] The technical solution of this invention is as follows: This invention provides a method for designing a microwave absorbing structure based on artificial intelligence assistance, comprising the following steps: S1. Data collection and processing, including: A large amount of experimental and simulation data of absorbing materials are collected and preprocessed to form a dataset that can be used for AI training; wherein the experimental and simulation data include the structural parameters of the absorbing materials, as well as reflection loss curves or absorptivity data. S2, AI model training, including: Using machine learning algorithms, the preprocessed dataset in S1 is trained to establish a bidirectional fast mapping relationship between the structural parameters of the absorbing material and its reflection loss curve or absorptivity data, thus training an AI model. S3. Design the wave-absorbing structure, including: Based on the AI ​​model trained in S2, input the target reflection loss curve or absorptivity data, and automatically output the structural parameters of the absorbing material that meet the performance requirements of the absorbing material through the AI ​​model. S4. Verification and optimization, including: Based on the simulation results of the structural parameters of the absorbing material, the structural parameters of the absorbing material output by the AI ​​model in S3 are verified and optimized to obtain the structural parameters of the absorbing material that meet the performance requirements.

[0006] Optionally, in the AI-assisted microwave absorbing structure design method described above, step S1 includes the following steps: S11. Data collection, including: Experimental data were collected from existing microwave absorbing material experiments, including the chemical composition, structural parameters, electromagnetic parameters, and reflection loss curves or absorptivity data of the microwave absorbing materials; a large amount of simulation data of microwave absorbing structures was generated using electromagnetic simulation software. The simulation data covered the microwave absorption performance under different design parameters, and the microwave absorption performance was represented by reflection loss curves or absorptivity data. S12. Data preprocessing, including: Remove incomplete, abnormal, or duplicate data records to ensure data quality; extract characteristic variables that have a significant impact on the absorption performance, including at least one of the chemical composition, structural parameters, and electromagnetic parameters of the absorbing material, and normalize the characteristic variables to facilitate AI model processing; use the reflection loss curve or absorption rate data as labels for AI model prediction.

[0007] Optionally, in the AI-assisted microwave absorbing structure design method described above, step S2 includes the following steps: S21. Select a model, including: Convolutional Neural Network (CNN) was chosen as the core algorithm. The structural parameters of the absorbing material were treated as image input, and the reflection loss curve or absorption rate data were treated as label output. A CNN model was constructed, consisting of an input layer, multiple convolutional layers, pooling layers, fully connected layers, and an output layer. The input layer received preprocessed feature variables, the convolutional and pooling layers were used to extract features from the feature variables received by the input layer, the fully connected layer was used to realize the nonlinear mapping between the image input and the label output, and the output layer was used to output the predicted reflection loss curve or absorption rate data. S22. Training the CNN model, including: The preprocessed dataset in S1 is divided into training, validation, and test sets to evaluate the generalization ability of the CNN model. The CNN model is trained using the training set data, and the model parameters are adjusted through the backpropagation algorithm to minimize the error between the predicted and true values. The learning rate, batch size, and kernel size hyperparameters are adjusted through cross-validation to optimize the performance of the CNN model and train the AI ​​model.

[0008] Optionally, in the AI-assisted microwave absorption structure design method described above, step S3 includes the following steps: S31. Set inputs, including: Based on application requirements, set the target reflection loss curve or target absorption rate data, and consider the availability, processing cost and weight limitations of the absorbing material in actual applications to set the range of structural parameters of the absorbing material. S32. Model predictions, including: The target reflection loss curve or target absorption rate data is input into the trained AI model, and the AI ​​model outputs multiple structural parameters that satisfy the target reflection loss curve or target absorption rate data. Since the AI ​​model generates multiple solutions that meet the conditions, the optimal solution is selected using genetic algorithms and particle swarm optimization strategies.

[0009] Optionally, in the AI-assisted microwave absorbing structure design method described above, step S4 includes the following steps: S41. Simulation verification, including: Electromagnetic simulation software was used to simulate the structural parameters of the microwave absorbing material output by the AI ​​model to verify whether its microwave absorbing performance met the design requirements. Based on the simulation results, the combination of structural parameters was adjusted to ensure that the performance met the requirements. S42. Experimental verification, including: Based on the optimized structural parameters, microwave absorbing material samples were prepared. The reflection loss curves or absorption rate data of the samples were tested using a vector network analyzer and compared with the simulation results to verify the accuracy of AI-assisted design.

