A Deep Learning-Based Method and System for Topology Optimization of Industrial Equipment

Abstract

This invention provides a deep learning-based method and system for topology optimization of industrial equipment. The method includes: constructing a model of the equipment to be optimized based on a finite element modeling algorithm, wherein the model equipment includes at least the model's geometric parameters and load parameters in the design domain; setting displacement boundary conditions for model optimization; constructing a convolutional neural network model, which is used to process the input multi-channel tensor to obtain the density distribution optimization result of the equipment model; determining the multi-channel tensor input to the convolutional neural network based on the variable density method, according to the model's geometric parameters, load parameters, and displacement boundary conditions; inputting the multi-channel tensor into the trained convolutional neural network model to obtain the target density distribution of the equipment model, and combining it with the density parameters in the material set to obtain the topology optimization result of the equipment model. The method provided by this invention can improve the efficiency of topology optimization of industrial equipment structures.

Description

Technical Field

[0001] This invention belongs to the field of product design technology, specifically relating to a method and system for optimizing the topology of industrial equipment based on deep learning. Background Technology

[0002] Equipment product design is a crucial task in modern engineering, as its performance and effectiveness directly impact the quality, efficiency, and safety of industrial production. However, traditional equipment product design methods are often limited by experience and trial and error, resulting in problems such as long design times, resource waste, and high costs. To address these issues and promote innovation in equipment design, topology optimization design, as an advanced design method, has gained widespread attention and application.

[0003] Topology optimization is an optimization design method for equipment manufacturing products, combining structural mechanics and optimization techniques. By optimizing the spatial distribution of materials, structural configuration, and component dimensions, the optimal structural form can be obtained from multiple structural design schemes, achieving weight reduction, cost reduction, and performance improvement in products such as aerospace and automobiles.

[0004] The crucial role of topology optimization software in product design lies primarily in improving product performance, enhancing design efficiency, and accelerating product releases. Therefore, developing topology optimization software for specific scenarios not only helps engineers quickly master its use but also better promotes the development of related industries.

[0005] Since topology optimization software is generally based on the finite element method, the computational efficiency of topology optimization often fails to meet the needs of practical engineering.

[0006] Improving the efficiency of topology optimization in the design of industrial equipment is a technical problem that urgently needs to be solved. Summary of the Invention

[0007] In view of the above-mentioned problems in the prior art, the purpose of this invention is to provide a method and system for topology optimization of industrial equipment based on deep learning, which can improve the efficiency of topology optimization in product design of industrial equipment.

[0008] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0009] On one hand, the present invention provides a deep learning-based method for optimizing the topology of industrial equipment, the method comprising:

[0010] Based on the finite element modeling algorithm, a device model to be optimized is constructed. The device model includes at least the model's geometric parameters and load parameters in the design domain.

[0011] Set the displacement boundary conditions for model optimization;

[0012] A convolutional neural network (CNN) model is constructed to process the input multi-channel tensor to obtain the density distribution optimization result of the device model. The CNN model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected sequentially. The current convolutional module is used to extract features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected sequentially and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and restores the spatial information of the multi-channel tensor based on the concatenated data. The output layer is used to obtain the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules.

[0013] Based on the variable density method, the multi-channel tensor input to the convolutional neural network is determined according to the model geometric parameters and load parameters of the device model, as well as the displacement boundary conditions. The multi-channel tensor includes at least the horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction.

[0014] The multi-channel tensor is input into the trained convolutional neural network model to obtain the target density distribution of the device model;

[0015] Based on the target density distribution of the device model and the density parameters in the material set, the topology optimization result of the device model is obtained.

[0016] The convolutional module includes convolutional layers, activation functions, and max pooling layers; the input layer is connected to multiple convolutional layers, and each convolutional layer is sequentially connected to an activation function and a max pooling layer. The convolutional layers are used to extract feature maps of the input multi-channel tensors, and the activation functions are used to perform non-linear operations on the multi-channel tensors after convolution.

[0017] The transposed convolution module includes a transposed convolutional layer, a Dropout layer, and a concatenation layer. The transposed convolutional layer is used to upsample and increase the size of the feature map. The Dropout layer is used to randomly discard neurons through the Dropout mechanism. The concatenation layer is used to concatenate the features extracted from different layers and send the concatenation result to the output layer.

[0018] Furthermore, the variable density method is expressed by the following formula:

[0019]

[0020] F = K·U;

[0021]

[0022] 0 <x min ≤x≤1;

[0023] Where U is the global displacement vector, F is the force vector, K is the global stiffness matrix, and u e k and k0 are the element displacement vector and stiffness matrix, respectively, and x is the design variable vector. min It is the minimum relative density vector, N is the number of elements used for the discrete design domain, P is the penalty factor, and V is the minimum relative density vector. (x) V0 and V0 are the material volume and design domain volume, respectively, and f is the set volume fraction.

[0024] Furthermore, the multi-channel tensor is obtained through the following steps:

[0025] Based on the variable density method and the initial design variable vector of the material in the device model, the displacement vector of the model in the horizontal direction and the displacement vector in the vertical direction are determined.

[0026] The horizontal displacement vector and the vertical displacement vector are respectively converted into horizontal displacement matrices and vertical displacement matrices of the same magnitude.

[0027] Extract the initial volume fraction of the design domain and expand it into an initial volume fraction matrix with the same size as the displacement matrix in the horizontal direction and the displacement matrix in the vertical direction;

[0028] The horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction matrix are combined into the multichannel tensor.

[0029] Furthermore, the method also includes:

[0030] The target density distribution of the device model is normalized.

[0031] Based on pre-set identifiers for different density values, the normalized device model is visualized, and the identifiers include colors.

[0032] Furthermore, the convolutional neural network model is trained through the following steps:

[0033] Construct the initial convolutional neural network model;

[0034] Acquire training data, which includes multiple optimized equipment models, density distribution results and optimization parameters corresponding to each equipment module, and the optimization parameters include at least model geometric parameters, load parameters and displacement boundary conditions;

[0035] The optimized parameters are processed to obtain a multi-channel tensor;

[0036] Using the multi-channel tensor as input data and the density distribution result of the device model as the target output, the initial convolutional neural network model is trained using a preset loss function until the convergence condition is met, thus obtaining the trained convolutional neural network model.

[0037] Furthermore, the convolutional neural network model is the U-net network model.

[0038] On the other hand, the present invention also provides a deep learning-based industrial equipment topology optimization system, the system comprising:

[0039] The equipment model construction module is used to construct an equipment model to be optimized based on the finite element modeling algorithm. The equipment model includes at least the model's geometric parameters and load parameters in the design domain.

[0040] The condition setting module is used to set the displacement boundary conditions for model optimization;

[0041] A convolutional neural network (CNN) model building module is used to construct a CNN model. This CNN model processes the input multi-channel tensor to obtain the density distribution optimization result of the device model. The CNN model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected sequentially. The current convolutional module extracts features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected sequentially and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and recovers the spatial information of the multi-channel tensor based on the concatenated data. The output layer obtains the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules.

[0042] The input module is used to determine the multi-channel tensor of the input convolutional neural network based on the variable density method, according to the model geometric parameters and load parameters of the device model, and the displacement boundary conditions. The multi-channel tensor includes at least a horizontal displacement matrix, a vertical displacement matrix, and an initial volume fraction.

[0043] The target density distribution calculation module is used to input the multi-channel tensor into the trained convolutional neural network model to obtain the target density distribution of the device model.

[0044] The determination module is used to obtain the topology optimization result of the device model based on the target density distribution of the device model and the density parameters in the material set.

[0045] On the other hand, the present invention also provides an industrial equipment topology optimization system, the system comprising:

[0046] The modeling module is used to construct a geometric model of the industrial equipment to be optimized based on the finite element modeling algorithm, including at least the following controls for user input: geometric parameters, load parameters, and displacement boundary conditions;

[0047] The optimization module is used to perform topology optimization processing according to the deep learning-based industrial equipment topology optimization method described above, so as to obtain the target density distribution of the industrial equipment to be optimized.

[0048] The visualization module is used to visualize the industrial equipment based on the target density distribution and a preset identifier matching rule.

[0049] On the other hand, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the deep learning-based industrial equipment topology optimization method as described above.

[0050] Using the above technical solution, the present invention provides a deep learning-based method for optimizing the topology of industrial equipment. The method includes: constructing a device model to be optimized based on a finite element modeling algorithm; the device model includes at least geometric parameters and load parameters of the model in the design domain; setting displacement boundary conditions for model optimization; and constructing a convolutional neural network model. This convolutional neural network model processes the input multi-channel tensor to obtain the density distribution optimization result of the device model. The convolutional neural network model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected sequentially. The current convolutional module is used to extract features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected sequentially and correspond one-to-one with the multiple convolutional modules. The block concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and recovers the spatial information of the multi-channel tensor based on the concatenated data. The output layer is used to obtain the density distribution optimization result of the equipment model based on the processing result of the transposed convolutional module. Based on the variable density method, the multi-channel tensor input to the convolutional neural network is determined according to the model geometric parameters and load parameters of the equipment model, as well as the displacement boundary conditions. The multi-channel tensor includes at least the horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction. The multi-channel tensor is input into the trained convolutional neural network model to obtain the target density distribution of the equipment model. Based on the target density distribution of the equipment model and the density parameters in the material set, the topology optimization result of the equipment model is obtained. The method provided by this invention can improve the efficiency of topology optimization of industrial equipment structures.

[0051] To make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are 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.

[0053] Figure 1 The diagram illustrates the steps of a deep learning-based industrial equipment topology optimization method provided by an embodiment of the present invention.

[0054] Figure 2A schematic diagram of the network structure of the convolutional neural network model in the embodiments of this specification is shown;

[0055] Figure 3 This specification shows a schematic diagram illustrating the visualization effects in the embodiments of this specification;

[0056] Figure 4 A schematic diagram of the structure of an industrial equipment topology optimization system based on deep learning, provided in an embodiment of this specification, is shown.

[0057] Figure 5 A schematic diagram of the structure of an industrial equipment topology optimization system based on deep learning, provided in an embodiment of this specification, is shown.

[0058] Figure 6 The workflow of the built-in software of the industrial equipment topology optimization system in the embodiments of this specification is shown;

[0059] Figure 7 This document shows a schematic diagram of a finite element modeling operation page in one embodiment of the present specification;

[0060] Figure 8 This document shows a schematic diagram of a topology optimization parameter input operation page in one embodiment of the present specification;

[0061] Figure 9 A schematic diagram of the system provided in the embodiments of this specification is shown.

[0062] Explanation of symbols in the attached drawings:

[0063] 410. Equipment model construction module; 420. Condition setting module; 430. Convolutional neural network model construction module; 440. Input module; 450. Target density distribution calculation module; 460. Determination module; 510. Modeling module; 520. Optimization module; 530. Visualization module. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0066] In existing technologies, topology optimization is an optimization design method for equipment manufacturing products, combining structural mechanics and optimization methods. By optimizing the spatial distribution of materials, structural configuration, and component dimensions, the optimal structural form can be obtained from various structural design schemes, achieving weight reduction, cost reduction, and performance improvement for products such as aerospace and automobiles.

[0067] The crucial role of topology optimization software in product design lies primarily in improving product performance, enhancing design efficiency, and accelerating product releases. Therefore, developing topology optimization software for specific scenarios not only helps engineers quickly master its use but also better promotes the development of related industries.

[0068] Since topology optimization software is generally based solely on finite element methods, its computational efficiency often falls short of the needs of practical engineering projects.

[0069] To address the aforementioned issues, this invention provides a deep learning-based method for optimizing the topology of industrial equipment, which can improve the efficiency of topology optimization in product design. Figure 1 This is a schematic diagram illustrating the steps of a deep learning-based industrial equipment topology optimization method provided in an embodiment of the present invention. This specification provides the operational steps of the method described in the embodiments or flowcharts, but based on conventional or non-creative labor, more or fewer operational steps may be included. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual system or device products, the methods shown in the embodiments or drawings can be executed sequentially or in parallel. Specifically, as shown in the attached drawings... Figure 1 As shown, the method may include:

[0070] S101: Based on the finite element modeling algorithm, construct the equipment model to be optimized. The equipment model includes at least the model geometric parameters and load parameters in the design domain.

[0071] S102: Set the displacement boundary conditions for model optimization;

[0072] S103: Construct a convolutional neural network model. The convolutional neural network model is used to process the input multi-channel tensor to obtain the density distribution optimization result of the device model. The convolutional neural network model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected in sequence. The current convolutional module is used to extract features based on the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected in sequence and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and restores the spatial information of the multi-channel tensor based on the concatenated data. The output layer is used to obtain the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules.

[0073] S104: Based on the variable density method, according to the model geometric parameters and load parameters of the device model, as well as the displacement boundary conditions, determine the multi-channel tensor input to the convolutional neural network. The multi-channel tensor includes at least the horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction.

[0074] S105: Input the multi-channel tensor into the trained convolutional neural network model to obtain the target density distribution of the device model;

[0075] S106: Based on the target density distribution of the device model and the density parameters in the material set, the topology optimization result of the device model is obtained.

[0076] This invention can be understood as follows: By utilizing deep learning algorithms, such as convolutional neural network models, to predict the topology optimization of industrial equipment during the product design process, the optimal topology structure can be obtained. Compared with the traditional SIMP method based on isotropic solid materials, the prediction results based on deep learning models have more accurate density values ​​and more intuitive topology boundary display.

[0077] Furthermore, the convolution module includes convolutional layers, activation functions, and max pooling layers; the input layer is connected to multiple convolutional layers, and each convolutional layer is sequentially connected to an activation function and a max pooling layer. The convolutional layers are used to extract feature maps of the input multi-channel tensors, and the activation functions are used to perform nonlinear operations on the multi-channel tensors after convolution.

[0078] The transposed convolution module includes a transposed convolutional layer, a Dropout layer, and a concatenation layer. The transposed convolutional layer is used to upsample and increase the size of the feature map. The Dropout layer is used to randomly discard neurons through the Dropout mechanism. The concatenation layer is used to concatenate the features extracted from different layers and send the concatenation result to the output layer.

[0079] For example, deep learning models are trained based on convolutional neural networks, such as Figure 2 As shown, a convolutional neural network includes an input layer (imageinput), a convolutional layer (conv), an activation function (relu), a pooling layer (maxpooling), a transposed convolutional layer (transconv), dropout, a concatenation layer (concatenate), and an output layer (regressionout).

[0080] The input to the neural network is a 33×65×3 tensor, which includes the horizontal displacement matrix, the vertical displacement matrix, and the initial volume of the image.

[0081] Multiple convolutional layers are concatenated after the input layer, with each convolutional layer followed by an activation function (ReLU) and a max-pooling layer. The ReLU activation function performs non-linear operations on the input tensor after convolution, preventing gradient loss due to increased model depth. The max-pooling layer further extracts important features and further reduces the tensor size, thus reducing computational cost.

[0082] In the latter part of the network, transposed convolutional layers are used for upsampling to increase the size of the feature maps, restoring the spatial information of the tensor data after being reduced in size in the previous step. In the latter part, a Dropout mechanism is used to prevent overfitting by randomly dropping some neurons to increase the model's generalization ability.

[0083] The final output layer concatenates the features extracted from different layers, inputting more feature information into the Sigmoid and regression functions to output the final prediction result. The model's output provides the optimized structural layout under given conditions in the form of relative density, i.e., a 32×64 matrix with element values ​​ranging from 0 to 1, representing the relative density value.

[0084] In other words, the convolutional neural network model outputs the density distribution of the designed product, and then combines it with the already determined material density to match the corresponding materials, ultimately obtaining an optimized solution for the designed product.

[0085] Furthermore, the application of the variable density method in topology optimization allows for the flexible allocation of materials within a given design space, thereby optimizing structural performance. By adjusting the density function, the distribution of materials within the design domain can be controlled, thus influencing the mechanical properties of the structure.

[0086] The variable density method describes material distribution by introducing a density variable. It expresses the relationship between the relative density of an element and the elastic modulus of the material as a density function of a continuous variable. The density referred to in the variable density method is a pseudo-density, reflecting the relationship between material density and material properties, and its value in each design element is between 0 and 1. The flexible optimization topology optimization problem can be written as:

[0087]

[0088] F = K·U;

[0089]

[0090] 0 < x min ≤x≤1;

[0091] Where U is the global displacement vector, F is the force vector, K is the global stiffness matrix, and u e k and k0 are the element displacement vector and stiffness matrix, respectively, and x is the design variable vector. min It is the minimum relative density vector, N is the number of elements used for the discrete design domain, P is the penalty factor, and V is the minimum relative density vector. (x) V0 and V0 are the material volume and design domain volume, respectively, and f is the set volume fraction.

[0092] Furthermore, the multi-channel tensor is obtained through the following steps:

[0093] Based on the variable density method and the initial design variable vector of the material in the device model, the displacement vector of the model in the horizontal direction and the displacement vector in the vertical direction are determined.

[0094] The horizontal displacement vector and the vertical displacement vector are respectively converted into horizontal displacement matrices and vertical displacement matrices of the same magnitude.

