Â鶹¾«Ñ¡

Join us as Xuanhong An shares insights about Optimization and Control of Unsteady Fluid Dynamic Systems.

Professor Xuanhong An joined the College of Aeronautics and Engineering at Â鶹¾«Ñ¡ in 2024 as an Assistant Professor. His research interests are at the intersection of control systems and fluid dynamics, focusing on the control of applications in unsteady flows. He has extensive experience in optimal control and reduced-order modeling of unsteady flows. Currently, his research is centered on enhancing aircraft performance.

Abstract

Systems in fluid dynamics, such as aerial/marine bio-inspired robots, aircraft and energy harvesting systems, often encounter unsteady flows, which can result in undesired performance reduction as well as catastrophic consequences. On the other hand, by controlling the unsteady flow, we can not only mitigate its negative impact but also utilize control strategies to optimize the performance of such applications. However, controlling an unsteady flow is fraught with the following challenges: 1) its high dimensionality requires large amounts of computational resources and 2) the nonlinear governing equations (Navier–Stokes equations) that describe the fluid mechanics make it difficult to apply linear mathematical tools to obtain low-order models. Several advances in the control and optimization of fluid flows have been made, including the development of efficient high-fidelity computational methods to optimize unsteady flows under known working conditions, and low-order modeling techniques as a part of real-time controllers in the presence of unpredictable environmental disturbances. This talk explores the research landscape of applications of optimization methods and mathematical modeling methods to fluid mechanics in two parts. The first part of this talk will discuss a high-fidelity adjoint-based optimization technique for unsteady fluid applications involving high dimensional parameter spaces. The second part will focus on low-order modeling, including data-driven modeling techniques as the key to the efficient use of real-time control of unsteady flows.