Physics-constrained adaptive learning for multi-physics optimization: Time-accurate prediction of chaotic and turbulent flows

用于多物理优化的物理约束自适应学习：混沌流和湍流的时间精确预测

• 批准号: 2606505
• 金额: --
• 依托单位: Imperial College London
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2606505
• 关键词: Physics; constrained; adaptive; learning; multi

The ability of fluid mechanics modelling to predict the evolution of a flow is enabled both by physical principles and empirical approaches. On the one hand, physical principles (for example conservation laws) are extrapolative - they provide predictions on phenomena that have not been observed. On the other hand, empirical modelling provides correlation functions within data. Artificial intelligence and machine learning are excellent at empirical modelling. In this project, the student will combine physical principles and empirical modelling into a unified approach: physics-constrained data-driven methods for multi-physics optimisation. The objectives are to constrain the governing equations of turbulence in machine learning. This will enable more accuracy, robustness, and generalization.

流体力学建模预测流动演变的能力既可以通过物理原理和经验方法。一方面，物理原理（例如保护法）是外反性的 - 它们提供了尚未观察到的现象的预测。另一方面，经验建模在数据中提供了相关功能。人工智能和机器学习在经验建模方面非常出色。在这个项目中，学生将将物理原理和经验模型结合到统一的方法中：由物理受限的数据驱动方法进行多物理优化。这些目标是限制机器学习中的湍流方程。这将使更准确，鲁棒性和概括。