Deep learning for determining a near-optimal topological design without any iteration

Abstract

In this study, we propose a novel deep learning-based method to predict an optimized structure for a given boundary condition and optimization setting without using any iterative scheme. For this purpose, first, using open-source topology optimization code, datasets of the optimized structures paired with the corresponding information on boundary conditions and optimization settings are generated at low (32 x 32) and high (128 x 128) resolutions. To construct the artificial neural network for the proposed method, a convolutional neural network (CNN)-based encoder and decoder network is trained using the training dataset generated at low resolution. Then, as a two-stage refinement, the conditional generative adversarial network (cGAN) is trained with the optimized structures paired at both low and high resolutions, and is connected to the trained CNN-based encoder and decoder network. The performance evaluation results of the integrated network demonstrate that the proposed method can determine a near-optimal structure in terms of pixel values and compliance with negligible computational time.

27 page, 11 figures, The paper is accepted in the Structural and Multidisciplinary Optimization journal, Springer

References in corpus (4)

• Conditional Generative Adversarial Nets
• CaloGAN: Simulating 3D High Energy Particle Showers in Multi-Layer Electromagnetic Calorimeters with Generative Adversarial Networks
• MaskGAN: Better Text Generation via Filling in the______
• Deep Learning the Physics of Transport Phenomena

Cited by in corpus (25)

• Deep Generative Design: Integration of Topology Optimization and Generative Models
• Topology optimization of 2D structures with nonlinearities using deep learning
• Data-driven topology optimization of spinodoid metamaterials with seamlessly tunable anisotropy
• On the use of Artificial Neural Networks in Topology Optimisation
• Deep Learning in Deterministic Computational Mechanics
• Self-Directed Online Machine Learning for Topology Optimization
• Generative Design by Reinforcement Learning: Enhancing the Diversity of Topology Optimization Designs
• Understanding the effect of hyperparameter optimization on machine learning models for structure design problems
• Algorithmically-Consistent Deep Learning Frameworks for Structural Topology Optimization
• Data-driven topology design using a deep generative model
• 3D Topology Optimization using Convolutional Neural Networks
• An AI-Assisted Design Method for Topology Optimization Without Pre-Optimized Training Data
• De-homogenization using Convolutional Neural Networks
• A novel topology design approach using an integrated deep learning network architecture
• Neural Networks for Generating Better Local Optima in Topology Optimization
• A New CGAN Technique for Constrained Topology Design Optimization
• Deep Generative Model-based Synthesis of Four-bar Linkage Mechanisms with Target Conditions
• A mechanistic-based data-driven approach to accelerate structural topology optimization through finite element convolutional neural network (FE-CNN)
• Deep learning-based topological optimization for representing a user-specified design area
• Image-based reconstruction for the impact problems by using DPNNs
• A generative adversarial network optimization method for damage detection and digital twinning by deep AI fault learning: Z24 Bridge structural health monitoring benchmark validation
• A surrogate model for topology optimisation of elastic structures via parametric autoencoders
• A Parametric Level Set Method for Topology Optimization based on Deep Neural Network (DNN)
• Generating various airfoil shapes with required lift coefficient using conditional variational autoencoders
• Can CNN Construct Highly Accurate Models Efficiently for High-Dimensional Problems in Complex Product Designs?