Method and system for optimizing flywheel energy storage by using mixed material

Abstract

The invention discloses a method and system for optimizing flywheel energy storage by using a mixed material, and relates to the field of flywheel energy storage, and the method comprises the steps: collecting data, screening candidate materials, and generating a material list; carrying out attribute modeling, and dividing the rotor into a plurality of stress areas through finite element analysis; calculating the volume of each stress area and the maximum stress of each area at the working rotating speed, distributing materials, and generating an initial material distribution scheme; iteration is carried out, and material distribution is optimized until the energy storage density and the limit rotating speed target are met. According to the method, multiple performance parameters of the material are comprehensively considered, the rotor material is reasonably selected, it is ensured that the rotor is not damaged due to insufficient strength in the working process, meanwhile, the rotor quality is optimized, and local stress concentration is reduced. In the material distribution optimization process, the maximum stress does not exceed the material strength limit, and meanwhile, the rotor mass is reduced, the energy storage density is improved, and the magnetic bearing load is reduced. Performance of the flywheel energy storage system rotor is integrally improved, and operation efficiency, stability and reliability of the system are enhanced.

Description

Technical Field

[0001] This invention relates to the field of flywheel energy storage, specifically to a method and system for optimizing flywheel energy storage using hybrid materials. Background Technology

[0002] In existing flywheel energy storage systems, the rotor generally adopts a uniform disk design. This design is relatively mature in technology due to long-term practical application, and the manufacturing process is also relatively simple. However, its uniform structural characteristics bring many adverse effects.

[0003] From a mass perspective, due to its uniform design, the overall rotor mass is relatively large. According to the formula for calculating disk mass, this large mass directly limits the system's energy storage density, making it unable to meet the high energy storage density requirements of modern energy storage systems.

[0004] Regarding the interaction with the magnetic bearing, a larger rotor mass significantly increases the load-bearing pressure on the magnetic bearing. The magnetic bearing requires higher performance to maintain stable rotor levitation and rotation, which not only increases the design difficulty of the magnetic bearing but also greatly increases the complexity of the entire system's design and maintenance, while also increasing costs. From a stress distribution perspective, when the rotor rotates at high speed, the uniform disk design leads to stress concentration in specific areas. According to the stress calculation formula for a rotating disk in elasticity, this stress concentration phenomenon increases the maximum stress on the rotor, thereby increasing the risk of rotor failure and severely limiting the system's maximum speed, hindering further performance improvements of the flywheel energy storage system. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by this invention is: how to optimize the performance of the flywheel energy storage system rotor, improve the energy storage density and ultimate speed by rationally selecting and allocating mixed materials, while ensuring structural safety and quality optimization.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for optimizing flywheel energy storage using hybrid materials, comprising the following steps, Collect data to screen candidate materials and generate a material list; Based on the materials in the material list, property modeling is performed, and the rotor is divided into multiple stress regions through finite element analysis. Calculate the volume of each stress region and the maximum stress of each region at the operating speed, allocate materials, and generate an initial material allocation scheme; The initial material allocation scheme is iterated and optimized until the energy storage density and ultimate rotational speed targets are met.

[0008] As a preferred embodiment of the flywheel energy storage method using hybrid materials described in this invention, the data collection involves obtaining the density, ultimate tensile strength, and elastic modulus of various candidate materials. Based on the maximum expected stress value of the flywheel energy storage system rotor during operation, materials that meet the strength limit are selected. Then, materials with the lowest density are prioritized. By calculating the ratio of the difference in elastic modulus of the materials to the average value, a ratio threshold is set to screen material combinations and draw a material list.

[0009] As a preferred embodiment of the flywheel energy storage method using hybrid materials described in this invention, the property modeling based on the materials in the material list involves constructing a geometric model of a uniform disk rotor based on the selected materials, meshing the geometric model, discretizing the continuous model into a finite number of elements and nodes, determining the mesh type and size, setting boundary conditions according to the rotor's operating conditions, running finite element analysis, calculating the stress distribution of the rotor under the set operating conditions, generating a stress cloud map, and dividing the region according to the stress magnitude represented by the color in the stress cloud map.

