Λεπτομέρειες βιβλιογραφικής εγγραφής
| Τίτλος: |
An agentic knowledge-infused reasoning framework for engineering design based on large language models. |
| Συγγραφείς: |
Li, Ruihao1 (AUTHOR), Chen, Chong1 (AUTHOR) chenc2021@gdut.edu.cn, Li, Xinhao1 (AUTHOR), Zhang, Hongbin1 (AUTHOR), Wang, Tao1 (AUTHOR), Cheng, Lianglun1 (AUTHOR) |
| Πηγή: |
Journal of Engineering Design. May2026, p1-29. 29p. 8 Illustrations. |
| Θεματικοί όροι: |
*Engineering design, Language models, Multiagent systems, Knowledge transfer, Knowledge graphs |
| Περίληψη: |
The rapid advancement of generative artificial intelligence, particularly large language models (LLMs), is transforming intelligent design from traditional paradigms such as Knowledge-Based Engineering and intelligent engineering design into a new form deeply integrated with artificial intelligence. However, design knowledge remains fragmented and distributed across multiple sources, while design tasks often require complex and high-level reasoning. The limited domain expertise of current LLMs further restricts their ability to produce accurate and reliable design output. To address these challenges, this study proposes an Agentic Knowledge-Infused Reasoning Framework (AKIRF) that integrates knowledge graph (KG) construction with multi-agent reasoning for engineering design. AKIRF first leverages LLMs to automatically construct a domain-specific design KG from diverse textual corpora. It employs knowledge distillation to train two specialised small language models (SLMs): one optimised for precise design knowledge extraction, and the other tailored for design-oriented Chain-of-Thought (CoT) reasoning, reducing computational costs. Building on this foundation, AKIRF introduces a cooperative multi-agent architecture to achieve accurate reasoning in engineering design. We validate the framework through a case study in ship equipment design. Experimental results demonstrate that AKIRF can enhance reasoning accuracy and provide an efficient pathway toward intelligent design in complex engineering domains. [ABSTRACT FROM AUTHOR] |
|
Copyright of Journal of Engineering Design is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Βάση Δεδομένων: |
Business Source Index |