J. Queißer, B. Hammer, H. Ishihara, M. Asada, Jochen J. Steil
{"title":"Skill Memories for Parameterized Dynamic Action Primitives on the Pneumatically Driven Humanoid Robot Child Affetto","authors":"J. Queißer, B. Hammer, H. Ishihara, M. Asada, Jochen J. Steil","doi":"10.1109/DEVLRN.2018.8761040","DOIUrl":null,"url":null,"abstract":"In this work, we propose an extension of parameterized skills to achieve generalization of forward control signals for action primitives that result in an enhanced control quality of complex robotic systems. We argue to shift the complexity of learning the full dynamics of the robot to a lower dimensional task related learning problem. Due to generalization over task variability, online learning for complex robots as well as complex scenarios becomes feasible. We perform an experimental evaluation of the generalization capabilities of the proposed online learning system through simulation of a compliant 2DOF arm. Scalability to a complex robotic system is demonstrated on the pneumatically driven humanoid robot Affetto including 6DOF.","PeriodicalId":236346,"journal":{"name":"2018 Joint IEEE 8th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"78 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 Joint IEEE 8th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DEVLRN.2018.8761040","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 5
Abstract
In this work, we propose an extension of parameterized skills to achieve generalization of forward control signals for action primitives that result in an enhanced control quality of complex robotic systems. We argue to shift the complexity of learning the full dynamics of the robot to a lower dimensional task related learning problem. Due to generalization over task variability, online learning for complex robots as well as complex scenarios becomes feasible. We perform an experimental evaluation of the generalization capabilities of the proposed online learning system through simulation of a compliant 2DOF arm. Scalability to a complex robotic system is demonstrated on the pneumatically driven humanoid robot Affetto including 6DOF.