Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Structure Topology Configuration Prediction Method Based on Convolutional Neural Network

A convolutional neural network and network model technology, applied in the field of structural optimization, can solve the problem of low prediction accuracy, and achieve the effects of improving prediction accuracy, increasing depth, and improving attention and freedom.

Active Publication Date: 2022-08-05
HUAZHONG UNIV OF SCI & TECH
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the defects and improvement needs of the prior art, the present invention provides a structural topology configuration prediction method based on convolutional neural network, which aims to solve the technical problem of low prediction accuracy of the existing structural topology optimization method based on deep learning

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Structure Topology Configuration Prediction Method Based on Convolutional Neural Network
  • A Structure Topology Configuration Prediction Method Based on Convolutional Neural Network
  • A Structure Topology Configuration Prediction Method Based on Convolutional Neural Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] A method for building optimal topology configuration prediction model based on convolutional neural network, such as figure 1 shown, including:

[0060] Build a training data set, each of which contains the optimal topology of the structure and the corresponding multi-channel tensor; when any parameter variable of the structure changes, at least one channel in the multi-channel tensor will change accordingly. Parametric variables include structural preset volume fraction, load location and load direction;

[0061] A network model including multiple encoding modules, one or more SE-ResNet modules, and multiple decoding modules connected in sequence is established to predict the optimal topological configuration of the corresponding structure according to the input multi-channel tensor. The model structure is as follows figure 2 shown; multiple encoding modules are used to extract multiple feature maps of different scales of the input multi-channel tensor; SE-ResNet mod...

Embodiment 2

[0089] A method for building optimal topology configuration prediction model based on convolutional neural network, such as Image 6 As shown, this embodiment is similar to the above-mentioned Embodiment 1, the difference lies in that, when training the network model by using the training data set, this embodiment adopts a two-stage training method, and the specific training process includes:

[0090] Using the multi-channel tensor in the training sample as the input data of the model, and using the corresponding optimal topology configuration as the label information, the network model is trained, and the first-stage model is obtained after the training is completed;

[0091] The multi-channel tensors in each training sample are input into the first-stage model, and the first-stage model outputs the optimal topological configuration prediction result corresponding to each training sample, and the optimal topological configuration prediction result corresponding to the training...

Embodiment 3

[0106] A method for predicting the optimal topology configuration of a structure based on a convolutional neural network is provided, including:

[0107] The parameter variables of the target structure are preprocessed into multi-channel tensors; the parameter variables include parameter variables including preset volume fraction, load position and load direction. When any parameter variable changes, at least one channel in the multi-channel tensor will occur. corresponding changes;

[0108] The multi-channel tensor is input into the structural optimal topology configuration prediction model established by the method for establishing the optimal structural topology configuration prediction model based on the convolutional neural network provided by the above-mentioned embodiment 1 or 2, so as to predict the optimal topology configuration of the structure. optimal topology;

[0109] For the specific implementation manner of preprocessing the parameter variables of the target s...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method for predicting a structure topology configuration based on a convolutional neural network, belonging to the field of structure optimization. tensor; when any one of the structural preset volume fraction, load position and load direction changes, at least one channel will change accordingly; establish a network model, in which multiple encoding modules are used to extract the multi-channel tensors. feature maps of different scales; the SE-ResNet module is used to obtain the attention weights of each channel in the feature map and fuse them into the feature map; the decoding module takes the feature map output by the previous module and the corresponding encoding module as input, and is used to convert the feature map Expand to the target size; use the training data set to train the network model, after the training, obtain the optimal topological configuration prediction model of the structure. The invention can improve the prediction accuracy of the optimal topological configuration of the structure.

Description

technical field [0001] The invention belongs to the field of structure optimization, and more particularly, relates to a structure topology configuration prediction method based on a convolutional neural network. Background technique [0002] Structural design is the key link in transforming abstract ideas into concrete designs. Designing a structure that meets the requirements with as little material (production cost) as possible is always the goal pursued by designers. The more traditional method is to first give a variety of initial designs based on experience and intuition, then calculate and compare each scheme, and decide the final scheme according to the analysis results. This method has obvious limitations: on the one hand, the initial The quality of the alternatives depends heavily on the level of designers, and more on the designers' past experience and engineering intuition; on the other hand, the whole process requires a lot of time and labor costs, and the opti...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F30/20G06N3/04G06F119/14
CPCG06F30/20G06F2119/14G06N3/045
Inventor 肖蜜张红扬崔芙铭汪逸晖高亮
Owner HUAZHONG UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com