Machine learning-based design of triply periodic minimal surface structures

基于机器学习的三周期最小表面结构设计

• 批准号: DE210101676
• 负责人: Dr Jianguang Fang
• 金额: $ 30.64万
• 依托单位: University of Technology Sydney
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Early Career Researcher Award
• 财政年份: 2021
• 资助国家: 澳大利亚
• 起止时间: 2021-03-01 至 2024-02-29
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DE210101676
• 关键词: Machine; learning; based; design; triply

This project aims to develop a new approach to design of new lightweight, crashworthy and manufacturable structures by taking advantage of the latest technologies in computational optimisation, artificial intelligence and additive manufacturing. The study intends to develop a new machine learning-based multiscale design framework to seek optimal triply periodic minimal surface structures, considering fabrication-induced defects and uncertainty. The expected outcome of this project is new methodologies for generating eco-friendly structures with robust mechanical properties in crashing applications. This should provide significant benefits to transport industries by enhancing structural safety and energy saving for next generation vehicles.

该项目旨在通过利用计算优化、人工智能和增材制造方面的最新技术，开发一种设计新型轻量化、耐碰撞和可制造结构的新方法。该研究旨在开发一种新的基于机器学习的多尺度设计框架，以寻求最佳的三周期最小表面结构，同时考虑到制造引起的缺陷和不确定性。该项目的预期结果是在碰撞应用中产生具有强大机械性能的环保结构的新方法。这将通过提高下一代车辆的结构安全性和节能性，为运输行业带来显著的好处。