kth.sePublications KTH
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Learning to Land on Flexible Structures
KTH, School of Electrical Engineering and Computer Science (EECS).
2022 (English)Independent thesis Advanced level (degree of Master of Fine Arts (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Lära sig att landa på flexibla strukturer (Swedish)
Abstract [en]

Forests cover 30% of the Earth’s surface area, but most of the tree canopy reaches up to tens of meters above the ground, making it challenging to explore, and thus our knowledge of it is very limited. To help scientists further grasp the biological information in the tree canopy and sample the tree branches, we need a control algorithm that can land drones on branches with different flexibility. Because the flexibility of branches is unpredictable before landing, conventional model-based control methods are not up to the task. For this reason, we developed a reinforcement learning-based landing strategy. By controlling the UAV, interacting with different flexible branches in the simulation to collect data, and then applying the PPO and SAC algorithms for training, we obtained a control strategy that can land the UAV on arbitrarily flexible branches. Finally, the control algorithm is verified in the simulation.

Abstract [sv]

Skogarna täcker 30% av jordens yta, men det mesta av trädkronorna når upp till tiotals meter över marken, vilket gör det svårt att utforska dem, och därför är vår kunskap om dem mycket begränsad. För att hjälpa forskarna att ytterligare förstå den biologiska informationen i trädkronorna och ta prover från trädgrenarna behöver vi en kontrollalgoritm som kan landa drönare på grenar med olika flexibilitet. Eftersom grenarnas flexibilitet är oförutsägbar före landning klarar konventionella modellbaserade styrmetoder inte uppgiften. Därför har vi utvecklat en förstärkningsinlärningsbaserad landningsstrategi. Genom att styra drönaren, interagera med olika flexibla grenar i simuleringen för att samla in data och sedan tillämpa PPO- och SAC-algoritmerna för träning fick vi en kontrollstrategi som kan landa drönaren på godtyckligt flexibla grenar. Slutligen verifieras kontrollalgoritmen i simuleringen.

Place, publisher, year, edition, pages
2022. , p. 48
Series
TRITA-EECS-EX ; 2022:613
Keywords [en]
Reinforcement Learning, Aerial Physical Interaction, Landing, Aerial Robots
Keywords [sv]
Förstärkningsinlärning, fysisk interaktion i luften, landning, flygrobotar
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:kth:diva-321088OAI: oai:DiVA.org:kth-321088DiVA, id: diva2:1708706
External cooperation
ETH zurich
Educational program
Master of Science - Embedded Systems
Supervisors
Examiners
Available from: 2022-11-07 Created: 2022-11-05 Last updated: 2022-11-07Bibliographically approved

Open Access in DiVA

fulltext(15837 kB)534 downloads
File information
File name FULLTEXT01.pdfFile size 15837 kBChecksum SHA-512
6c83e656b37fb12c86d6f2b805d42a7f5d3879b176450e169bd97df31a53cc1810cc030298bc3edd189155b5c52aa97b82e0e539744ed4cd8529e7a5684214ba
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 535 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 338 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf