Terrain Traversal Cost Learning with Knowledge Transfer Between Multi-legged Walking Robot Gaits

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F22%3A00359487" target="_blank" >RIV/68407700:21230/22:00359487 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/ICARSC55462.2022.9784790" target="_blank" >https://doi.org/10.1109/ICARSC55462.2022.9784790</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICARSC55462.2022.9784790" target="_blank" >10.1109/ICARSC55462.2022.9784790</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Terrain Traversal Cost Learning with Knowledge Transfer Between Multi-legged Walking Robot Gaits

Popis výsledku v původním jazyce

The terrain traversal abilities of multi-legged walking robots are affected by gaits, the walking patterns that enable adaptation to various operational environments. Fast and lowset gaits are suited to flat ground, while cautious and highset gaits enable traversing rough areas. A suitable gait can be selected using prior experience with a particular terrain type. However, experience alone is insufficient in practical setups, where the robot experiences each terrain with only one or just a few gaits and thus would infer novel gait-terrain interactions from insufficient data. Therefore, we use knowledge transfer to address unsampled gait-terrain interactions and infer the traversal cost for every gait. The proposed solution combines gaitterrain cost models using inferred gait-to-gait models projecting the robot experiences between different gaits. We implement the cost models as Gaussian Mixture regressors providing certainty to identify unknown terrains where knowledge transfer is desirable. The presented method has been verified in synthetic showcase scenarios and deployment with a real walking robot. The proposed knowledge transfer demonstrates improved cost prediction and selection of the appropriate gait for specific terrains.

Název v anglickém jazyce

Terrain Traversal Cost Learning with Knowledge Transfer Between Multi-legged Walking Robot Gaits

Popis výsledku anglicky

The terrain traversal abilities of multi-legged walking robots are affected by gaits, the walking patterns that enable adaptation to various operational environments. Fast and lowset gaits are suited to flat ground, while cautious and highset gaits enable traversing rough areas. A suitable gait can be selected using prior experience with a particular terrain type. However, experience alone is insufficient in practical setups, where the robot experiences each terrain with only one or just a few gaits and thus would infer novel gait-terrain interactions from insufficient data. Therefore, we use knowledge transfer to address unsampled gait-terrain interactions and infer the traversal cost for every gait. The proposed solution combines gaitterrain cost models using inferred gait-to-gait models projecting the robot experiences between different gaits. We implement the cost models as Gaussian Mixture regressors providing certainty to identify unknown terrains where knowledge transfer is desirable. The presented method has been verified in synthetic showcase scenarios and deployment with a real walking robot. The proposed knowledge transfer demonstrates improved cost prediction and selection of the appropriate gait for specific terrains.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2022 IEEE INTERNATIONAL CONFERENCE ON AUTONOMOUS ROBOT SYSTEMS AND COMPETITIONS (ICARSC)

ISBN

978-1-6654-8217-2

ISSN

2573-9360

e-ISSN

2573-9387

Počet stran výsledku

6

Strana od-do

148-153

Název nakladatele

Institute of Electrical and Electronics Engineers

Místo vydání

New York

Místo konání akce

Santa Maria da Feira

Datum konání akce

29. 4. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000838705300026