Learned parametrized dynamic movement primitives with shared synergies for controlling robotic and musculoskeletal systems
Publikationen: Beitrag in Fachzeitschrift › Artikel › Forschung › (peer-reviewed)
Standard
in: Frontiers in computational neuroscience, Jahrgang 7.2013, Nr. October, 17.10.2013.
Publikationen: Beitrag in Fachzeitschrift › Artikel › Forschung › (peer-reviewed)
Harvard
APA
Vancouver
Author
Bibtex - Download
}
RIS (suitable for import to EndNote) - Download
TY - JOUR
T1 - Learned parametrized dynamic movement primitives with shared synergies for controlling robotic and musculoskeletal systems
AU - Rückert, Elmar
AU - d'Avella, Andrea
PY - 2013/10/17
Y1 - 2013/10/17
N2 - Asalientfeatureofhumanmotorskilllearningistheabilitytoexploitsimilaritiesacrossrelatedtasks.Inbiologicalmotorcontrol,ithasbeenhypothesizedthatmusclesynergies,coherentactivationsofgroupsofmuscles,allowforexploitingsharedknowledge.Recentstudieshaveshownthatarichsetofcomplexmotorskillscanbegeneratedbyacombinationofasmallnumberofmusclesynergies.Inrobotics,dynamicmovementprimitivesarecommonlyusedformotorskilllearning.Thismachinelearningapproachimplementsastableattractorsystemthatfacilitateslearninganditcanbeusedinhigh-dimensionalcontinuousspaces.However,itdoesnotallowforreusingsharedknowledge,i.e.,foreachtaskanindividualsetofparametershastobelearned.Weproposeanovelmovementprimitiverepresentationthatemploysparametrizedbasisfunctions,whichcombinesthebenefitsofmusclesynergiesanddynamicmovementprimitives.Foreachtaskasuperpositionofsynergiesmodulatesastableattractorsystem.Thisapproachleadstoacompactrepresentationofmultiplemotorskillsandatthesametimeenablesefficientlearninginhigh-dimensionalcontinuoussystems.Themovementrepresentationsupportsdiscreteandrhythmicmovementsandinparticularincludesthedynamicmovementprimitiveapproachasaspecialcase.Wedemonstratethefeasibilityofthemovementrepresentationinthreemulti-tasklearningsimulatedscenarios.First,thecharacteristicsoftheproposedrepresentationareillustratedinapoint-masstask.Second,incomplexhumanoidwalkingexperiments,multiplewalkingpatternswithdifferentstepheightsarelearnedrobustlyandefficiently.Finally,inamulti-directionalreachingtasksimulatedwithamusculoskeletalmodelofthehumanarm,weshowhowtheproposedmovementprimitivescanbeusedtolearnappropriatemuscleexcitationpatternsandtogeneralizeeffectivelytonewreachingskills.
AB - Asalientfeatureofhumanmotorskilllearningistheabilitytoexploitsimilaritiesacrossrelatedtasks.Inbiologicalmotorcontrol,ithasbeenhypothesizedthatmusclesynergies,coherentactivationsofgroupsofmuscles,allowforexploitingsharedknowledge.Recentstudieshaveshownthatarichsetofcomplexmotorskillscanbegeneratedbyacombinationofasmallnumberofmusclesynergies.Inrobotics,dynamicmovementprimitivesarecommonlyusedformotorskilllearning.Thismachinelearningapproachimplementsastableattractorsystemthatfacilitateslearninganditcanbeusedinhigh-dimensionalcontinuousspaces.However,itdoesnotallowforreusingsharedknowledge,i.e.,foreachtaskanindividualsetofparametershastobelearned.Weproposeanovelmovementprimitiverepresentationthatemploysparametrizedbasisfunctions,whichcombinesthebenefitsofmusclesynergiesanddynamicmovementprimitives.Foreachtaskasuperpositionofsynergiesmodulatesastableattractorsystem.Thisapproachleadstoacompactrepresentationofmultiplemotorskillsandatthesametimeenablesefficientlearninginhigh-dimensionalcontinuoussystems.Themovementrepresentationsupportsdiscreteandrhythmicmovementsandinparticularincludesthedynamicmovementprimitiveapproachasaspecialcase.Wedemonstratethefeasibilityofthemovementrepresentationinthreemulti-tasklearningsimulatedscenarios.First,thecharacteristicsoftheproposedrepresentationareillustratedinapoint-masstask.Second,incomplexhumanoidwalkingexperiments,multiplewalkingpatternswithdifferentstepheightsarelearnedrobustlyandefficiently.Finally,inamulti-directionalreachingtasksimulatedwithamusculoskeletalmodelofthehumanarm,weshowhowtheproposedmovementprimitivescanbeusedtolearnappropriatemuscleexcitationpatternsandtogeneralizeeffectivelytonewreachingskills.
UR - https://cps.unileoben.ac.at/wp/Frontiers2013bRueckert.pdf
U2 - 10.3389/fncom.2013.00138
DO - 10.3389/fncom.2013.00138
M3 - Article
VL - 7.2013
JO - Frontiers in computational neuroscience
JF - Frontiers in computational neuroscience
SN - 1662-5188
IS - October
ER -