[0010] Optionally, in the AI-assisted absorbing structure design method described above, the normalization process in S12 adopts the Min-Max normalization method to scale the feature variables to the [0,1] interval in order to eliminate the influence of dimensional differences on model training.

[0011] Optionally, in the AI-assisted microwave absorption structure design method described above, the hyperparameter optimization in S22 adopts the K-fold cross-validation method, and the model's generalization ability is ensured by adjusting the learning rate, batch size, and convolution kernel size.

[0012] Optionally, in the AI-assisted microwave absorption structure design method described above, the genetic algorithm in S32 adopts a tournament selection strategy, sets the crossover probability to 0.8 and the mutation probability to 0.1, and combines the fitness function to evaluate candidate solutions in order to screen the globally optimal combination of structural parameters.

[0013] Optionally, the AI-assisted microwave absorbing structure design method described above further includes: S5. Iteratively optimize the AI ​​model trained in S2, including: The experimental verification results in S4 are fed back to the AI ​​model for training and optimization, forming a closed-loop iteration to continuously improve design efficiency and accuracy.

[0014] The beneficial effects of this invention are as follows: This invention provides an artificial intelligence-assisted method for designing microwave absorbing structures, which has at least the following beneficial effects: (1) By introducing advanced artificial intelligence technology, especially deep learning models such as convolutional neural networks (CNN), this invention can automatically learn the complex relationship between the chemical composition, physical structure parameters and electromagnetic properties of the absorbing material. This feature greatly shortens the cycle from material design to performance verification, reduces the tedious trial and error process, and thus significantly improves the overall efficiency of absorbing structure design. (2) By utilizing the predictive capabilities of artificial intelligence models, this invention can accurately predict the absorption performance under different combinations of structural parameters, including reflection loss curves or absorption rate data. Furthermore, through optimization algorithms (such as genetic algorithms, particle swarm optimization, etc.), the optimal combination of structural parameters that meets specific application requirements can be further screened out, thereby achieving the optimization of absorption performance. In addition, the artificial intelligence model provided by this invention can explore a wider parameter space, discover new material compositions and structural configurations, and provide unprecedented flexibility and innovation for the design of absorbing materials. (3) By reducing the number of experiments and shortening the R&D cycle, the present invention effectively reduces the material cost, labor cost and time cost in the R&D process of absorbing materials. In addition, since the model can accurately predict the performance, unnecessary sample preparation and testing are reduced, further reducing the overall R&D cost. (4) By optimizing the material composition and structure, the present invention is expected to contribute to reducing resource consumption and environmental pollution. For example, by designing efficient microwave absorbing materials, the impact of electromagnetic pollution on the environment and human health can be reduced. At the same time, by precisely controlling the amount and structure of materials, energy consumption and waste generation in the production process can be reduced. Attached Figure Description

[0015] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0016] Figure 1 A flowchart illustrating an artificial intelligence-assisted design method for a microwave absorbing structure, as provided in an embodiment of the present invention; Figure 2 This is a model structure diagram of an artificial intelligence-assisted microwave absorbing structure design method provided in an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

[0018] As explained in the background section above, traditional design and optimization schemes for microwave absorbing materials typically employ experimental trial-and-error methods and theoretical calculation methods. Experimental trial-and-error methods are time-consuming, costly, and difficult to fully explore all possible material combinations and structures. Theoretical calculation methods, when dealing with complex multilayer structures or nonlinear materials, involve a huge amount of computation and require high computational resources.

[0019] In recent years, with the rapid development of computer technology and the arrival of the big data era, artificial intelligence (AI) technology, especially machine learning and deep learning algorithms, has shown enormous potential in the field of materials science. AI technology can learn the relationship between the structure and properties of materials from massive amounts of data, thereby enabling the efficient design and optimization of new materials. Significant results have been achieved in fields such as metamaterials, optoelectronic materials, and catalysts, greatly accelerating the discovery and application of new materials.

[0020] However, in the field of microwave absorbing materials, although some studies have attempted to utilize AI technology for performance prediction and design optimization, these works have mostly focused on specific types of absorbing materials (such as frequency-selective surfaces and metamaterials). The application of AI technology in the design optimization of more widely used traditional absorbing materials (such as ferrites and carbon-based materials) remains limited. Furthermore, existing research often neglects the complex nonlinear relationship between the structure and performance of absorbing materials, as well as the influence of process parameters during material preparation on performance, when constructing AI models, resulting in limited predictive accuracy and generalization ability of the models.

[0021] Therefore, developing a method for efficiently and accurately designing traditional microwave absorbing materials with excellent absorption performance is of great significance for promoting the development and application of microwave absorbing material technology. Based on this need, this invention proposes an artificial intelligence-assisted microwave absorbing structure design method. This method aims to learn the relationship between the structure and performance of microwave absorbing materials from a large amount of experimental data and simulation results using AI technology, quickly identifying the material structure and composition with good microwave absorption performance, thereby overcoming the limitations of traditional methods and improving design efficiency and performance.

[0022] The present invention provides the following specific embodiments, which can be combined with each other. For the same or similar concepts or processes, they may not be described again in some embodiments.

[0023] Figure 1 A flowchart illustrating an artificial intelligence-assisted design method for a microwave absorbing structure, as provided in an embodiment of the present invention; Figure 2 This is a model structure diagram of an artificial intelligence-assisted microwave absorbing structure design method provided in an embodiment of the present invention. (Refer to...) Figure 1 and Figure 2 As shown, the AI-assisted microwave absorbing structure design method provided by this invention includes the following steps: S1. Data collection and processing, including: We collected a large amount of experimental and simulation data on microwave absorbing materials and preprocessed them to form a dataset that can be used for AI training. The experimental and simulation data in S1 include the structural parameters of the absorbing material, as well as the reflection loss curve or absorptivity data. S2, AI model training, including: Using machine learning algorithms, such as convolutional neural networks (CNN), the preprocessed dataset in S1 is trained to establish a bidirectional fast mapping relationship between the structural parameters of the absorbing material and its reflection loss curve or absorption rate data, thus training an AI model. S3. Design the wave-absorbing structure, including: Based on the AI ​​model trained in S2, input the target reflection loss curve or absorptivity data, and automatically output the structural parameters of the absorbing material that meet the performance requirements of the absorbing material through the AI ​​model. S4. Verification and optimization, including: Based on the simulation results of the structural parameters of the absorbing material, the structural parameters of the absorbing material output by the AI ​​model in S3 are verified and optimized to obtain the structural parameters of the absorbing material that meet the performance requirements, thereby ensuring the accuracy and reliability of the design.

[0024] In one implementation of this invention, the process of S1 described above may include: S11, Data Collection Experimental data will be collected from existing research on microwave absorbing materials, including but not limited to the chemical composition, physical structure, electromagnetic parameters, and corresponding reflection loss curves or absorptivity data of microwave absorbing materials. In addition to existing research literature, microwave absorbing material data from public databases (such as the Materials Project and the NIST Electromagnetic Materials Library) will be integrated, and measured data will be shared with collaborative laboratories to ensure that the data covers traditional microwave absorbing materials such as ferrite, carbon-based composite materials, and conductive polymers. A large number of simulation data of microwave absorbing structures will be generated using electromagnetic simulation software, covering the microwave absorption performance under different design parameters. By using CSTStudio Suite or ANSYS HFSS electromagnetic simulation software, multiple parameter combinations (such as thickness 0.1-5mm, number of layers 1-5, dielectric constant 2-20, and permeability 1-5) will be set to generate reflection loss curves covering the 1GHz-40GHz frequency band. The amount of simulation data will be no less than 100,000 sets, covering typical application scenarios such as low-frequency broadband and high-frequency narrowband. S12, Data Preprocessing To ensure data quality, incomplete, abnormal, or duplicate data records were removed. Box plot analysis was used to eliminate outlier data with reflection loss values ​​exceeding ±3 standard deviations. For missing values, KNN (K-nearest neighbor) interpolation was used to fill in the missing values. Feature variables that significantly affect the absorption performance were extracted. Pearson correlation coefficient analysis was used to screen key features (such as thickness, interlayer spacing, and real / imaginary part of dielectric constant) with a correlation >0.8 with reflection loss. Principal component analysis (PCA) was then used to reduce the dimensionality to within 10 dimensions to avoid overfitting, such as material structural parameters and electromagnetic parameters, and normalization was performed.

