Neural Networks for Dynamic Flight Control

Abstract

This thesis examines the application of artificial neural networks (NNs) to the problem of dynamic flight control. The specific application is the control of a flying model helicopter. The control interface is provided through a hardware and software test bed called the Fast Adaptive Maneuvering Experiment (FAME). The NN design approach uses two NNs: one trained as an emulator of the plant and the other trained to control the emulator. The emulator neural network is designed to reproduce the flight dynamics of the experimental plant. The controller is then designed to produce the appropriate control inputs to drive the emulator to a desired final state. Previous research in the area of NNs for controls has almost exclusively been applied to simulations. To develop a controller for a real plant, a neural network must be created which will accurately recreate the dynamics of the plant. This thesis demonstrates the ability of a neural network to emulate a real, dynamic, nonlinear plant.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1993
Accession Number
ADA274089

Entities

People

  • Ronald E. Setzer

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Air Force
  • Aircrafts
  • Computer Programming
  • Computer Programs
  • Computer Science
  • Computer Simulations
  • Computers
  • Control Systems
  • Fixed Wing Aircraft
  • Helicopters
  • Human Factors Engineering
  • Mathematical Analysis
  • Mathematical Models
  • Measurement
  • Neural Networks
  • Pattern Recognition
  • Simulations

Readers

  • Neural Network Machine Learning.
  • Parallel and Distributed Computing.
  • Robotics and Automation.

Technology Areas

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