[0010] As a preferred embodiment of the flywheel energy storage method using hybrid materials described in this invention, the mesh type and size are determined based on the complexity of the model and the required computational accuracy, and a suitable mesh density is determined through convergence analysis. Different colors in the stress cloud diagram represent different stress magnitudes.

[0011] As a preferred embodiment of the flywheel energy storage method using hybrid materials described in this invention, the initial material distribution scheme is generated by calculating the corresponding volume and the maximum stress at the corresponding operating speed based on the region in the stress cloud diagram. The maximum stress should be less than the material's strength limit. Select the material with the lowest density, iterate through all candidate materials, and generate a suitable initial material allocation scheme.

[0012] As a preferred embodiment of the method for optimizing flywheel energy storage using hybrid materials described in this invention, the optimization of material allocation until the energy storage density and the ultimate speed target are met involves inputting an initial material allocation scheme into an XGBoost model, allocating several candidate materials to the region, and representing a material allocation scheme using an integer vector of a certain length, with the vector added to the dataset as a feature column. Calculate the volume of each region, and combine the volume values ​​of each region into a feature vector and add it to the dataset; For each region, calculate the average radius, output a basic stress estimate, and combine it into a feature vector and add it to the dataset. By generating a large number of different material allocation scheme examples, a dataset containing multiple samples is constructed; The target variables are constructed by including labels for energy density difference and limit speed difference, forming a label vector.

[0013] As a preferred embodiment of the method for optimizing flywheel energy storage using hybrid materials described in this invention, the optimization of material allocation until the energy storage density and ultimate speed targets are met includes determining the parameters of the XGBoost model and dividing the constructed dataset into a training set, a validation set, and a test set according to a certain ratio. The XGBoost model is trained using the training set, and the objective function is minimized, expressed as: , , , in, It is a loss function, which is constructed by summing the loss functions for the energy density difference label and the limiting speed difference label using the mean square error. , These are the energy storage density difference labels and the XGBoost model predictions, respectively. These are the labels for the difference in extreme speeds and the predicted values ​​from the XGBoost model, respectively. Using the trained XGBoost model, the newly generated or evaluated material allocation schemes are predicted, and the output material allocation schemes with energy density difference and limit speed difference both approaching 0 are generated. The material allocation is optimized and adjusted, the maximum stress in each region is calculated and compared with the corresponding material strength limit, and a new set of material allocation scheme candidates is generated. The newly generated candidate material allocation schemes are converted into feature vectors and input into the trained XGBoost model for prediction. This yields the corresponding predicted energy density difference and limiting speed difference. The scheme that best combines the predicted energy density difference and limiting speed difference is selected to achieve the optimal result. The minimum material allocation scheme will serve as the basis for the next iteration. Repeat the steps of generating candidate material allocation schemes, model evaluation and selection, and optimize material allocation in the direction that makes energy storage density and limiting rotational speed closer to the target values. Stop repeating when the maximum number of iterations reaches the requirement or the accuracy meets the requirement, and output the final optimal material allocation scheme.

[0014] Another objective of this invention is to provide a flywheel energy storage system optimized using hybrid materials. This system can be optimized by comprehensively considering parameters such as the material's strength limit, density, and elastic modulus, combined with finite element analysis and the XGBoost algorithm, thus addressing the shortcomings of existing flywheel energy storage systems in terms of material distribution and performance improvement.

[0015] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a flywheel energy storage system optimized using hybrid materials, comprising: A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method for optimizing flywheel energy storage using hybrid materials as described above.

[0016] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a flywheel energy storage method using hybrid materials as described above.

