摘要
近年来,以深度学习为代表的人工智能技术在民用领域飞速发展。在航空应用方面,人工智能技术将发挥重要的作用。人工智能技术在航空领域的应用,必须要考虑机载环境的约束和限制,尤其对智能计算处理器更是有着严格的要求。综合考虑人工智能技术的机载应用场景、计算特性和算力需求,通过分析研究通用处理器以及智能专用处理器的架构和适用性,基于异构融合的设计思想,提出了一条可行的技术实施路线,为机载智能应用提供计算支撑服务。
In recent years, artificial intelligence technology represented by deep learning has developed rapidly in the civil field. While for military application, such as aviation, AI technology will also play an important role. The AI application in the field of aviation must consider the constraints and limitations of the airborne environment, especially for the intelligent computing processors. Considering the airborne application scenario, computing characteristics and computing power requirements, the architecture and applicability of general purpose processors and AI processors were analyzed and studied. Then a feasible technical implementation route was proposed based on the idea of heterogeneous integration. It could provide basic computing services for airborne intelligent applications.
作者
文鹏程
白林亭
高泽
程陶然
Wen Pengcheng;Bai Linting;Gao Ze;Cheng Taoran(Aviation Key Laboratory of Science and Technology on Airborne and Missileborne Computer,AVIC Xi’an Aeronautics Computing Technique Research Institute,Xi’an 710065,China)
出处
《航空科学技术》
2020年第10期81-86,共6页
Aeronautical Science & Technology
关键词
航空人工智能
机载智能应用
OODA
智能计算
智能专用处理器
通用处理器
aeronautical artificial intelligence
airborne intelligent application
OODA
intelligent computing
AI processor
general purpose processor