This website requires JavaScript.

Biologically Inspired Design Concept Generation Using Generative Pre-Trained Transformers

Qihao ZhuXinyu ZhangJianxi Luo
Dec 2022
摘要
Biological systems in nature have evolved for millions of years to adapt andsurvive the environment. Many features they developed can be inspirational andbeneficial for solving technical problems in modern industries. This leads to aspecific form of design-by-analogy called bio-inspired design (BID). AlthoughBID as a design method has been proven beneficial, the gap between biology andengineering continuously hinders designers from effectively applying themethod. Therefore, we explore the recent advance of artificial intelligence(AI) for a data-driven approach to bridge the gap. This paper proposes agenerative design approach based on the generative pre-trained language model(PLM) to automatically retrieve and map biological analogy and generate BID inthe form of natural language. The latest generative pre-trained transformer,namely GPT-3, is used as the base PLM. Three types of design concept generatorsare identified and fine-tuned from the PLM according to the looseness of theproblem space representation. Machine evaluators are also fine-tuned to assessthe mapping relevancy between the domains within the generated BID concepts.The approach is evaluated and then employed in a real-world project ofdesigning light-weighted flying cars during its conceptual design phase Theresults show our approach can generate BID concepts with good performance.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答