[0017] The beneficial effects of this invention are as follows: By comprehensively considering multiple performance parameters such as the material's ultimate tensile strength, density, and elastic modulus, this invention scientifically and rationally selects rotor materials, ensuring that the rotor is not damaged due to insufficient strength during operation, while simultaneously optimizing rotor mass and reducing local stress concentration. Finite element analysis is used to accurately delineate rotor stress regions, ensuring that the region division reflects the actual stress distribution and is both manufacturing-feasible. During material allocation optimization, the maximum stress is kept within the material's ultimate tensile strength while also reducing rotor mass, increasing energy storage density, and reducing magnetic bearing load. The XGBoost algorithm is used for efficient optimization, combined with multi-dimensional feature information, ensuring that the material allocation scheme achieves an optimal balance between energy storage density and limiting speed while meeting strength requirements. Overall, this improves the performance of the flywheel energy storage system rotor, enhancing the system's operating efficiency, stability, and reliability. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The first embodiment of the present invention provides an overall flowchart of a method for optimizing flywheel energy storage using hybrid materials. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0021] Example 1, referring to Figure 1As one embodiment of the present invention, a method for optimizing flywheel energy storage using hybrid materials is provided, comprising: Collect data to screen candidate materials and generate a material list; Based on the materials in the material list, property modeling is performed, and the rotor is divided into multiple stress regions through finite element analysis. Calculate the volume of each stress region and the maximum stress of each region at the operating speed, allocate materials, and generate an initial material allocation scheme; The initial material allocation scheme is iterated and optimized until the energy storage density and ultimate rotational speed targets are met.

[0022] It should be further explained that: Obtain the density of various candidate materials Ultimate strength Performance parameters such as elastic modulus (E) are used as the basis for material selection.

[0023] Strength requirement: Based on the maximum expected stress value of the flywheel energy storage system rotor during operation. Select the strength limit The material is [specifically, the type of material]. This ensures that the rotor will not break down due to insufficient material strength during operation.

[0024] To optimize rotor quality, materials with lower density are prioritized while meeting strength requirements, and a density threshold is set. Choose density from materials that meet strength requirements. Materials.

[0025] Considering the deformation coordination of the rotor during rotation, a combination of materials with similar elastic moduli is selected to avoid localized stress concentration caused by excessive differences in elastic moduli. This is achieved by calculating the ratio of the difference in elastic moduli to the average value of the materials. ( and (This represents different materials) and sets an allowable ratio threshold to filter material combinations.

[0026] Taking into account the above criteria, n suitable materials are selected from the numerous candidate materials and labeled as follows: Their densities are respectively The strength limits are respectively The elastic moduli are respectively .

[0027] A geometric model of a uniform disk rotor was created using the finite element analysis software ANSYS. The model was created based on the actual dimensions of the rotor (diameter D and thickness H).

[0028] Assign material properties to the model, including density, elastic modulus, and Poisson's ratio, based on the actual material used.

[0029] The geometric model is meshed, discretizing the continuous model into a finite number of elements and nodes. The mesh type and size are selected. The mesh size is determined based on the complexity of the model and the required computational accuracy, and a suitable mesh density is determined through convergence analysis. Initially, a coarser mesh is used for calculation, and then the mesh is gradually refined, observing the changes in the calculated stress values. When the results tend to stabilize, the corresponding mesh density is the suitable mesh density.

[0030] Boundary conditions are set according to the rotor's operating conditions. For rotors mounted on a shaft, axial and radial displacement constraints are applied at the contact point between the rotor's central hole and the shaft to simulate an actual fixed support state. The rotor's rotational speed load is set, and the rotational speed is... (Revolutions per minute) converted to angular velocity And apply it to the entire rotor model.

[0031] Run a finite element analysis to calculate the stress distribution of the rotor under specified operating conditions. Obtain the stress values ​​at each node, including radial stress. and tangential stress

[0032] , , Where r is the radius from the node to the center of rotation, and A and C are constants determined according to the boundary conditions. A stress contour map is generated, with different colors representing different stress magnitudes.

[0033] The stress region is divided according to the stress magnitude represented by the colors in the stress cloud map. A stress range threshold is set. The stress value is in ( Nodes within the range of minimum stress are divided into regions. Simultaneously, considering process feasibility, if the division of a region leads to excessive manufacturing complexity, the region division needs to be appropriately adjusted.

[0034] The rotor is divided into m distinct regions, each labeled as follows: Record node information or describe it through geometry within each region. Regularly shaped regions are described using parameters such as center coordinates, radius, and angle, while irregularly shaped regions are described using boundary node coordinates.

[0035] Calculate the volume of each region Disk-shaped area According to the formula for the volume of a disk Calculate the volume. Wherein, It is a region The corresponding equivalent disk diameter, It refers to the thickness of that region. If the region is part of the entire rotor disk, its diameter and thickness may be a certain proportion of the original rotor diameter and thickness.

