人机认知协同的复杂装备概念设计:挑战、进展和展望

Human-computer Cognitive Collaboration-driven Conceptual Design of Complex Equipment:Research Progress and Challenges

传统的计算机辅助设计以几何特征为主导、以定量表征为依据、以经验试错为手段,难以有效支持设计信息残缺且设计认知混沌的概念设计阶段。我国正步入“十四五”发展的新时期,认知科学和人工智能的高度融合为复杂装备概念设计带来了全新的机遇与挑战。人机认知协同的概念设计将设计对象和设计主体有机结合,从经验试错推理发展到设计规律外显认知,揭示设计对象递归迭代与设计主体认知演化的内在联系,实现概念设计抽象认知过程的计算机可理解和可计算。阐述侧重设计对象和侧重设计主体的复杂装备概念设计研究现状,分析客户需求语义认知辨识、思维认知规律神经成像、功构映射智能认知推理和概念方案协同认知决策等人机认知协同的概念设计关键技术。在分析现有计算机辅助概念设计局限性的基础上,对新一代基于人机认知协同的计算机辅助概念设计进行了展望。

Traditional computer-aided design relies on geometric features,quantitative characterization,and trial-and-error.It does not conform to the conceptual design stage with incomplete design information and chaotic design cognition.China is entering a new development period of the 14th Five-Year Plan.The fusion of cognitive science and artificial intelligence brings new opportunities and challenges to the conceptual design of complex equipment.Human-computer cognitive collaboration-driven conceptual design organically combines object-oriented and subject-oriented aspects.It develops from experiential trial-and-error reasoning to conceptual design with explicit cognition of design laws.The inherent connection between the recursive iteration of design objects and the cognitive evolution of design subjects is revealed to render abstract design procedures comprehensible and operational for computers.The state-of-the-art in object-oriented and subject-oriented conceptual design of complex equipment is illustrated firstly.The key technologies such as semantic cognitive identification of customer needs,neuroimaging of thinking cognitive laws,intelligent cognitive reasoning of function-structure mapping,and collaborative cognitive decision-making of concept schemes are analyzed.Through revealing the limitations of existing computer-aided conceptual design methods,a new generation of computer-aided conceptual design based on human-machine cognitive collaboration has prospected.