Structure optimization method and system based on bionics, electronic equipment and medium

Abstract

The invention discloses an optimization-based structure optimization method and system, electronic equipment and a storage medium, and the method comprises the steps: carrying out the three-dimensional modeling according to preset geometric parameters, and obtaining an initial three-dimensional model; performing bionic processing on the initial three-dimensional model according to a preset structure requirement and a topological optimization target to obtain a plurality of initial bionic structure models; thirdly, finite element analysis is conducted on the multiple initial bionic structure models according to preset engineering requirements, and a first bionic structure model is determined; performing simulation analysis according to a preset algorithm, a preset working condition, the first bionic structure model and the plurality of initial bionic structure models, and determining a sample data set; performing parameter optimization on the first bionic structure model according to the sample data set and a preset model, determining a target optimization parameter combination, and determining an optimized structure model according to the target optimization parameter combination as a final optimization result; the optimization efficiency can be improved, and the optimization precision is improved; the embodiment of the invention can be widely applied to the technical field of structure optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of structural optimization, and in particular to a bionic-based structural optimization method, system, electronic equipment and storage medium. Background Art

[0002] The wheel hub is the "heart" of the rail transit system, and its performance directly affects the various performance and safety of the vehicle. Continuous optimization of the wheel hub can provide key support for the intelligent and sustainable development of the rail transit system. Traditional optimization methods use proxy model optimization methods to optimize existing wheel hub solutions. However, traditional proxy model optimization methods require a large number of samples for model training. The trained proxy model is used to perform inference and prediction on the existing wheel hub solution to achieve optimization. The training cost of the proxy model is high, the training time is long, and the inference and prediction accuracy is not high. Therefore, the traditional solution has low optimization efficiency and low optimization accuracy. Summary of the Invention

[0003] The main purpose of the embodiments of the present invention is to provide a bionic-based structural optimization method, system, electronic device and storage medium, which can improve optimization efficiency and improve optimization accuracy.

[0004] To achieve the above objectives, an embodiment of the present invention provides a bionics-based structural optimization method, the method comprising:

[0005] Establish a model according to preset geometric parameters to obtain an initial three-dimensional model;

[0006] Performing bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain a plurality of initial bionic structural models;

[0007] Performing finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; performing simulation analysis on the first bionic structure model and the plurality of initial bionic structure models according to a preset algorithm, a preset working condition, to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set;

[0008] Parameters of the first bionic structure model are optimized according to the sample data set and the preset model, a target optimization parameter combination is determined, and an optimization result is determined according to the target optimization parameter combination.

[0009] In some embodiments, performing simulation analysis based on a preset algorithm, preset working conditions, the first bionic structure model, and several of the initial bionic structure models to determine a sample data set specifically includes:

[0010] Performing simulation analysis on the initial bionic structure models according to the preset working conditions to determine a bionic sample data set;

[0011] Calculating the first bionic structure model according to the preset algorithm to generate several groups of shape variable parameters;

[0012] Performing simulation analysis on the first bionic structure model according to the preset working conditions and the plurality of groups of shape variable parameters to determine an optimized sample data set;

[0013] The sample data set is determined according to the bionic sample data and the optimized sample data set.

[0014] In some embodiments, the simulation analysis of the plurality of initial bionic structure models according to the preset working conditions to determine the bionic sample data set specifically includes:

[0015] performing stress analysis on the plurality of initial bionic structure models according to the preset working conditions to determine a first stress data set;

[0016] performing mass calculations on a plurality of the initial bionic structure models respectively to determine a second mass data set;

[0017] The bionic sample dataset is determined according to the first stress dataset and the second mass dataset.

[0018] In some embodiments, the simulation analysis of the first bionic structure model according to the preset working conditions and the plurality of sets of shape variable parameters to determine the optimized sample data set specifically includes:

[0019] Adjusting the first bionic structure model according to the plurality of groups of shape variable parameters to determine a plurality of second bionic structure models;

[0020] performing stress analysis on the plurality of second bionic structure models according to the preset working conditions to determine a second stress data set, and performing mass calculation on the plurality of second bionic structure models to determine a second mass data set;

[0021] The optimized sample dataset is determined according to the second stress dataset and the second quality dataset.

[0022] In some embodiments, the performing parameter optimization on the first bionic structure model according to the sample data set and the preset model to determine a target optimization parameter combination specifically includes:

[0023] Training the preset model according to the sample data set to determine an optimized model;

[0024] According to the matching between the first bionic structure model and the sample data set, a plurality of sets of target shape variable parameters are determined; wherein the target shape variable parameters are design shape parameters corresponding to the first bionic structure model;

[0025] Optimizing and calculating several groups of target shape variable parameters according to the optimization model to determine the target optimization parameter combination.

[0026] In some embodiments, the training of the preset model based on the sample data set to determine the optimized model specifically includes:

[0027] Preprocessing the sample data set to obtain a standard data set, and dividing the standard data set according to a preset ratio to determine a training set, a validation set, and a test set;

[0028] The preset model is trained according to the training set and the validation set to obtain an intermediate model; a model accuracy value is obtained by performing calculations according to the test set and the intermediate model; and the model accuracy value is compared with a preset threshold;

[0029] If the model accuracy value is greater than or equal to the preset threshold, the intermediate model is used as the optimization model;

[0030] If the model accuracy value is less than the preset threshold, the parameters of the intermediate model are adjusted, and the intermediate model after parameter adjustment is used as the preset model, and the process of training the preset model according to the training set and the validation set is returned to obtain the intermediate model; until the model accuracy value is greater than or equal to the preset threshold.

[0031] In some embodiments, the method further comprises:

[0032] Analyzing the optimization results to determine working condition response data, and performing image recognition on the optimization results to obtain bionic structure image data;

[0033] The bionic sample data set in the sample data set is supplemented according to the bionic structure image data and the working condition response data.

[0034] To achieve the above objectives, another aspect of an embodiment of the present invention provides a bionics-based structural optimization system, the system comprising:

[0035] The first module is used to build a model according to preset geometric parameters to obtain an initial three-dimensional model;

[0036] The second module is used to perform bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain a plurality of initial bionic structural models;

[0037] The third module is used to perform finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; perform simulation analysis based on a preset algorithm, preset working conditions, the first bionic structure model, and the plurality of initial bionic structure models to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set;

[0038] The fourth module is used to optimize the parameters of the first bionic structure model according to the sample data set and the preset model, determine the target optimization parameter combination, and determine the optimization result according to the target optimization parameter combination.

[0039] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned method when executing the computer program.

[0040] To achieve the above objectives, another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0041] The implementation of the embodiments of the present invention includes the following beneficial effects: the embodiments of the present invention provide a bionic-based structural optimization method, system, electronic device and storage medium, which performs three-dimensional modeling according to preset geometric parameters to obtain an initial three-dimensional model; the initial three-dimensional model is biomimetically processed according to preset structural requirements and topology optimization goals to obtain several different initial bionic structural models; then, finite element analysis is performed on the several initial bionic structural models according to preset engineering requirements to determine the first bionic structural model; then, simulation analysis is performed according to the preset algorithm, preset working conditions, the determined first bionic structural model and several initial bionic structural models to determine a sample data set; parameter optimization is performed on the determined first bionic structural model according to the sample data set and the preset model to determine the target optimization parameter combination, and the optimized structural model is determined according to the target optimization parameter combination as the final optimization result. Through bionic processing, a variety of bionic structural models are provided, and then the generated bionic structures are simulated and analyzed based on engineering requirements and preset working conditions, the optimal bionic structural model is screened and the corresponding sample data is generated for model training. The optimal bionic structural model is optimized based on the trained model, enriching the diversity of sample data, reducing the number of samples for model training, reducing training costs and time, and improving overall optimization efficiency; at the same time, sample data is generated based on engineering requirements and preset working conditions, and the diversity of data is enriched through bionic processing, thereby improving model prediction accuracy and thus improving optimization accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic flow chart of the steps of a bionic-based structural optimization method provided by an embodiment of the present invention;

[0043] Figure 2 This is a schematic flow chart of the steps of determining a sample data set in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0044] Figure 3 This is a schematic flow chart of the steps of determining a bionic sample data set in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0045] Figure 4 This is a schematic flow chart of the steps of determining an optimized sample data set in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0046] Figure 5 This is a schematic flow chart of the steps for determining a target optimization parameter combination in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0047] Figure 6 This is a schematic flow chart of the steps of determining an optimization model in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0048] Figure 7 This is a schematic flow chart of the steps of enriching a bionic sample data set in a bionic-based structural optimization method provided by an embodiment of the present invention;

[0049] Figure 8 This is a schematic diagram of a specific embodiment of the present invention;

[0050] Figure 9 is a schematic diagram of a bionic structure model obtained in a specific embodiment provided by an embodiment of the present invention;

[0051] Figure 10 3D structural diagram of a bionic water droplet weight loss hole model 4 in a specific embodiment provided by an embodiment of the present invention;

[0052] Figure 11 Schematic diagram of a free vertex in a bionic water droplet weight loss hole model 4 in a specific embodiment provided by an embodiment of the present invention;

[0053] Figure 12 is a schematic diagram of calculation parameters for different road conditions in a specific embodiment provided by an embodiment of the present invention;

[0054] Figure 13 This is a flowchart of the steps for training a deep learning model in a specific embodiment provided by an embodiment of the present invention;

[0055] Figure 14 This is a structural diagram of an optimized wheel weight-reducing hole shape in a specific embodiment provided by an embodiment of the present invention;

[0056] Figure 15 It is a schematic diagram of a stress cloud diagram of an optimized bionic water drop weight-reducing hole model 4 under a curve working load in a specific embodiment provided by an embodiment of the present invention;

[0057] Figure 16 This is a flowchart of the steps of adding the final optimization solution to the release sample data set to generate a bionic structure in a specific embodiment provided by an embodiment of the present invention;

[0058] Figure 17 This is a structural block diagram of a bionic-based structural optimization system provided by an embodiment of the present invention;

[0059] Figure 18 The figure is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The step numbers in the following embodiments are provided for ease of description only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted based on the understanding of those skilled in the art.

