首页 >

AI驱动高速飞行器多学科发展知识图谱分析

安凯1,黄伟2,王振国3,徐小平1,孟玉珊1   

  1. 1. 中国人民解放军国防科技大学
    2. 湖南长沙国防科技大学空天科学学院临空所
    3. 国防科技大学
  • 收稿日期:2024-04-22 修回日期:2024-06-21 出版日期:2024-06-21 发布日期:2024-06-21
  • 通讯作者: 黄伟
  • 基金资助:
    湖南省杰出青年自然科学基金;国家自然科学基金项目

Knowledge atlas analysis of AI-Driven multidisciplinary development of hypersonic aircrafts

  • Received:2024-04-22 Revised:2024-06-21 Online:2024-06-21 Published:2024-06-21

摘要: 近年来,AI(Artificial Intelligence)技术的不断涌现为推动高速飞行器多学科发展提供了新的求解思路和实现方法。为系统梳理高速飞行器多学科发展的研究脉络、热点和趋势,探析AI方法对高速飞行器学科发展的长远影响,本文首先利用CiteSpace6.3.R1以及VOSviewer1.6.20软件对2000年至2024年中国知网(CNKI)和 Web of Science(WoS)数据库中的相关文献进行了调查,并对发文量、研究机构以及关键词聚类图谱等知识矩阵进行了多角度分析。基于此,介绍了流场特性智能预测、无模型自适应制导与控制、偏微分方程智能求解以及知识数据融合的不确定性多学科设计优化四个热点领域的研究概况。最后,本文总结了AI驱动下高速飞行器学科发展新的研究趋势,强调了该领域现有的挑战,并得出以下结论:1)AI融合知识已成为高速飞行器多学科研究新的科技范式;2)深度学习方法进一步拓展了高速飞行器各学科技术理论边界和应用范围,但在精确模型建立求解以及试验应用上仍具有很大探索空间。

关键词: AI技术, 高速飞行器, 多学科, 知识图谱, CiteSpace, VOSviewer

Abstract: In recent years, the development of AI (Artificial Intelligence) technology has provided new solutions and implementation for promoting the multidisciplinary development of hypersonic aircrafts. To sort out the research progress, hotspots, and trends of multidisciplinary development of high-speed aircraft systematically, and explore the long-term impact of AI methods, this article firstly uses software CiteSpace 6.3. R1 and VOSviewer 1.6.20 to investigate relevant literatures in the databases of China National Knowledge Infrastructure (CNKI) and Web of Science (WoS) from 2000 to 2024, and conducts a multi-dimensional analysis of knowledge matrices such as publication numbers, research institutions, and keyword clustering graphs. Based on this, this article introduces the research overview of four hot views: intelligent pre-diction of flow field characteristics, model-free adaptive guidance and control, intelligent applications of partial differential equations, and uncertain multidisciplinary design optimization through knowledge data fusion. Finally, this article summa-rizes new research trends in the development of hypersonic aircraft disciplines driven by AI, emphasizing the existing challenges in this field, and draws the following conclusions: 1) AI combining with knowledge has become a new techno-logical paradigm for multidisciplinary research on hypersonic aircraft; 2) The introduction of deep learning methods has further expanded the theoretical boundaries and application scope of various disciplines, but there is still an extensive exploration space in the establishment and solution of accurate models and experimental applications.

Key words: AI, hypersonic aircraft, multidiscipline, knowledge atlas, CiteSpace, VOSviewer

中图分类号: