智能赋能流体力学展望

Artificial intelligence (AI) is a cutting-edge technology in the 21st century. How fluid mechanics rejuvenates in the intelligent era is a topic worthy of consideration for researchers in this field. From the perspective of intelligent empowerment of fluid mechanics, its research connotations, research contents, recent research and difficulties are summarized, and the future development of intelligent fluid mechanics is prospected. The research points out that the data generated in fluid mechanics calculations or experiments is inherently big data. How to utilize these data through machine learning methods such as deep neural networks, random forests, and reinforcement learning, relieve or even replace the dependence on the human brain at the theoretical and methodological levels, and mine new knowledge has become a new research paradigm; related research will cover machine learning of flow control equations, machine learning of turbulence models, intelligent physical dimension analysis and scaling, and intelligent numerical simulation methods; with the aid of artificial intelligence technology, developing the intelligence of flow information feature extraction and multi-source data fusion is an urgent need for the development of fluid mechanics; the research contents should cover at least massive data mining methods and intelligent fusion of multi-source aerodynamic data; developing data-driven multi-disciplinary and multi-physical field coupling modeling and control of fluid mechanics is an urgent need for engineering applications, and related work involves multi-field coupling modeling, intelligent optimization design of aerodynamic shapes, and intelligent adaptive control of flows, etc.

人工智能(AI)是21世纪的前沿科技,流体力学如何在智能化时代焕发青春是值得本领域研究者思考的话题。从智能赋能流体力学角度,就其研究内涵、研究内容、近期研究及难点进行了总结,并对智能流体力学未来的发展进行了展望。研究指出,流体力学计算或试验中所产生的数据是天生的大数据,如何通过深度神经网络、随机森林、强化学习等机器学习方法来利用这些数据,缓解甚至替代理论和方法层面对人脑的依赖,挖掘新的知识,成为一种新的研究范式;相关研究将涵盖流动控制方程的机器学习、湍流模型的机器学习、物理量纲分析与标度的智能化以及数值模拟方法的智能化;借助人工智能技术,发展流动信息特征提取与多源数据融合的智能化是流体力学发展的迫切需求;研究内容应至少涵盖海量数据挖掘方法以及多源气动数据的智能融合;发展数据驱动的流体力学多学科、多物理场耦合建模与控制是工程应用的迫切需求,相关工作涉及多场耦合建模、气动外形智能优化设计以及流动智能自适应控制等方面。