[0025] In specific implementation, the data normalization process in S12 can adopt the Min-Max normalization method to scale the feature variables to the [0,1] interval to eliminate the influence of dimensional differences on model training, so as to facilitate AI model processing; the reflection loss curve or the absorption rate at a specific frequency is used as the label for model prediction.

[0026] In one implementation of this invention, the process of S2 described above may include: S21, Model Selection Given that the design of microwave absorbing materials involves a complex multi-dimensional spatial search problem, a convolutional neural network (CNN) is chosen as the core algorithm. The structural parameters of the microwave absorbing material are regarded as "image" input and the reflection loss curve is regarded as "label" output. A CNN model containing an input layer, multiple convolutional layers, pooling layers, fully connected layers and an output layer is designed. The input layer receives the preprocessed feature vector, the convolutional and pooling layers are used for feature extraction, the fully connected layer is used for nonlinear mapping, and the output layer outputs the predicted reflection loss curve or absorption rate value. The specific principle of CNN architecture design is as follows: The input layer receives a 10-dimensional feature vector, which is then processed by 3 convolutional layers (3×3 kernel size, stride 1, activation function ReLU) and 2 max pooling layers (pooling window 2×2) to extract features. The fully connected layer contains 128 neurons (Dropout rate 0.5 to prevent overfitting). The output layer uses a linear activation function to predict the reflection loss curve.

[0027] S22. Train the CNN model The preprocessed dataset in S1 is divided into training, validation, and test sets to evaluate the model's generalization ability. The CNN model is trained using the training set data, and the model parameters are adjusted through the backpropagation algorithm to minimize the error between the predicted and true values. The learning rate, batch size, and kernel size hyperparameters are adjusted through methods such as cross-validation to optimize the performance of the CNN model and train the AI ​​model.

[0028] In practice, the hyperparameter optimization in S22 can be achieved using the k-fold cross-validation method. By adjusting the learning rate, batch size, and convolutional kernel size, the generalization ability of the model can be ensured.

[0029] In one implementation of this invention, the process of S3 described above may include: S31, Setting Input Based on application requirements, set the target reflection loss curve or target absorption rate data, and consider the availability, processing cost and weight limitations of the absorbing material in actual applications to set the range of structural parameters of the absorbing material. S32, Model Prediction The target reflection loss curve or target absorption rate data is input into the trained AI model, and the AI ​​model outputs multiple structural parameters that satisfy the target reflection loss curve or target absorption rate data. Since the AI ​​model generates multiple solutions that meet the conditions, the optimal solution is selected using genetic algorithms and particle swarm optimization strategies.

[0030] For example, a genetic algorithm uses a tournament selection strategy, setting the crossover probability to 0.8 and the mutation probability to 0.1, and combines this with a fitness function to evaluate candidate solutions in order to select the globally optimal combination of structural parameters. The fitness function is defined as: ; in , To balance performance and cost.

[0031] In one implementation of this invention, in step S4, the microwave absorbing structure output by the AI ​​model in step S3 is further optimized and verified based on the high accuracy of simulation calculations to ensure the accuracy and reliability of the design. The implementation process of step S4 in this embodiment may include: S41, Simulation Verification Electromagnetic simulation software was used to simulate the structural parameters of the microwave absorbing material output by the AI ​​model to verify whether its microwave absorbing performance met the design requirements. Based on the simulation results, the combination of structural parameters was adjusted to ensure that the performance met the requirements.

[0032] In practice, for example, in CST Studio Suite, plane wave incidence is set, the boundary condition is open space, the mesh accuracy is λ / 20, the simulation frequency band is 1-40GHz, and the step size is 0.1GHz. If the deviation between the simulated absorption rate and the predicted value is >5%, the gradient descent method is used to adjust the structural parameters (such as thickness ±0.05mm) until the error is <2%.

[0033] S42, Experimental Verification Based on the optimized structural parameters, microwave absorbing material samples were prepared. The reflection loss curves or absorption rate data of the samples were tested using a vector network analyzer and compared with the simulation results to verify the accuracy of AI-assisted design.

[0034] In specific implementation, for example, ferrite-based microwave absorbing materials are prepared using powder metallurgy (sintering temperature 1200℃, pressure 50MPa), and the interlayer spacing is controlled to an accuracy of ±0.01mm using the lamination process; in a shielded anechoic chamber (temperature 25±1℃, humidity <30%), the 2-18GHz reflection loss is tested using a Keysight N5224B vector network analyzer, and the calibration is performed using the TRL method.