[0061] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0062] In the following description, the terms "first\second\third" are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It is understandable that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present invention described herein can be implemented in an order other than that illustrated or described herein.

[0063] Unless otherwise defined, all technical and scientific terms used in the embodiments of the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention pertains. The terms used in the embodiments of the present invention are for the purpose of describing the embodiments of the present invention only and are not intended to limit the present invention.

[0064] Figure 1 The embodiment of the present invention provides an optional flow chart of a bionic-based structural optimization method. Figure 1 The method may include but is not limited to steps S101 to S104.

[0065] Step S101, building a model according to preset geometric parameters to obtain an initial three-dimensional model;

[0066] Step S102, performing bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain several initial bionic structural models;

[0067] Step S103: performing finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; performing simulation analysis based on a preset algorithm, preset working conditions, the first bionic structure model, and the plurality of initial bionic structure models to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set;

[0068] Step S104 , performing parameter optimization on the first bionic structure model according to the sample data set and the preset model, determining a target optimization parameter combination, and determining an optimization result according to the target optimization parameter combination.

[0069] In the steps S101 to S104 shown in the embodiment of the present application, three-dimensional modeling is performed according to the geometric parameters of the wheel hub model to be optimized and the corresponding modeling requirements to obtain an unoptimized initial three-dimensional model; then, based on the unoptimized initial three-dimensional model, based on the bionic principle, bionic topological processing is performed according to the actual wheel hub structural requirements and topological optimization goals to generate different bionic topological structures and obtain a finite number of initial bionic structural models; these initial bionic structural models are obtained by performing preliminary structural optimization of the unoptimized initial three-dimensional model through the bionic principle, and while meeting the structural requirements, they have certain design requirements and performance requirements; then, finite element analysis is performed on the finite number of initial bionic structural models, and according to the actual engineering technical requirements, a finite number of initial bionic structural models are obtained. An optimal bionic structure model is selected from the bionic structure models as the first bionic structure model; at the same time, a finite number of initial bionic structure models are simulated and analyzed according to several preset wheel hub working conditions and a preset optimization algorithm to obtain analysis data of a finite number of bionic structure models under different working conditions, and the obtained analysis data is used as a sample data set; a preset deep learning model is trained and tested according to the obtained sample data set to obtain a trained deep learning model; the trained deep learning model is then used to optimize the parameters of the selected first bionic structure model to obtain a corresponding optimized structural parameter combination; the first bionic structure model is further optimized according to the optimized structural parameter combination to obtain a final optimized bionic structure model. In this embodiment, the optimized structural parameter combination needs to be rounded according to production requirements before further optimizing the first bionic structure model.

[0070] In step S101 of some embodiments, modeling can be performed according to preset geometric parameters to obtain an initial three-dimensional model. Alternatively, the existing three-dimensional model can be input in the form of an electronic file in other ways, without limitation thereto.

[0071] See also Figure 2 In some embodiments, step S103 may include but is not limited to steps S201 to S204:

[0072] Step S201, performing simulation analysis on a plurality of initial bionic structure models according to preset working conditions to determine a bionic sample data set;

[0073] Step S202, calculating the first bionic structure model according to a preset algorithm to generate several sets of shape variable parameters;

[0074] Step S203, performing simulation analysis on the first bionic structure model according to preset working conditions and several sets of shape variable parameters to determine an optimized sample data set;

[0075] Step S204 : determining a sample data set according to the bionic sample data and the optimized sample data set.

[0076] In step S201 of some embodiments, several working conditions are selected as preset working conditions based on the working conditions that the wheel hub may encounter in actual applications, such as the working condition with the highest probability of occurrence, the working condition with the greatest impact on the wheel hub structure, etc.; based on the preset working conditions, the generated finite number of initial bionic structural models are simulated according to the corresponding working conditions, and the response values ​​corresponding to various model parameters of the finite number of initial bionic structural models under the preset working conditions are analyzed as a bionic sample data set.

[0077] In step S202 of some embodiments, a preset optimization algorithm is used to optimize the corresponding shape parameters of the selected first bionic structure model to obtain several groups of shape design parameters under the bionic topological structure corresponding to the first bionic structure model; the first bionic structure model is structurally designed according to different shape design parameters to obtain different bionic models that meet the same bionic topological structure; in this embodiment, a free vertex control method is used to optimize the parameters of the area to be optimized in the selected first bionic structure model to obtain several groups of different shape variable parameters; such as several groups of different shape variable parameters corresponding to the teardrop-shaped structure, including the area of ​​the teardrop-shaped structure, the focal coordinates, etc.

[0078] In step S203 of some embodiments, a preliminary structural optimization is performed on the first bionic structural model based on the obtained several groups of different shape variable parameters to obtain several different structural models; then, the obtained several different structural models are simulated and analyzed according to preset working conditions, and the response values ​​corresponding to the model parameters of the several different structural models under the preset working conditions are recorded as optimization sample data sets.

[0079] In step S204 of some embodiments, the previously obtained bionic sample data set and optimized sample data set are aggregated to obtain a sample data set for subsequent training of a deep learning model to learn the relationship between different bionic structural models and working conditions and shape parameters, and then make corresponding adjustments to the areas that need to be optimized in the input bionic structural model.

[0080] See also Figure 3 In some embodiments, step S201 may include but is not limited to steps S301 to S303:

[0081] Step S301, performing stress analysis on a plurality of initial bionic structure models according to preset working conditions to determine a first stress data set;

[0082] Step S302, performing mass calculation on a plurality of initial bionic structure models to determine a second mass data set;

[0083] Step S303 : determining a bionic sample data set according to the first stress data set and the second mass data set.

[0084] In step S301 of some embodiments, the optimized design of the wheel hub needs to consider whether the overall performance requirements of the optimized wheel hub meet the corresponding requirements and have a certain lightweight design to meet the production needs in actual applications; based on the bionic principle, the initial three-dimensional model is biomimetically processed according to the structural requirements and topology optimization goals to obtain several different initial bionic structural models, and then the corresponding working conditions are simulated for several different initial bionic structural models according to a plurality of preset working conditions, and the corresponding stress response values ​​of each initial bionic structural model under different working conditions are recorded to analyze and determine whether each bionic structural model meets the actual engineering requirements.

[0085] In step S302 of some embodiments, for the obtained several different initial bionic structure models, the corresponding mass sizes are calculated to determine whether the corresponding lightweight requirements are met, thereby obtaining mass data corresponding to each initial bionic structure model.

[0086] In step S303 of some embodiments, the stress response values ​​and quality data obtained above are summarized and classified according to the type of the initial bionic structure model to obtain a bionic sample data set.

[0087] See also Figure 4 In some embodiments, step S203 may include but is not limited to steps S401 to S403:

[0088] Step S401, adjusting the first bionic structure model according to several groups of shape variable parameters to determine several second bionic structure models;

[0089] Step S402: performing stress analysis on the plurality of second bionic structure models according to the preset working conditions to determine a second stress data set, and performing mass calculation on the plurality of second bionic structure models to determine a second mass data set;

[0090] Step S403: determining an optimized sample data set according to the second stress data set and the second quality data set.

[0091] In step S401 of some embodiments, parameters of the area to be optimized in the first target bionic structure model obtained by screening are optimized by an optimization algorithm to obtain several groups of different shape variable parameters; parameters of the first target bionic structure model obtained by screening are adjusted according to the several groups of different shape variable parameters to obtain several different second bionic structure models; each second bionic structure model belongs to the same bionic topological structure, but the area to be optimized in the bionic structure model has different shape parameters.

[0092] In step S402 of some embodiments, corresponding stress analysis is performed on several different second bionic structure models obtained according to multiple preset working conditions, and the stress response value corresponding to each second bionic structure model under different working conditions is recorded; at the same time, the mass of each second bionic structure model is calculated to obtain corresponding mass data.

[0093] In step S403 of some embodiments, the stress response values ​​obtained by stress analysis of each second bionic structure model and the quality data of each second bionic structure model are summarized and processed to obtain an optimized sample data set for subsequent training of a deep learning model to learn the relationship between different shape variable parameters, bionic structures, and working conditions.

[0094] See also Figure 5 In some embodiments, step S104 may include but is not limited to steps S501 to S503:

[0095] Step S501: training a preset model based on a sample data set to determine an optimized model;

[0096] Step S502, determining a plurality of sets of target shape variable parameters by matching the first bionic structure model with the sample data set; wherein the target shape variable parameters are design shape parameters corresponding to the first bionic structure model;

[0097] Step S503 , performing optimization calculations on several groups of target shape variable parameters according to the optimization model to determine a target optimization parameter combination.

[0098] In step S501 of some embodiments, after obtaining a sample data set, the deep learning model is trained using the sample data set so that the deep learning model learns the relationship between relevant parameters of different bionic topological structures in the sample data set and factors such as working conditions, and obtains a corresponding optimization model; in this embodiment, the deep learning model can also be used to generate different bionic structure models, and according to different actual optimization requirements, simulation analysis is performed based on the generated different bionic structure models, and corresponding sample data sets are collected, and the sample data sets are used to fine-tune the parameters of the deep learning model so as to subsequently optimize the bionic structure model.

[0099] In step S502 of some embodiments, after completing the training of the deep learning model, an optimization model is obtained; the first target bionic structure model obtained by finite element analysis is matched in the sample data set, and the corresponding shape variable parameters are screened out as input parameters of the optimization model.

[0100] In step S503 of some embodiments, the shape variable parameters corresponding to the first target bionic structure model are input into the optimization model as input parameters of the optimization model; the shape variable parameters corresponding to the first target bionic structure model are evaluated and calculated using the optimization model, and the best optimization solution is selected.

[0101] See also Figure 6 In some embodiments, step S501 may include but is not limited to steps S601 to S604:

[0102] Step S601: pre-process the sample data set to obtain a standard data set, and divide the standard data set according to a preset ratio to determine a training set, a validation set, and a test set;

[0103] Step S602: Train the preset model based on the training set and the validation set to obtain an intermediate model; perform calculations based on the test set and the intermediate model to obtain a model accuracy value; and compare the model accuracy value with a preset threshold.