[0095] Extract the initial volume fraction of the design domain and expand it into an initial volume fraction matrix with the same size as the displacement matrix in the horizontal direction and the displacement matrix in the vertical direction;

[0096] The horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction matrix are combined into the multichannel tensor.

[0097] Furthermore, the convolutional neural network model is trained through the following steps:

[0098] Construct the initial convolutional neural network model;

[0099] Acquire training data, which includes multiple optimized equipment models, density distribution results and optimization parameters corresponding to each equipment module, and the optimization parameters include at least model geometric parameters, load parameters and displacement boundary conditions;

[0100] The optimized parameters are processed to obtain a multi-channel tensor;

[0101] Using the multi-channel tensor as input data and the density distribution result of the device model as the target output, the initial convolutional neural network model is trained using a preset loss function until the convergence condition is met, thus obtaining the trained convolutional neural network model.

[0102] For example, 6000 samples with different boundary conditions were generated using topology optimization software to build a dataset for training a deep learning model. The input to the neural network is a 33×65×3 tensor, which includes the horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction. Therefore, for the input data, U1 and U2 are converted into matrices of size [number of row nodes, number of column nodes], while the initial volume fraction of the design domain is extracted and expanded into matrices of the same size as U1 and U2. Then, U1, U2, and the matrices are merged into three-channel image data of size [number of row nodes, number of column nodes, 3, number of samples].

[0103] The output of the deep learning model presents the optimized structural layout under given conditions in the form of relative density, i.e., a 32×64 matrix with element values ​​ranging from 0 to 1. Regarding the output data, to improve training efficiency and accuracy, the output vectors with values ​​between 0 and 1 are normalized and then converted into image data of [number of cell rows, number of cell columns, 1, number of samples]. To effectively train the neural network and accurately evaluate its performance, all samples are divided into training, validation, and test sets in an 8:1:1 ratio.

[0104] During model training, it is necessary to continuously adjust parameters such as learning rate, weight initialization, regularization, batch size, and number of iterations, as well as data augmentation techniques, to enhance the model's performance and generalization ability, improve convergence speed, avoid overfitting, and gradually minimize the loss function to improve model accuracy. Compared to traditional SIMP-based methods, the prediction results based on deep learning models offer more accurate density values ​​and more intuitive topological boundary displays.

[0105] In one embodiment of this specification, the neural network uses the unetLayers code included with Matlab to generate the U-Net network.

[0106] Furthermore, the method also includes:

[0107] The target density distribution of the device model is normalized.

[0108] Based on pre-set identifiers for different density values, the normalized device model is visualized, and the identifiers include colors.

[0109] This can be understood as follows: in order to improve training efficiency and accuracy, the output vector with values ​​between 0 and 1 is normalized, and different density values ​​are represented by different colors, making the boundary conditions more obvious.

[0110] like Figure 3 The image shown is a visual illustration of an example from an embodiment of this specification.

[0111] Based on the methods provided above, this specification also provides an industrial equipment topology optimization system based on deep learning, such as... Figure 4 As shown, the system includes:

[0112] The equipment model construction module 410 is used to construct an equipment model to be optimized based on the finite element modeling algorithm. The equipment model includes at least the model geometric parameters and load parameters in the design domain.

[0113] The condition setting module 420 is used to set the displacement boundary conditions for model optimization;

[0114] A convolutional neural network model building module 430 is used to build a convolutional neural network model. The convolutional neural network model is used to process the input multi-channel tensor to obtain the density distribution optimization result of the device model. The convolutional neural network model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected in sequence. The current convolutional module is used to extract features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected in sequence and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and restores the spatial information of the multi-channel tensor based on the concatenated data. The output layer is used to obtain the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules.

[0115] The input module 440 is used to determine the multi-channel tensor of the input convolutional neural network based on the variable density method, according to the model geometric parameters and load parameters of the device model, and the displacement boundary conditions. The multi-channel tensor includes at least a horizontal displacement matrix, a vertical displacement matrix, and an initial volume fraction.

[0116] The target density distribution calculation module 450 is used to input the multi-channel tensor into the trained convolutional neural network model to obtain the target density distribution of the device model.

[0117] The determination module 460 is used to obtain the topology optimization result of the device model based on the target density distribution of the device model and the density parameters in the material set.

[0118] The effects achieved by the above-described device are the same as those achieved by the above-described method, and will not be repeated in the embodiments of this specification.

[0119] This specification also provides an industrial equipment topology optimization system in its embodiments, such as... Figure 5 As shown, the system includes:

[0120] Modeling module 510 is used to construct a geometric model of the industrial equipment to be optimized based on the finite element modeling algorithm, including at least the following controls for user input: geometric parameters, load parameters, and displacement boundary conditions;

[0121] The optimization module 520 is used to perform topology optimization processing according to the deep learning-based industrial equipment topology optimization method described above, so as to obtain the target density distribution of the industrial equipment to be optimized.

[0122] The visualization module 530 is used to visualize the industrial equipment based on the target density distribution and a preset identifier matching rule.

[0123] For example, corresponding processing software can be configured within the topology optimization system to implement the functions of the above modules. The software is divided into three main modules: modeling, topology optimization algorithms, and software visualization, such as... Figure 6 As shown.

[0124] The software, through its finite element modeling capabilities, can quickly and accurately create geometric models of industrial equipment. For example... Figure 7 As shown, users can specify the geometric parameters of the equipment, the magnitude, direction, and location of the loads, and pass the equipment's boundary parameters to the topology optimization algorithm model via the "Apply" button. Displacement boundary conditions (MMB and cantilever beams) can be set via radio buttons and are visually displayed in the image according to the corresponding displacement boundary conditions. With these inputs, the software can generate an accurate finite element model, preparing it for the subsequent optimization process.

[0125] After inputting the corresponding geometric lengths x and y in the geometric region, the geometric parameters are passed to the topology optimization algorithm model. In the load region, the load position coordinates Position_x and Position_y, as well as the load components Force_x and Force_y in the x and y directions, are input, and then the load parameters are passed to the topology optimization algorithm model.

[0126] Based on the results of the finite element analysis, optimization iterations are performed. The parameters required for topology optimization iterations include volume fraction, number of iterations, penalty factor, and filter radius, as follows: Figure 8 As shown. The software will pass these parameters to the topology optimization algorithm model to calculate the topology optimization.

[0127] After each iteration, the optimization results are evaluated to check if they meet the design requirements. The optimization process ends when predetermined termination conditions are met, such as reaching the maximum number of iterations or result convergence. After the topology optimization calculation is complete, the results of the optimization iterations can be viewed, such as... Figure 3 As shown.

[0128] This embodiment provides a computer device, the internal structure of which can be shown in the following diagram. Figure 9 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection.

[0129] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0130] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0131] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0132] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0133] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0134] It should also be understood that, in the embodiments of the present invention, the term "and / or" is merely a description of the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the present invention, the character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0135] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0136] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0137] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, or may be electrical, mechanical, or other forms of connection.

[0138] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0139] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A deep learning-based method for optimizing the topology of industrial equipment, characterized in that, The method includes: Based on the finite element modeling algorithm, a device model to be optimized is constructed. The device model includes at least the model's geometric parameters and load parameters in the design domain. Set the displacement boundary conditions for model optimization; A convolutional neural network (CNN) model is constructed to process the input multi-channel tensor to obtain the density distribution optimization result of the device model. The CNN model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected sequentially. The current convolutional module is used to extract features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected sequentially and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and restores the spatial information of the multi-channel tensor based on the concatenated data. The output layer is used to obtain the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules. Based on the variable density method, the multi-channel tensor input to the convolutional neural network is determined according to the model geometric parameters and load parameters of the device model, as well as the displacement boundary conditions. The multi-channel tensor includes at least the horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction. The multi-channel tensor is input into the trained convolutional neural network model to obtain the target density distribution of the device model; Based on the target density distribution of the device model and the density parameters in the material set, the topology optimization result of the device model is obtained.

2. The method according to claim 1, characterized in that, The convolutional module includes convolutional layers, activation functions, and max pooling layers; the input layer is connected to multiple convolutional layers, and each convolutional layer is sequentially connected to an activation function and a max pooling layer. The convolutional layers are used to extract feature maps of the input multi-channel tensors, and the activation functions are used to perform non-linear operations on the multi-channel tensors after convolution. The transposed convolution module includes a transposed convolutional layer, a Dropout layer, and a concatenation layer. The transposed convolutional layer is used to upsample and increase the size of the feature map. The Dropout layer is used to randomly discard neurons through the Dropout mechanism. The concatenation layer is used to concatenate the features extracted from different layers and send the concatenation result to the output layer.

3. The method according to claim 1, characterized in that, The variable density method is expressed by the following formula: F = K·U; 0<x min ≤x≤1; Where U is the global displacement vector, F is the force vector, K is the global stiffness matrix, and u e k and k0 are the element displacement vector and stiffness matrix, respectively, and x is the design variable vector. min It is the minimum relative density vector, N is the number of elements used for the discrete design domain, P is the penalty factor, and V is the minimum relative density vector. (x) V0 and V0 are the material volume and design domain volume, respectively, and f is the set volume fraction.

4. The method according to claim 3, characterized in that, The multichannel tensor is obtained through the following steps: Based on the variable density method and the initial design variable vector of the material in the device model, the displacement vector of the model in the horizontal direction and the displacement vector in the vertical direction are determined. The horizontal displacement vector and the vertical displacement vector are respectively converted into horizontal displacement matrices and vertical displacement matrices of the same magnitude. Extract the initial volume fraction of the design domain and expand it into an initial volume fraction matrix with the same size as the displacement matrix in the horizontal direction and the displacement matrix in the vertical direction; The horizontal displacement matrix, the vertical displacement matrix, and the initial volume fraction matrix are combined into the multichannel tensor.

5. The method according to claim 1, characterized in that, The method further includes: The target density distribution of the device model is normalized. Based on pre-set identifiers for different density values, the normalized device model is visualized, and the identifiers include colors.

6. The method according to claim 1, characterized in that, The convolutional neural network model is trained through the following steps: constructing an initial convolutional neural network model; Acquire training data, which includes multiple optimized equipment models, density distribution results and optimization parameters corresponding to each equipment module, and the optimization parameters include at least model geometric parameters, load parameters and displacement boundary conditions; The optimized parameters are processed to obtain a multi-channel tensor; Using the multi-channel tensor as input data and the density distribution result of the device model as the target output, the initial convolutional neural network model is trained using a preset loss function until the convergence condition is met, thus obtaining the trained convolutional neural network model.

7. The method according to claim 1, characterized in that, The convolutional neural network model is the U-net network model.

8. A deep learning-based industrial equipment topology optimization system, characterized in that, The system includes: The equipment model construction module is used to construct an equipment model to be optimized based on the finite element modeling algorithm. The equipment model includes at least the model's geometric parameters and load parameters in the design domain. The condition setting module is used to set the displacement boundary conditions for model optimization; A convolutional neural network (CNN) model building module is used to construct a CNN model. This CNN model processes the input multi-channel tensor to obtain the density distribution optimization result of the device model. The CNN model includes an input layer, multiple convolutional modules, multiple transposed convolutional modules, and an output layer. The input layer is used to input the multi-channel tensor. The multiple convolutional modules are connected sequentially. The current convolutional module extracts features from the output of the previous convolutional module to obtain the feature data of the multi-channel tensor. The multiple transposed convolutional modules are connected sequentially and correspond one-to-one with the multiple convolutional modules. The current transposed convolutional module concatenates the output of the previous transposed convolutional module and the output of the corresponding convolutional module, and recovers the spatial information of the multi-channel tensor based on the concatenated data. The output layer obtains the density distribution optimization result of the device model based on the processing result of the transposed convolutional modules. The input module is used to determine the multi-channel tensor of the input convolutional neural network based on the variable density method, according to the model geometric parameters and load parameters of the device model, and the displacement boundary conditions. The multi-channel tensor includes at least a horizontal displacement matrix, a vertical displacement matrix, and an initial volume fraction. The target density distribution calculation module is used to input the multi-channel tensor into the trained convolutional neural network model to obtain the target density distribution of the device model. The determination module is used to obtain the topology optimization result of the device model based on the target density distribution of the device model and the density parameters in the material set.

9. An industrial equipment topology optimization system, characterized in that, The system includes: The modeling module is used to construct a geometric model of the industrial equipment to be optimized based on the finite element modeling algorithm, including at least the following controls for user input: geometric parameters, load parameters, and displacement boundary conditions; An optimization module is used to perform topology optimization processing on the deep learning-based industrial equipment topology optimization method according to any one of claims 1 to 7, so as to obtain the target density distribution of the industrial equipment to be optimized. The visualization module is used to visualize the industrial equipment based on the target density distribution and a preset identifier matching rule.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the deep learning-based industrial equipment topology optimization method as described in any one of claims 1 to 7.

Images