[0035] Furthermore, the AI-assisted microwave absorbing structure design method provided in this embodiment of the invention may further include: S5. Iteratively optimize the AI ​​model trained in S2, including: The experimental verification results in S4 are fed back to the AI ​​model for training and optimization, forming a closed-loop iteration to continuously improve design efficiency and accuracy.

[0036] In practice, the training set is updated with a 5:1 weight (simulation:experiment) for the experimental data, and the generalization ability of the model is re-evaluated every 10 iterations. A closed loop of "simulation-experiment-feedback" is implemented through Python scripts to automatically trigger model retraining and reduce manual intervention.

[0037] The design method provided by the above embodiments of the present invention can significantly improve the design efficiency of microwave absorbing materials and reduce R&D costs. At the same time, AI technology can be used to accurately predict and optimize the structural parameters of microwave absorbing materials, thereby significantly improving their electromagnetic performance. In addition, the present invention promotes the deep integration of artificial intelligence technology and materials science, and drives the development of interdisciplinary research.

[0038] The AI-assisted microwave absorbing structure design method provided in this invention introduces advanced AI technologies, particularly deep learning models such as convolutional neural networks (CNNs). This invention can automatically learn the complex relationships between the chemical composition, physical structural parameters, and electromagnetic properties of microwave absorbing materials. This feature greatly shortens the cycle from material design to performance verification, reduces tedious trial-and-error processes, and thus significantly improves the overall efficiency of microwave absorbing structure design. By utilizing the predictive capabilities of AI models, this invention can accurately predict the microwave absorbing performance under different parameter combinations, including reflection loss curves and absorption rates at specific frequencies. Through optimization algorithms (such as genetic algorithms and particle swarm optimization), the optimal structural parameters that meet specific application requirements can be further screened, thereby achieving optimal microwave absorbing performance. At the same time, the AI ​​model of this invention can explore a wider parameter space, discover new material compositions and structural configurations, and provide unprecedented flexibility and innovation for the design of microwave absorbing materials.

[0039] Furthermore, by reducing the number of experiments and shortening the R&D cycle, this invention effectively reduces material costs, labor costs, and time costs in the R&D process of microwave absorbing materials. In addition, since the model can accurately predict performance, unnecessary sample preparation and testing are reduced, further reducing the overall R&D cost. By optimizing the material composition and structure, this invention is expected to contribute to reducing resource consumption and environmental pollution. For example, by designing highly efficient microwave absorbing materials, the impact of electromagnetic pollution on the environment and human health can be reduced. At the same time, by precisely controlling the amount and structure of materials, energy consumption and waste generation in the production process can be reduced.

[0040] While the embodiments disclosed in this invention are as described above, they are merely illustrative of the embodiments to facilitate understanding of the invention and are not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in the form and details of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for designing a microwave absorbing structure based on artificial intelligence assistance, characterized in that, Includes the following steps: S1. Data collection and processing, including: A large amount of experimental and simulation data of absorbing materials are collected and preprocessed to form a dataset that can be used for AI training; wherein the experimental and simulation data include the structural parameters of the absorbing materials, as well as reflection loss curves or absorptivity data. S2, AI model training, including: Using machine learning algorithms, the preprocessed dataset in S1 is trained to establish a bidirectional fast mapping relationship between the structural parameters of the absorbing material and its reflection loss curve or absorptivity data, thus training an AI model. S3. Design the wave-absorbing structure, including: Based on the AI ​​model trained in S2, input the target reflection loss curve or absorptivity data, and automatically output the structural parameters of the absorbing material that meet the performance requirements of the absorbing material through the AI ​​model; S4. Verification and optimization, including: Based on the simulation results of the structural parameters of the absorbing material, the structural parameters of the absorbing material output by the AI ​​model in S3 are verified and optimized to obtain the structural parameters of the absorbing material that meet the performance requirements.

2. The method for designing a microwave absorbing structure based on artificial intelligence assistance according to claim 1, characterized in that, S1 includes the following steps: S11. Data collection, including: Experimental data were collected from existing microwave absorbing material experiments, including the chemical composition, structural parameters, electromagnetic parameters, and reflection loss curves or absorptivity data of the microwave absorbing materials; a large amount of simulation data of microwave absorbing structures was generated using electromagnetic simulation software. The simulation data covered the microwave absorption performance under different design parameters, and the microwave absorption performance was represented by reflection loss curves or absorptivity data. S12. Data preprocessing, including: Remove incomplete, abnormal, or duplicate data records to ensure data quality; extract characteristic variables that have a significant impact on the absorption performance, including at least one of the chemical composition, structural parameters, and electromagnetic parameters of the absorbing material, and normalize the characteristic variables to facilitate AI model processing; use the reflection loss curve or absorption rate data as labels for AI model prediction.

3. The method for designing a microwave absorbing structure based on artificial intelligence assistance according to claim 2, characterized in that, S2 includes the following steps: S21. Select a model, including: Convolutional Neural Network (CNN) was chosen as the core algorithm. The structural parameters of the absorbing material were treated as image input, and the reflection loss curve or absorption rate data were treated as label output. A CNN model was constructed, consisting of an input layer, multiple convolutional layers, pooling layers, fully connected layers, and an output layer. The input layer received preprocessed feature variables, the convolutional and pooling layers were used to extract features from the feature variables received by the input layer, the fully connected layer was used to realize the nonlinear mapping between the image input and the label output, and the output layer was used to output the predicted reflection loss curve or absorption rate data. S22. Training the CNN model, including: The preprocessed dataset in S1 is divided into training, validation, and test sets to evaluate the generalization ability of the CNN model. The CNN model is trained using the training set data, and the model parameters are adjusted through the backpropagation algorithm to minimize the error between the predicted and true values. The learning rate, batch size, and kernel size hyperparameters are adjusted through cross-validation to optimize the performance of the CNN model and train the AI ​​model.

4. The method for designing an absorbing structure based on artificial intelligence assistance according to claim 3, characterized in that, S3 includes the following steps: S31. Set inputs, including: Based on application requirements, set the target reflection loss curve or target absorption rate data, and consider the availability, processing cost and weight limitations of the absorbing material in actual applications to set the range of structural parameters of the absorbing material. S32. Model predictions, including: The target reflection loss curve or target absorption rate data is input into the trained AI model, and the AI ​​model outputs multiple structural parameters that satisfy the target reflection loss curve or target absorption rate data. Since the AI ​​model generates multiple solutions that meet the conditions, the optimal solution is selected using genetic algorithms and particle swarm optimization strategies.

5. The method for designing an absorbing structure based on artificial intelligence assistance according to claim 4, characterized in that, S4 includes the following steps: S41. Simulation verification, including: Electromagnetic simulation software was used to simulate the structural parameters of the microwave absorbing material output by the AI ​​model to verify whether its microwave absorbing performance met the design requirements. Based on the simulation results, the combination of structural parameters was adjusted to ensure that the performance met the requirements. S42. Experimental verification, including: Based on the optimized structural parameters, microwave absorbing material samples were prepared. The reflection loss curves or absorption rate data of the samples were tested using a vector network analyzer and compared with the simulation results to verify the accuracy of AI-assisted design.

6. The method for designing a microwave absorbing structure based on artificial intelligence assistance according to claim 2, characterized in that, The normalization process in S12 uses the Min-Max normalization method to scale the feature variables to the [0,1] interval to eliminate the impact of dimensional differences on model training.

7. The method for designing a microwave absorbing structure based on artificial intelligence assistance according to claim 3, characterized in that, The hyperparameter optimization in S22 uses the K-fold cross-validation method, which adjusts the learning rate, batch size, and convolutional kernel size to ensure the model's generalization ability.

8. The method for designing an absorbing structure based on artificial intelligence assistance according to claim 4, characterized in that, The genetic algorithm in S32 adopts a tournament selection strategy, sets the crossover probability to 0.8 and the mutation probability to 0.1, and combines the fitness function to evaluate candidate solutions in order to screen the globally optimal combination of structural parameters.

9. A method for designing a microwave absorbing structure based on artificial intelligence according to any one of claims 1 to 5, characterized in that, Also includes: S5. Iteratively optimize the AI ​​model trained in S2, including: The experimental verification results in S4 are fed back to the AI ​​model for training and optimization, forming a closed-loop iteration to continuously improve design efficiency and accuracy.

Images