Publication detail

Evolution Assisted Flight Control System Design

CHUDÝ, P. VLK, J. DITTRICH, P.

Original Title

Evolution Assisted Flight Control System Design

English Title

Evolution Assisted Flight Control System Design

Type

conference paper

Language

en

Original Abstract

An evolution driven controller design approach has been applied to a rigid body aircraft model of a light sport aircraft. The model comprises inertial, aerodynamic and flight dynamics related elements with the controller architecture based on Classical Control Theory. The evolution driven concept has played a significant role in the optimization of the proposed controller structure by providing tuned controller parameters, which meet the designed fitness function criteria imposed through the optimization problem formulation. The proposed fitness function combines significant controller stability evaluation conditions into a single abstraction. The use of a robust optimization framework based on the genetic algorithms has allowed the suggested form of multi-criteria optimization definition. The suitability of the evolutionary optimization has been successfully tested on a model with rigid body aircraft dynamics. Time-domain simulation results have shown the compliance of the tuned controller performance with its anticipated behavior.

English abstract

An evolution driven controller design approach has been applied to a rigid body aircraft model of a light sport aircraft. The model comprises inertial, aerodynamic and flight dynamics related elements with the controller architecture based on Classical Control Theory. The evolution driven concept has played a significant role in the optimization of the proposed controller structure by providing tuned controller parameters, which meet the designed fitness function criteria imposed through the optimization problem formulation. The proposed fitness function combines significant controller stability evaluation conditions into a single abstraction. The use of a robust optimization framework based on the genetic algorithms has allowed the suggested form of multi-criteria optimization definition. The suitability of the evolutionary optimization has been successfully tested on a model with rigid body aircraft dynamics. Time-domain simulation results have shown the compliance of the tuned controller performance with its anticipated behavior.

Keywords

Evolution, light sport aircraft, framework, time-domain, control, genetic algrithm.

RIV year

2013

Released

14.09.2013

Publisher

IEEE Computer Society

Location

Syracuse, NY

ISBN

978-1-4799-1536-1

Book

Proceedings of 32nd Digital Avionics Systems Conference

Edition

NEUVEDEN

Edition number

NEUVEDEN

Pages from

1

Pages to

9

Pages count

9

URL

Documents

BibTex


@inproceedings{BUT103566,
  author="Peter {Chudý} and Jan {Vlk} and Petr {Dittrich}",
  title="Evolution Assisted Flight Control System Design",
  annote="An evolution driven controller design approach has been applied to a rigid body
aircraft model of a light sport aircraft. The model comprises inertial,
aerodynamic and flight dynamics related elements with the controller architecture
based on Classical Control Theory. The evolution driven concept has played
a significant role in the optimization of the proposed controller structure by
providing tuned controller parameters, which meet the designed fitness function
criteria imposed through the optimization problem formulation. The proposed
fitness function combines significant controller stability evaluation conditions
into a single abstraction. The use of a robust optimization framework based on
the genetic algorithms has allowed the suggested form of multi-criteria
optimization definition. The suitability of the evolutionary optimization has
been successfully tested on a model with rigid body aircraft dynamics.
Time-domain simulation results have shown the compliance of the tuned controller
performance with its anticipated behavior.",
  address="IEEE Computer Society",
  booktitle="Proceedings of 32nd Digital Avionics Systems Conference",
  chapter="103566",
  doi="10.1109/DASC.2013.6712641",
  edition="NEUVEDEN",
  howpublished="electronic, physical medium",
  institution="IEEE Computer Society",
  year="2013",
  month="september",
  pages="1--9",
  publisher="IEEE Computer Society",
  type="conference paper"
}