Physics-Informed Reinforcement Learning-based Multiple-Input Multiple-Output Flow Control

Abstract

The proposed research aims to develop a multiple-input multiple-output (MIMO) control framework based on reinforcement learning and informed by physical insights from theoretical analysis. The proposed MIMO flow control will be decentralized, as each controller independently works towards a uniform objective or separated objectives based on specified learning models. With this design philosophy, potential risks associated with centralized flow control frameworks are mitigated, leading to effective, efficient, and robust control.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 05, 2025
Source ID
FA95502410069

Entities

People

  • Qiong Liu

Organizations

  • Air Force Office of Scientific Research
  • New Mexico State University
  • United States Air Force

Tags

Readers

  • Neural Network Machine Learning.
  • Radio communications and signal processing.
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms