Retrospective-Predictive Flow Control with Sensor/Actuator Constraints

Abstract

The purpose of this action is to provide FY24 CR3 funds, in the amount of $20K, for a new start grant award.--Approved for Public ReleaseThis project will apply adaptive control to flow control problems that are directly relevant toNaval applications. These computational investigations are based on retrospective-predictive adaptive control (RPAC) and computational fluid dynamics with output-based adaptive mesh refinement (CFD/OAMR). The project will investigate the achievable performance of RPAC in three subsonic scenarios that have direct Naval relevance, namely delay of separation/stall, maximizing the lift-to-drag ratio for anairfoil, and reducing non-uniformity in channel flows with recirculation. A key focus of these studies will be the incorporation ofrealistic sensors and actuators. We will also consider data-interconnection constraints, where RPAC is implemented with feedforward/feedback action and decentralized and sparse controller structures. Key goals of the project are to investigate the ability of RPACto learn the driving features of the fluid dynamics with minimal prior modeling, and to use the adjoint capability of CFD/OAMR to quantify discretization errors and investigate the sensitivity of the achievable performance as afunction of sensor/actuators specifications and placement.

Document Details

Document Type
DoD Grant Award
Publication Date
Apr 11, 2024
Source ID
N000142412285

Entities

People

  • Dennis S. Bernstein

Organizations

  • Board of Regents of the University of Michigan
  • Office of Naval Research
  • United States Navy

Tags

Fields of Study

  • Physics

Readers

  • Computational Fluid Dynamics (CFD)
  • Fluid Mechanics and Fluid Dynamics.
  • Robotics and Automation.