Flow Control with 'Smart Walls'.

Abstract

We have conducted a study of 'smart walls' which combines microelectromechanical sensors and actuators with a neural network controller in a single layer of silicon for active flow control. We have determined the optimum arrangement of sensors and actuators for a single neural network controller under the influence of noise. With sensors and actuator implemented in hardware and the neural network simulated by computer, wind-tunnel experiments have demonstrated that the control unit is capable of controlling TS waves and wave packets. The experimental results are in good agreement with our previous numerical simulations of the control process. Using DNS for realistic actuator configurations, we have developed an efficient procedure to provide training data for the network controller. In preparation for a smart wall with multiple control units, we have developed DNS-based techniques to analyze the flow response to multiple actuator movements

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 29, 1995
Accession Number
ADA304464

Entities

People

  • Thorwald Herbert

Organizations

  • Ohio State University

Tags

Communities of Interest

  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Actuators
  • Agreements
  • Computers
  • Flow
  • Hypervelocity Flow
  • Neural Networks
  • Simulations
  • Simulators
  • Training
  • Wave Packets
  • Wind Tunnels

Fields of Study

  • Physics

Readers

  • Combustion and Flow Dynamics.
  • Computer Networking
  • Materials Science and Engineering.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - Bayesian Inference
  • AI & ML - Neural Networks
  • Microelectronics
  • Microelectronics - Microelectromechanical Systems