Neural network based trajectory optimization for unmanned aerial vehicles

Brian R. Geiger, Joseph Francis Horn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

A neural network approximation to direct trajectory optimization methods is presented. The method uses neural networks to approximate the dynamics and objective equations integrated over a given time interval. The trajectory is then built recursively and treated as a nonlinear programming problem. The method is compared to a direct collocation method as well as more recent pseudospectral methods and shows competitive results while being computationally faster. In addition, a neural network provides a continuously differentiable function approximation which may be advantageous when a discontinuous objective function is used in a nonlinear solver. A surveillance trajectory planning problem for an unmanned aerial vehicle is given as an example application and results are presented for all three methods.

Original languageEnglish (US)
Title of host publication47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition
StatePublished - 2009
Event47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition - Orlando, FL, United States
Duration: Jan 5 2009Jan 8 2009

Other

Other47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition
CountryUnited States
CityOrlando, FL
Period1/5/091/8/09

Fingerprint

trajectory optimization
pilotless aircraft
Unmanned aerial vehicles (UAV)
trajectory
Trajectories
Neural networks
nonlinear programming
trajectory planning
collocation
Nonlinear programming
surveillance
approximation
trajectories
intervals
Planning
method
vehicle

All Science Journal Classification (ASJC) codes

  • Space and Planetary Science
  • Aerospace Engineering

Cite this

Geiger, B. R., & Horn, J. F. (2009). Neural network based trajectory optimization for unmanned aerial vehicles. In 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition [2009-0054]
Geiger, Brian R. ; Horn, Joseph Francis. / Neural network based trajectory optimization for unmanned aerial vehicles. 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition. 2009.
@inproceedings{507bda03c1bf41bdb5497b876db0364e,
title = "Neural network based trajectory optimization for unmanned aerial vehicles",
abstract = "A neural network approximation to direct trajectory optimization methods is presented. The method uses neural networks to approximate the dynamics and objective equations integrated over a given time interval. The trajectory is then built recursively and treated as a nonlinear programming problem. The method is compared to a direct collocation method as well as more recent pseudospectral methods and shows competitive results while being computationally faster. In addition, a neural network provides a continuously differentiable function approximation which may be advantageous when a discontinuous objective function is used in a nonlinear solver. A surveillance trajectory planning problem for an unmanned aerial vehicle is given as an example application and results are presented for all three methods.",
author = "Geiger, {Brian R.} and Horn, {Joseph Francis}",
year = "2009",
language = "English (US)",
isbn = "9781563479694",
booktitle = "47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition",

}

Geiger, BR & Horn, JF 2009, Neural network based trajectory optimization for unmanned aerial vehicles. in 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition., 2009-0054, 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, Orlando, FL, United States, 1/5/09.

Neural network based trajectory optimization for unmanned aerial vehicles. / Geiger, Brian R.; Horn, Joseph Francis.

47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition. 2009. 2009-0054.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Neural network based trajectory optimization for unmanned aerial vehicles

AU - Geiger, Brian R.

AU - Horn, Joseph Francis

PY - 2009

Y1 - 2009

N2 - A neural network approximation to direct trajectory optimization methods is presented. The method uses neural networks to approximate the dynamics and objective equations integrated over a given time interval. The trajectory is then built recursively and treated as a nonlinear programming problem. The method is compared to a direct collocation method as well as more recent pseudospectral methods and shows competitive results while being computationally faster. In addition, a neural network provides a continuously differentiable function approximation which may be advantageous when a discontinuous objective function is used in a nonlinear solver. A surveillance trajectory planning problem for an unmanned aerial vehicle is given as an example application and results are presented for all three methods.

AB - A neural network approximation to direct trajectory optimization methods is presented. The method uses neural networks to approximate the dynamics and objective equations integrated over a given time interval. The trajectory is then built recursively and treated as a nonlinear programming problem. The method is compared to a direct collocation method as well as more recent pseudospectral methods and shows competitive results while being computationally faster. In addition, a neural network provides a continuously differentiable function approximation which may be advantageous when a discontinuous objective function is used in a nonlinear solver. A surveillance trajectory planning problem for an unmanned aerial vehicle is given as an example application and results are presented for all three methods.

UR - http://www.scopus.com/inward/record.url?scp=78549263010&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78549263010&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:78549263010

SN - 9781563479694

BT - 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition

ER -

Geiger BR, Horn JF. Neural network based trajectory optimization for unmanned aerial vehicles. In 47th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition. 2009. 2009-0054