[0104] Step S603: If the model accuracy value is greater than or equal to the preset threshold, the intermediate model is used as the optimization model;

[0105] In step S604, if the model accuracy value is less than the preset threshold, the parameters of the intermediate model are adjusted, and the intermediate model after parameter adjustment is used as the preset model, and the process returns to execute the training of the preset model according to the training set and the validation set to obtain the intermediate model; until the model accuracy value is greater than or equal to the preset threshold.

[0106] In step S601 of some embodiments, the obtained sample data set is preprocessed to improve the data quality of the sample data set. In this embodiment, before training the preset deep learning model based on the sample data set, the sample data set is cleaned to delete missing values, error values, etc. in the sample data set. Then, the cleaned sample data set is normalized. The specific data normalization is performed according to the following formula:

[0107]

[0108] Among them, X * is the normalized value, X is the sample data in the sample data set, X min is the minimum value in the sample data set, X max is the maximum value in the sample data set. After preprocessing the sample data set, the preprocessed sample data set is divided according to a preset ratio to obtain a training set, a validation set, and a test set.

[0109] In step S602 of some embodiments, the preset deep learning model is trained according to the divided training set and validation set, and an intermediate model is obtained after the training is completed; the performance of the intermediate model obtained is tested using the test set to determine whether the intermediate model after training meets the set accuracy requirements, so as to determine whether it can be used for subsequent optimization processing by the user; in this embodiment, the RMSE value and R 2 The value is used to judge whether the prediction accuracy of the intermediate model meets the requirements.

[0110] In step S603 of some embodiments, the test set is input into the intermediate model for inference prediction to obtain corresponding prediction results, and the model accuracy value of the intermediate model is calculated based on the prediction results and the test set, and the calculated model accuracy value is compared with a preset threshold; if the model accuracy value is greater than or equal to the preset threshold, it is determined that the prediction accuracy of the intermediate model is qualified, and the current intermediate model can be used as an optimization model and put into subsequent optimization processing of the first target bionic structure model.

[0111] In step S604 of some embodiments, if the model accuracy value is less than the preset threshold, it means that the prediction accuracy of the intermediate model does not meet the requirements, and the parameters of the current intermediate model are adjusted, such as adjusting the learning rate, adjusting the number of training iterations, etc. The intermediate model after parameter adjustment is used as the preset model, and the preset model is trained using the training set and the validation set until the inference accuracy of the trained intermediate model meets the preset threshold.

[0112] See also Figure 7 In some embodiments, the bionic-based structural optimization method provided by the embodiment of the present invention may further include but is not limited to steps S701 to S702:

[0113] Step S701, analyzing the optimization results, determining the working condition response data, and performing image recognition on the optimization results to obtain bionic structure image data;

[0114] Step S702 : supplementing the bionic sample data set in the sample data set according to the bionic structure image data and the working condition response data.

[0115] In step S701 of some embodiments, the optimization model further optimizes the first target bionic structure model, selects the optimal shape variable parameter combination from several groups of shape variable parameters, and optimizes the first target bionic structure model based on the optimal shape variable parameter combination to obtain the final optimized structure; since it is the optimal structural solution under a specific bionic structure working condition, the optimal structural solution can be used as a supplement to the sample data to enrich the sample data set, thereby improving the diversity of bionic topological structures; therefore, the final optimization result is analyzed to determine the stress response value corresponding to the final optimization result under different working conditions, and image recognition is performed on the final optimization result to convert the three-dimensional model into corresponding image data to conform to the input data format of the optimization model or deep learning model.

[0116] In step S702 of some embodiments, the stress response values ​​corresponding to the final optimization results obtained by analysis and the data of the final optimization results after image recognition are summarized and analyzed to obtain corresponding sample data, and the sample data is supplemented to the sample data set to enrich the sample data set and improve the diversity of the generated topological bionic structure.

[0117] The following describes the solution of the embodiment of the present invention in detail with reference to specific application examples:

[0118] See also Figure 8, a bionic-based structural optimization method provided in an embodiment of the present application is applied to a wheel hub optimization program or terminal device to perform structural optimization design on an existing cavity wheel; relevant dimension data of the wheel hub and rim of the cavity wheel are input into the wheel hub optimization program or terminal device, corresponding modeling requirements are set, and the optimization program or terminal device performs modeling according to the input data to obtain an initial cavity wheel model. Specifically, the relevant dimension data of the wheel hub and rim of the cavity wheel are the outer diameter of the rolling circle of the wheel 860mm, the hub aperture 184mm, the hub length 180.5mm, and the inner hub-rim distance 12.5mm; then, based on the established initial cavity wheel model, according to the bionic principle, bionic processing is performed in accordance with the weight reduction structure requirements and topology optimization goals to obtain four bionic structural models, as shown in the following figure. Figure 9 As shown, it includes bionic water drop weight loss hole model 1, bionic water drop weight loss hole model 2, bionic water drop weight loss hole model 3, and bionic water drop weight loss hole model 4; Figure 9 The three-dimensional model of one of the bionic structure models is shown in Figure 10 As shown, according to Figure 10 As shown in the content, the bionic structure model is modified on the initial cavity wheel model based on the bionic principle to obtain a hub model with a bionic structure weight-reducing hole; the four bionic structure models obtained are subjected to finite element analysis to obtain the stress response values ​​of the four bionic structure models under the corresponding working conditions, and the mass data corresponding to the four bionic structure models are calculated to determine whether the weight-reduction requirements are met, and the stress response values ​​and mass data are used as bionic sample data sets; at the same time, the bionic water drop weight-reducing hole model 4 is selected as the optimal bionic structure model from the four bionic structure models through finite element analysis. This model has the lightest mass and can also meet the stress working condition requirements; the bionic water drop weight-reducing hole model 4 is parameterized by using the free vertex control method, and several free vertices are selected in the weight-reducing hole area that needs to be optimized in the bionic water drop weight-reducing hole model 4, such as Figure 11 As shown in the figure, the selected free vertices are vertex 31, vertex 32, vertex 33, vertex 34, vertex 35, vertex 36, vertex 37, vertex 38 and vertex 39; vertex 31 is taken as the free vertex to describe the shape of the optimized area of ​​the weight-reducing hole, and it is parameterized. The sample space is generated according to the value range, and the stress and mass of the bionic water drop weight-reducing hole model 4 under three road conditions are analyzed to obtain the optimized shape parameter combination. The shape parameters include X1, X2, X3, X4, X5, X6, X7, X8 and X9, which are respectively used as Figure 11 The optimization parameters of the free vertex shown in the figure are X. 10 Among them, along Figure 11 A pair of shape parameters symmetrical about the Y axis should be opposite to each other. In the optimization program or terminal device, such as Figure 12As shown in the figure, there are three different road working conditions according to the working condition load specified in the railway wheel standard, including straight line working condition load (F z1 ), Curve load (F z2 ,F y2 ) and turnout load (F z3 ,F y3 ), the optimization program or terminal device performs stress calculation on the bionic water drop weight reduction hole model 4 according to different road conditions. The specific calculation parameters are set as follows:

[0119] (1) Working condition 1: linear working condition, wheel-rail vertical force F z1 =1.25Pg / 2;

[0120] (2) Working condition 2: Curved working condition, wheel-rail vertical force F 22 =1.25Pg / 2, wheel-rail lateral force F y2 =0.7Pg / 2;

[0121] (3) Working condition 3: Turnout working condition, wheel-rail vertical force F z3 =1.25Pg / 2, wheel-rail lateral force F y3 =0.42Pg / 2.

[0122] Where: P is the axle weight, which is 17t; g is the acceleration due to gravity, which is 9.81m / s2.

[0123] The constraint conditions are set as follows: the stress in the spoke area under the three static working conditions does not exceed 150 MPa;

[0124] Set the variables as follows: the area of ​​the hub excluding the rim and axle, that is, the shape of the spokes;

[0125] The optimization goal is to achieve lightweighting of the total wheel weight under a total of three constraints in static working conditions.

[0126] The optimization program or terminal device calculates the bionic water droplet weight loss hole model 4 according to the constraints, optimization objectives and calculation parameters set above, obtains the stress response value of the bionic water droplet weight loss hole model 4 under different shape variable parameter combinations and different road working conditions, and at the same time calculates the corresponding quality data; the obtained data is used as the optimization sample data set.

[0127] See also Figure 13 , the obtained bionic sample data set and optimized sample data set are used to train the deep learning model, and after the training is completed, the RMSE value and R of the trained model are calculated. 2 Value, according to RMSE value and R 2 The accuracy of the model is judged by the value of RMSE < 0.2 and R 2When ≥0.9, the model accuracy is qualified and the optimized model is obtained. Specifically, during the model training process, the deformation value of each free vertex is input, the stress response value of different road conditions and the weight data of the bionic water drop weight reduction hole model 4 are calculated, and the accuracy of the model is calculated based on the output data to obtain the following table data:

[0128] Table 1

[0129] <![CDATA[R 2 Value]]> RMSE value Curve stress response value 0.9853 0.0268 Linear stress response value 0.9814 0.0262 Bending (turnout) stress response value 0.9976 0.01 weight 0.9987 0.0104

[0130] The optimization model is used to optimize the bionic water drop weight loss hole model 4, and the optimized parameter combination is obtained, as shown in the following table:

[0131] Table 2

[0132]

[0133] According to the parameter combination in the table, the final optimization solution can be obtained. In the final optimization solution, the shape of the optimized wheel weight reduction hole is as follows: Figure 14 As shown in the table, according to the data in the table, the maximum stress in the optimized area under the turnout working condition is 108.6MPa, the maximum stress under the straight working condition is 37.58MPa, the maximum stress under the curve working condition is 121.3MPa, and the mass is 259.8kg. Under the curve working condition load, the stress of the outer spoke is analyzed and the following is obtained: Figure 15 As shown in the stress cloud diagram, the maximum principal stress is 121.3 MPa, which is less than the design requirement of 150 MPa.

[0134] The performance of the wheel hub model obtained through pure bionic structure was compared with the performance of the optimal wheel hub model obtained through optimization: the stress of the bionic water drop weight reduction hole model 4 obtained through only bionic topology under three road conditions was 70.21MPa for the optimized area turnout condition, 22.71MPa for the straight condition, and 79.05MPa for the curved condition, with a mass of 321kg. The results are shown in the following table:

[0135] Table 3

[0136] Model Turnout working stress Linear stress Curve working condition stress quality Bionic structure model 70.21Mpa 22.71Mpa 79.05Mpa 321kg Optimized model 108.6Mpa 37.58Mpa 121.3Mpa 259.8kg Weight loss rate / / / 19.07％

[0137] See also Figure 16 The final optimization solution is added to the bionic sample data set for supplementation to obtain a richer bionic data set, thereby improving the diversity of bionic generated structures and obtaining structures that better meet engineering technology requirements, thereby improving the bionic accuracy and the efficiency of the overall optimization process.

[0138] The implementation of the embodiments of the present invention includes the following beneficial effects: the embodiments of the present invention provide a bionic-based structural optimization method, system, electronic device and storage medium, which performs three-dimensional modeling according to preset geometric parameters to obtain an initial three-dimensional model; the initial three-dimensional model is biomimetically processed according to preset structural requirements and topology optimization goals to obtain several different initial bionic structural models; then, finite element analysis is performed on the several initial bionic structural models according to preset engineering requirements to determine the first bionic structural model; then, simulation analysis is performed according to the preset algorithm, preset working conditions, the determined first bionic structural model and several initial bionic structural models to determine a sample data set; parameter optimization is performed on the determined first bionic structural model according to the sample data set and the preset model to determine the target optimization parameter combination, and the optimized structural model is determined according to the target optimization parameter combination as the final optimization result. Through bionic processing, a variety of bionic structural models are provided, and then the generated bionic structures are simulated and analyzed based on engineering requirements and preset working conditions, the optimal bionic structural model is screened and the corresponding sample data is generated for model training. The optimal bionic structural model is optimized based on the trained model, enriching the diversity of sample data, reducing the number of samples for model training, reducing training costs and time, and improving overall optimization efficiency; at the same time, sample data is generated based on engineering requirements and preset working conditions, and the diversity of data is enriched through bionic processing, thereby improving model prediction accuracy and thus improving optimization accuracy.

[0139] like Figure 17 As shown, an embodiment of the present invention further provides a bionic-based structural optimization system, which can implement the above-mentioned bionic-based structural optimization method, including:

[0140] The first module is used to build a model according to preset geometric parameters to obtain an initial three-dimensional model;

[0141] The second module is used to perform bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain a plurality of initial bionic structural models;

[0142] The third module is used to perform finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; perform simulation analysis based on a preset algorithm, preset working conditions, the first bionic structure model, and the plurality of initial bionic structure models to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set;

[0143] The fourth module is used to optimize the parameters of the first bionic structure model according to the sample data set and the preset model, determine the target optimization parameter combination, and determine the optimization result according to the target optimization parameter combination.

[0144] It can be seen that the contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0145] The present application also provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the aforementioned bionic-based structural optimization method. The electronic device can be any smart terminal, such as a tablet computer or an in-vehicle computer.

[0146] It can be understood that the contents of the above method embodiments are applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0147] See also Figure 18 , Figure 18 The hardware structure of an electronic device according to another embodiment is shown. The electronic device includes:

[0148] The processor 1801 can be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application.

[0149] The memory 1802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1802 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1802 and is called by the processor 1801 to execute a bionic-based structural optimization method according to the embodiments of this application.

[0150] Input / output interface 1803, used to implement information input and output;

[0151] Communication interface 1804, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0152] Bus 1805 , which transmits information between various components of the device (e.g., processor 1801 , memory 1802 , input / output interface 1803 , and communication interface 1804 );

[0153] The processor 1801 , the memory 1802 , the input / output interface 1803 and the communication interface 1804 are connected to each other in communication within the device via the bus 1805 .

[0154] Among them, the memory is a non-transient computer-readable storage medium that can be used to store non-transient software programs and non-transient computer executable programs. The memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory optionally includes a remote memory remotely arranged relative to the processor, and these remote memories can be connected to the processor via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0155] In addition, the embodiments of the present application further disclose a computer program product or computer program, which is stored in a computer-readable storage medium. The processor of a computer device can read the computer program from the computer-readable storage medium, and the processor executes the computer program, so that the computer device performs the above-mentioned method. Similarly, the contents of the above-mentioned method embodiment are all applicable to the present storage medium embodiment, and the functions specifically implemented by the present storage medium embodiment are the same as those of the above-mentioned method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned method embodiment.

[0156] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned bionic-based structural optimization method.

