Publications
Detailed Information
Learning and Generalization of Dynamic Movement Primitives by Hierarchical Deep Reinforcement Learning : 계층적 심층 강화학습을 활용한 동적 단위 동작의 학습 및 일반화
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | 김현진 | - |
dc.contributor.author | 김원철 | - |
dc.date.accessioned | 2018-12-03T01:46:34Z | - |
dc.date.available | 2018-12-03T01:46:34Z | - |
dc.date.issued | 2018-08 | - |
dc.identifier.other | 000000153294 | - |
dc.identifier.uri | https://hdl.handle.net/10371/143961 | - |
dc.description | 학위논문 (석사)-- 서울대학교 대학원 : 공과대학 기계항공공학부, 2018. 8. 김현진. | - |
dc.description.abstract | This paper presents an approach to learn and generalize robotic skills from a demonstration using deep
reinforcement learning (deep RL). Dynamic Movement Primitives (DMPs) formulate a nonlinear differential equation and produce the observed movement from a demonstration. However, it is hard to generate new behaviors from using DMPs. Thus, we apply DMPs framework into deep RL as an initial setting for learning the robotic skills. First, we build a network to represent this differential equation, and learn and generalize the movements by optimizing the shape of DMPs with respect to the rewards up to the end of each sequence of movement primitives. In order to do this, we consider a deterministic actor-critic algorithm for deep RL and we also apply a hierarchical strategy. This drastically reduces the search space for a robot by decomposing the task, which allows to solve the sparse reward problem from a complex task. In order to integrate DMPs with hierarchical deep RL, the differential equation is considered as temporal abstraction of option. The overall structure is mainly composed of two controllers: meta-controller and sub-controller. The meta-controller learns a policy over intrinsic goals and a sub-controller learns a policy over actions to accomplish the given goals. We demonstrate our approach on a 6 degree-of-freedom (DOF) arm with a 1-DOF gripper and evaluate that DMPs are learned and generalized using deep RL with a pick-and-place task. | - |
dc.description.tableofcontents | 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Thesis contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.3 Thesis outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2 Background Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1 Dynamic Movement Primitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.2 Reinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3 Model-free Reinforcement Learning with Dynamic Movement Primitives . . . . . . . . . . . . 10 3.1 Object detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.2 Pre-training DMPs for networks of deep RL . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.3 Reinforcement Learning to learn DMPs . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 4 Hierarchical Deep Reinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 5 Evaluation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 5.1 Learning and improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 5.2 Learning and Generalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 5.3 Pick and Place Manipulation Task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 | - |
dc.format | application/pdf | - |
dc.format.medium | application/pdf | - |
dc.language.iso | en | - |
dc.publisher | 서울대학교 대학원 | - |
dc.subject.ddc | 621 | - |
dc.title | Learning and Generalization of Dynamic Movement Primitives by Hierarchical Deep Reinforcement Learning | - |
dc.title.alternative | 계층적 심층 강화학습을 활용한 동적 단위 동작의 학습 및 일반화 | - |
dc.type | Thesis | - |
dc.contributor.AlternativeAuthor | Wonchul Kim | - |
dc.description.degree | Master | - |
dc.contributor.affiliation | 공과대학 기계항공공학부 | - |
dc.date.awarded | 2018-08 | - |
- Appears in Collections:
- Files in This Item:
Item View & Download Count
Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.