DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Noh, Jun-Yong | - |
dc.contributor.advisor | 노준용 | - |
dc.contributor.advisor | Shin, Sung-Yong | - |
dc.contributor.advisor | 신성용 | - |
dc.contributor.author | Han, Da-Seong | - |
dc.contributor.author | 한다성 | - |
dc.date.accessioned | 2015-04-23T08:30:40Z | - |
dc.date.available | 2015-04-23T08:30:40Z | - |
dc.date.issued | 2014 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=591851&flag=dissertation | - |
dc.identifier.uri | http://hdl.handle.net/10203/197830 | - |
dc.description | 학위논문(박사) - 한국과학기술원 : 전산학과, 2014.8, [ vi, 94 p. ] | - |
dc.description.abstract | Predictive control is indispensable in human locomotion, in particular, locomotion steering and balance recovery. Humans perform prediction fast so as to move as intended as well as to respond to an unexpected external perturbation or environmental change on the fly. However, it is challenging in physics-based animation to make prediction as fast as humans do so that a realistic motion can be synthesized in real time according to an on-line user input. The main difficulty in on-line real-time motion control arises from the complexity and high dimensionality of the dynamics for a full-body character. In this thesis, we present an on-line real-time physics-based approach to motion control with contact repositioning based on a low-dimensional dynamics model using example motion data. Our approach first generates a reference motion in run time according to an on-line user request by transforming an example motion extracted from a motion library. Our motion transformation method exploits biomechanical observations to improve the quality of the reference motion. Guided by the reference motion, it repeatedly generates an optimal control policy for a small time window one at a time for a sequence of partially overlapping windows, each covering a couple of footsteps of the reference motion, which supports an on-line performance. On top of this, our system dynamics and problem formulation allow to derive closed-form derivative functions by exploiting the low-dimensional dynamics model together with example motion data. These derivative functions and their sparse structures facilitate a real-time performance. Our approach also allows contact foot repositioning so as to robustly respond to an external perturbation or an environmental change as well as to perform locomotion tasks such as stepping on stones effectively. | eng |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | low-dimensional | - |
dc.subject | physics-based | - |
dc.subject | character animation | - |
dc.subject | real-time | - |
dc.subject | on-line | - |
dc.subject | data-driven | - |
dc.subject | motion synthesis | - |
dc.subject | motion control | - |
dc.subject | 보행 | - |
dc.subject | 모델예측제어 | - |
dc.subject | 저차원 | - |
dc.subject | 물리 기반 | - |
dc.subject | 캐릭터 애니메이션 | - |
dc.subject | 실시간 | - |
dc.subject | 온라인 | - |
dc.subject | 데이터 구동 | - |
dc.subject | 동작 합성 | - |
dc.subject | 동작 제어 | - |
dc.subject | locomotion | - |
dc.subject | model predictive control | - |
dc.title | Data-driven on-line real-time physics-based locomotion synthesis based on low-dimensional model predictive control | - |
dc.title.alternative | 저차원 모델예측제어를 이용한 데이터 구동 온라인 실시간 물리 기반 보행 동작 생성 | - |
dc.type | Thesis(Ph.D) | - |
dc.identifier.CNRN | 591851/325007 | - |
dc.description.department | 한국과학기술원 : 전산학과, | - |
dc.identifier.uid | 020085348 | - |
dc.contributor.localauthor | Noh, Jun-Yong | - |
dc.contributor.localauthor | 노준용 | - |
dc.contributor.localauthor | Shin, Sung-Yong | - |
dc.contributor.localauthor | 신성용 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.