[0157] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiment, the functions specifically implemented by the present storage medium embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0158] It is understood that all or some steps, systems in the disclosed method above can be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components can be implemented as software by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application specific integrated circuit. Such software can be distributed on a computer-readable medium, and the computer-readable medium can include computer storage media (or non-transitory media) and communication media (or temporary media). As known to those of ordinary skill in the art, the term computer storage medium is included in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data) and is volatile and non-volatile, removable and non-removable media. Computer storage media includes but is not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, disk storage or other magnetic storage device, or can be used to store desired information and any other medium that can be accessed by a computer. Furthermore, as is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0159] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A bionics-based structural optimization method, characterized in that: The method comprises: Establish a model according to preset geometric parameters to obtain an initial three-dimensional model; Performing bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain a plurality of initial bionic structural models; Performing finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; performing simulation analysis on the first bionic structure model and the plurality of initial bionic structure models according to a preset algorithm, a preset working condition, to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set; Parameters of the first bionic structure model are optimized according to the sample data set and the preset model, a target optimization parameter combination is determined, and an optimization result is determined according to the target optimization parameter combination.

2. The method according to claim 1, characterized in that The performing simulation analysis according to a preset algorithm, a preset working condition, the first bionic structure model, and the plurality of initial bionic structure models to determine a sample data set specifically includes: Performing simulation analysis on the initial bionic structure models according to the preset working conditions to determine a bionic sample data set; Calculating the first bionic structure model according to the preset algorithm to generate several groups of shape variable parameters; Performing simulation analysis on the first bionic structure model according to the preset working conditions and the plurality of groups of shape variable parameters to determine an optimized sample data set; The sample data set is determined according to the bionic sample data and the optimized sample data set.

3. The method according to claim 2, characterized in that The performing simulation analysis on the plurality of initial bionic structure models according to the preset working conditions to determine the bionic sample data set specifically includes: performing stress analysis on the plurality of initial bionic structure models according to the preset working conditions to determine a first stress data set; performing mass calculations on a plurality of the initial bionic structure models respectively to determine a second mass data set; The bionic sample dataset is determined according to the first stress dataset and the second mass dataset.

4. The method according to claim 2, characterized in that The performing simulation analysis on the first bionic structure model according to the preset working condition and the plurality of groups of shape variable parameters to determine the optimized sample data set specifically includes: Adjusting the first bionic structure model according to the plurality of groups of shape variable parameters to determine a plurality of second bionic structure models; performing stress analysis on the plurality of second bionic structure models according to the preset working conditions to determine a second stress data set, and performing mass calculation on the plurality of second bionic structure models to determine a second mass data set; The optimized sample dataset is determined according to the second stress dataset and the second quality dataset.

5. The method according to claim 1, characterized in that Optimizing the parameters of the first bionic structure model according to the sample data set and the preset model to determine a target optimization parameter combination specifically includes: Training the preset model according to the sample data set to determine an optimized model; According to the matching between the first bionic structure model and the sample data set, a plurality of sets of target shape variable parameters are determined; wherein the target shape variable parameters are design shape parameters corresponding to the first bionic structure model; Optimizing and calculating several groups of target shape variable parameters according to the optimization model to determine the target optimization parameter combination.

6. The method according to claim 5, characterized in that The training of the preset model according to the sample data set to determine the optimized model specifically includes: Preprocessing the sample data set to obtain a standard data set, and dividing the standard data set according to a preset ratio to determine a training set, a validation set, and a test set; The preset model is trained according to the training set and the validation set to obtain an intermediate model; a model accuracy value is obtained by performing calculations according to the test set and the intermediate model; and the model accuracy value is compared with a preset threshold; If the model accuracy value is greater than or equal to the preset threshold, the intermediate model is used as the optimization model; If the model accuracy value is less than the preset threshold, the parameters of the intermediate model are adjusted, and the intermediate model after parameter adjustment is used as the preset model, and the process of training the preset model according to the training set and the validation set is returned to obtain the intermediate model; until the model accuracy value is greater than or equal to the preset threshold.

7. The method according to claim 1, characterized in that The method further comprises: Analyzing the optimization results to determine working condition response data, and performing image recognition on the optimization results to obtain bionic structure image data; The bionic sample data set in the sample data set is supplemented according to the bionic structure image data and the working condition response data.

8. A bionics-based structural optimization system, characterized in that: include: The first module is used to build a model according to preset geometric parameters to obtain an initial three-dimensional model; The second module is used to perform bionic processing on the initial three-dimensional model according to preset structural requirements and topology optimization goals to obtain a plurality of initial bionic structural models; The third module is used to perform finite element analysis on the plurality of initial bionic structure models according to preset engineering requirements to determine a first bionic structure model; perform simulation analysis based on a preset algorithm, preset working conditions, the first bionic structure model, and the plurality of initial bionic structure models to determine a sample data set; wherein the sample data set includes a bionic sample data set and an optimized sample data set; The fourth module is used to optimize the parameters of the first bionic structure model according to the sample data set and the preset model, determine the target optimization parameter combination, and determine the optimization result according to the target optimization parameter combination.

9. An electronic device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is configured to perform the method according to any one of claims 1 to 7 when executed by the processor.

Images