Data-Driven Control of Unsteady Flows
Abstract
The objective of the proposed work is to develop techniques to control fluid flows with large amplitude unsteadiness, for example due to gust disturbances or large aggressive maneuvers. To achieve the goals outlined in this project, we will explore emerging data-driven techniques in machine learning to 1) extract dominant patterns from high-dimensional fluid flow data, 2) robustly sense and estimate the flow from limited measurements, 3) develop reduced-order models for the flows, and 4) design controllers to modify the behavior of the flow. This work relies on the observation that even complex, unsteady flows exhibit large-scale coherent patterns that may facilitate efficient modeling, estimation, and control. The sharp acceleration experienced by the aerodynamic body generates vorticity along its surface that forms into large coherent vortical structures. Characterizing, modeling, and controlling these patterns may be formulated as high-dimensional, nonconvex optimization problems. For solving such problems, emerging techniques in machine learning provide a powerful approach for generalizations of existing linear modeling (e.g., POD-DMD) and control (e.g., LQR-LQG) techniques. The adaptive control techniques developed from the proposed efforts will be tested on progressively complex unsteady fluid flows. We in particular will consider two- and three-dimensional laminar and turbulent flows over airfoil in unsteady maneuver or under the influence of gust. Advanced flow control using these techniques will enable practical engineering goals including lift increase, drag reduction, and stable flight, which are critical to improve performance and efficiency in various aerospace applications. We anticipate that the use of machine learning based adaptive techniques will be transformative in how we examine and manipulate unsteady fluid flows.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 07, 2023
- Source ID
- FA95502110178
Entities
People
- Kunihiko Taira
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of California, Los Angeles