Cristin-resultat-ID: 2039791
Sist endret: 27. juli 2022, 13:23
NVI-rapporteringsår: 2022
Resultat
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
2022

UAV Icing: Ice Shedding Detection Methods for an Electrothermal De-Icing System

Bidragsytere:
  • Bogdan Løw Hansen
  • Richard Hann og
  • Tor Arne Johansen

Bok

AIAA AVIATION 2022 Forum
ISBN:
  • 978-1-62410-635-4

Utgiver

American Institute of Aeronautics and Astronautics (AIAA)
NVI-nivå 1

Om resultatet

Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
Publiseringsår: 2022
ISBN:
  • 978-1-62410-635-4

Klassifisering

Fagfelt (NPI)

Fagfelt: Elektronikk og kybernetikk
- Fagområde: Realfag og teknologi

Beskrivelse Beskrivelse

Tittel

UAV Icing: Ice Shedding Detection Methods for an Electrothermal De-Icing System

Sammendrag

Among the identified problems with unmanned aerial vehicles (UAVs), especially with the medium-sized fixed-wing UAVs, is a lack of mature ice protection systems (IPSs). One of the emerging UAV ice protection technologies is based on electrothermal de-icing, where heating elements are integrated into the wings of a UAV. In icing conditions, this allows the UAV wings to be heated, resulting in weakened ice adhesion and subsequent ice shedding. A drawback of such de-icing system is that it requires substantial amounts of energy. This study addresses the issue of energy consumption by proposing two electrothermal ice shedding detection algorithms that make it possible for a de-icing system to be turned off shortly after ice shedding has occurred. This is in contrast to fixed de-icing periods where the heating elements operate irrespective of the ice shedding status. The first proposed algorithm is based on parameter change detection, which involves thresholding the innovation sequence of a first-order temperature-estimating Kalman filter (KF). The second algorithm is based on pattern matching using the continuous wavelet transform (CWT) of the KF innovation sequence. The results, based on icing wind tunnel experiments, show that both approaches can produce accurate detections with an F- measure of 1.0 and average detection times of 1.98 seconds and 2.53 seconds, respectively. Thus, making it possible to optimize the efficiency of electrothermal IPSs.

Bidragsytere

Bogdan Løw Hansen

  • Tilknyttet:
    Forfatter
    ved Institutt for teknisk kybernetikk ved Norges teknisk-naturvitenskapelige universitet
Aktiv cristin-person

Richard Hann

  • Tilknyttet:
    Forfatter
    ved Institutt for teknisk kybernetikk ved Norges teknisk-naturvitenskapelige universitet

Tor Arne Johansen

  • Tilknyttet:
    Forfatter
    ved Institutt for teknisk kybernetikk ved Norges teknisk-naturvitenskapelige universitet
1 - 3 av 3

Resultatet er en del av Resultatet er en del av

AIAA AVIATION 2022 Forum.

AIAA, Aviation. 2022, American Institute of Aeronautics and Astronautics (AIAA). Vitenskapelig antologi/Konferanseserie
1 - 1 av 1