[0036] The maximum stress in the calculation area at the operating speed.

[0037] For the region At rotation speed The stress calculation formula is determined based on its geometry.

[0038] According to elasticity, the radial stress in the disk region and tangential stress The calculation formula is: , , Set up a region The allocated materials are ,make ,in It is a material The strength limit. To minimize rotor mass, among the various available materials that meet the stress requirements, materials with lower density are preferred. All candidate materials are examined. For each region The maximum stress when using each material is calculated and compared with the material's ultimate strength, while also taking into account the material density, to ultimately determine a suitable material allocation scheme.

[0039] Calculate each region Use appropriate materials The quality after According to the formula ,in It is a material density, It is a region The volume of each region is calculated after using the allocated materials, which is then used for subsequent calculation of the total rotor mass and performance evaluation.

[0040] The XGBoost algorithm was used to optimize the rotor mass distribution.

[0041] Construct a feature set for training the XGBoost model.

[0042] Assuming there is a total Candidate materials ,right each region To allocate materials, a material allocation scheme can be implemented using a length of... Integer vector representation, such as ,in The value ranges from 1 to n, representing a region. The allocated materials are This vector is added to the dataset as a feature column.

[0043] Calculate each region volume Circular area The volume values ​​of each region are combined into a feature vector. Add to dataset.

[0044] For each region, calculate its theoretical average radius. , , Calculate an estimated value of the foundation stress. , forming feature vectors Add to dataset.

[0045] The goal is to optimize material allocation so that the energy storage density and limiting rotational speed meet the requirements. The following two quantities are used as target variables.

[0046] Energy density difference label Calculate the initial energy storage density corresponding to each material allocation scheme. From the formula Calculate the moment of inertia. , It is the difference between the target energy storage density and the actual energy storage density, and this difference should be made close to 0.

[0047] Limiting speed difference label The limiting speed under the current material distribution scheme was calculated using rotor dynamics formulas. , It is the difference between the target speed limit and the actual speed limit, making this difference approach 0.

[0048] By generating a large number of different material allocation scheme examples, a dataset containing multiple samples was constructed. , in, This represents the feature vector of the i-th sample. This represents a tag vector consisting of the corresponding energy storage density difference and the limiting speed difference.

[0049] Determine the key parameters of the XGBoost model.

[0050] Number of trees : Indicates the number of decision trees in the model. It is determined by selecting an appropriate value within a certain range through cross-validation and observing the model's performance on the validation set.

[0051] Learning rate The learning rate controls the step size of the model's learning in each iteration. A smaller learning rate may make training more stable but convergence slower, and needs to be determined according to the actual situation.

[0052] Maximum depth of tree Limiting the growth depth of the decision tree and avoiding overfitting requires adjustments and optimizations based on the performance of the validation set.

[0053] Other parameters include (Minimum weight of child nodes, used to control the growth of the tree) (Minimum loss reduction required for node splitting), using the default value, and then gradually fine-tuning and optimizing.

[0054] The constructed dataset D is divided into a training set according to a certain ratio. Validation set and test set The common division ratio is For training, Used for verification, Used for testing, for example, if dataset D has N samples, then the number of samples in the training set... Number of validation set samples Number of test set samples .

[0055] Using the training set The goal of training the XGBoost model is to minimize the objective function: , , in, The loss function is constructed using the mean squared error (MSE) for both labels and then summed. This applies to the energy storage density difference label. and model predictions and the limit speed difference label and predicted value ; The loss function can be expressed as:

[0056] During training, the XGBoost algorithm, based on the idea of ​​gradient boosting, constructs a new decision tree in each iteration by fitting the negative gradient between the current model's predicted value and the true label (when the loss function is differentiable), continuously updating the model to reduce the loss function value. The model is monitored on the validation set. The model's performance on the validation set is evaluated (by calculating the loss value and relevant evaluation metrics on the validation set). The model parameters are then adjusted based on the validation set performance to prevent overfitting, until the model achieves good performance on the validation set or reaches the set maximum number of training epochs or other stopping conditions.

[0057] Using a trained XGBoost model, predictions are made for newly generated or evaluated material allocation schemes to obtain the predicted energy density difference. Difference between and limit speed The goal of prediction is to find... Approaching 0 and A material allocation scheme that approaches zero.

[0058] Material allocation is optimized and adjusted based on model predictions. A method is used to calculate the maximum stress in each region and compare it with the corresponding material strength limit to ensure... A new set of candidate material allocation schemes is generated using a specific strategy. These newly generated candidate schemes are converted into feature vectors and input into a trained XGBoost model for prediction, yielding their corresponding predicted differences in energy density and maximum speed. The scheme is then selected to achieve the optimal combination of predicted differences in energy density and maximum speed. The minimum material allocation scheme will serve as the basis for the next iteration.

[0059] Repeat the steps of generating candidate material allocation schemes, model evaluation, and selection to optimize material allocation in a direction that brings energy storage density and limiting rotational speed closer to the target values, while always ensuring that stress requirements are met. Continue this process until one of the following termination conditions is met: (1) Reach the maximum number of iterations: Set a maximum number of iterations. When the optimization process reaches this number of iterations, stop the optimization and output the current optimal material allocation scheme.

[0060] (2) Meet accuracy requirements: Define an accuracy threshold for the difference in energy storage density and the difference in limiting rotational speed. When the predicted difference in energy storage density is less than 100%, the threshold is set at 100%. And the difference in limiting speeds is less than When sufficient optimization accuracy has been achieved, the optimization is stopped and the result is output.

[0061] Through the above steps and related calculations based on the XGBoost algorithm, the material allocation scheme is continuously optimized and adjusted using the model, ultimately finding the optimal material allocation scheme that meets the target energy storage density and limit speed requirements while also taking into account stress constraints.

[0062] Example 2, an embodiment of the present invention, provides a flywheel energy storage system optimized using hybrid materials, including: a data collection and screening module, a finite element analysis and property modeling module, an XGBoost optimization and material allocation optimization module, and an iterative optimization and scheme evaluation module; The data collection and screening module collects physical property data of various candidate materials and screens out materials that meet the requirements. The finite element analysis and property modeling module performs geometric modeling of the flywheel rotor based on the selections in the material list, and uses the finite element analysis method to calculate the stress distribution of the rotor under working conditions. The XGBoost optimization and material allocation optimization module constructs a feature dataset of material allocation schemes, uses the XGBoost model to predict the energy storage density difference and the limit speed difference, optimizes and adjusts the model based on the training set, iteratively improves the material allocation scheme, and finally makes the target value close to the ideal value. The iterative optimization and scheme evaluation module generates a new set of material allocation scheme candidates by comparing the maximum stress with the strength limit, and inputs them into the XGBoost model for prediction, calculation and evaluation of the performance of each material allocation scheme.

[0063] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0065] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0066] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0067] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing flywheel energy storage using hybrid materials, characterized in that, include: Collect data to screen candidate materials and generate a material list; Based on the materials in the material list, property modeling is performed, and the rotor is divided into multiple stress regions through finite element analysis. Calculate the volume of each stress region and the maximum stress of each region at the operating speed, allocate materials, and generate an initial material allocation scheme; The initial material allocation scheme is iterated and optimized until the energy storage density and ultimate rotational speed targets are met.

2. The method for optimizing flywheel energy storage using hybrid materials as described in claim 1, characterized in that: The collected data includes the density, ultimate tensile strength, and elastic modulus of various candidate materials. Based on the maximum expected stress value of the flywheel energy storage system rotor during operation, materials that meet the strength limit are selected. Then, materials with the lowest density are prioritized. By calculating the ratio of the difference in elastic modulus of the materials to the average value, a ratio threshold is set to screen material combinations and draw a material list.

3. The method for optimizing flywheel energy storage using hybrid materials as described in claim 2, characterized in that: The property modeling based on materials in the material list involves constructing a geometric model of a uniform disk rotor based on the selected materials, meshing the geometric model, discretizing the continuous model into a finite number of elements and nodes, determining the mesh type and size, setting boundary conditions according to the rotor's working conditions, running finite element analysis, calculating the stress distribution of the rotor under the set working conditions, generating a stress cloud map, and dividing the region according to the stress magnitude represented by the color in the stress cloud map.

4. The method for optimizing flywheel energy storage using hybrid materials as described in claim 3, characterized in that: The mesh type and size are determined based on the complexity of the model and the required computational accuracy, and a suitable mesh density is determined through convergence analysis. Different colors in the stress cloud diagram represent different stress magnitudes.

5. The method for optimizing flywheel energy storage using hybrid materials as described in claim 4, characterized in that: The initial material distribution scheme is generated by calculating the corresponding volume and the maximum stress at the corresponding operating speed based on the region in the stress cloud diagram. The maximum stress should be less than the material's strength limit. Select the material with the lowest density, iterate through all candidate materials, and generate a suitable initial material allocation scheme.

6. The method for optimizing flywheel energy storage using hybrid materials as described in claim 5, characterized in that: The goal of optimizing material allocation until the energy storage density and the ultimate rotational speed are met is to input the initial material allocation scheme into the XGBoost model, allocate several candidate materials to the region, and use an integer vector of one length to represent a material allocation scheme, and add the vector as a feature column to the dataset. Calculate the volume of each region, and combine the volume values ​​of each region into a feature vector and add it to the dataset; Calculate the average radius for each region and output an estimated base stress value. The feature vectors are added to the dataset; By generating a large number of different material allocation scheme examples, a dataset containing multiple samples is constructed; The target variables are constructed by including labels for energy density difference and limit speed difference, forming a label vector.

7. The method for optimizing flywheel energy storage using hybrid materials as described in claim 6, characterized in that: The optimization of material allocation until the energy storage density and ultimate speed targets are met includes determining the parameters of the XGBoost model and dividing the constructed dataset into training set, validation set and test set according to the proportions. The XGBoost model is trained using the training set, and the objective function is minimized, expressed as: ， ， ， in, It is a loss function, which is constructed by summing the loss functions for the energy density difference label and the limiting speed difference label using the mean square error. , These are the energy storage density difference labels and the XGBoost model predictions, respectively. These are the labels for the difference in extreme speeds and the predicted values ​​from the XGBoost model, respectively. Using the trained XGBoost model, the newly generated or evaluated material allocation schemes are predicted, and the output material allocation schemes with energy density difference and limit speed difference both approaching 0 are generated. The material allocation is optimized and adjusted, the maximum stress in each region is calculated and compared with the corresponding material strength limit, and a new set of material allocation scheme candidates is generated. The newly generated candidate material allocation schemes are converted into feature vectors and input into the trained XGBoost model for prediction. This yields the corresponding predicted energy density difference and limiting speed difference. The scheme is then selected to achieve the optimal combination of predicted energy density difference and limiting speed difference. The minimum material allocation scheme will serve as the basis for the next iteration. Repeat the steps of generating candidate material allocation schemes, model evaluation and selection, and optimize material allocation in the direction that makes energy storage density and limiting rotational speed closer to the target values. Stop repeating when the maximum number of iterations reaches the requirement or the accuracy meets the requirement, and output the final optimal material allocation scheme.

8. A system employing a method for optimizing flywheel energy storage using hybrid materials as described in any one of claims 1 to 7, characterized in that: It includes modules for data collection and filtering, finite element analysis and property modeling, XGBoost optimization and material allocation optimization, and iterative optimization and scheme evaluation. The data collection and screening module collects physical property data of various candidate materials and screens out materials that meet the requirements. The finite element analysis and property modeling module performs geometric modeling of the flywheel rotor based on the selections in the material list, and uses the finite element analysis method to calculate the stress distribution of the rotor under working conditions. The XGBoost optimization and material allocation optimization module constructs a feature dataset of material allocation schemes, uses the XGBoost model to predict the energy storage density difference and the limit speed difference, optimizes and adjusts the model based on the training set, iteratively improves the material allocation scheme, and finally makes the target value close to the ideal value. The iterative optimization and scheme evaluation module generates a new set of material allocation scheme candidates by comparing the maximum stress with the strength limit, and inputs them into the XGBoost model for prediction, calculation and evaluation of the performance of each material allocation scheme.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of any one of claims 1 to 7 for a method of optimizing flywheel energy storage using hybrid materials.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of any one of claims 1 to 7 of the method for optimizing flywheel energy storage using hybrid materials.

Images