204 0

Ontology-based Unified Robot Knowledge and Its Utilization for Sustainable Service

Title
Ontology-based Unified Robot Knowledge and Its Utilization for Sustainable Service
Author
임기현
Alternative Author(s)
Lim, Gi Hyun
Advisor(s)
서일홍
Issue Date
2010-08
Publisher
한양대학교
Degree
Doctor
Abstract
지능적인 서비스 로봇은 부분적으로 또는 완전히 자율적으로 인간에게 서비스를 제공하게 된다. 로봇에 의한 서비스 임무 수행은 로봇의 여러 분야에서 관심이 집중되고 있다. 서비스 로봇은 인간 생활을 지원하기 위하여 사용자의 요구 사항과 로봇이 생활하는 공간을 이해하여야 하고, 로봇이 수행할 수 있는 기본적인 동작을 조합하여 주어진 복잡한 임무를 수행 할 수 있어야 한다.실제 환경에서 서비스를 수행하기 위해, 서비스 로봇은 개방적이고, 동적이며 비 구조적인 환경에서 취득된 지식을 사용하여야 한다. 로봇 지능을 구현하기 위하여 통합된 로봇 지식 프레임워크와 이를 적용할 수 있는 지식 결합 방법을 필요로 하게 된다. 본 논문은 우선, 공유 및 증식이 가능한 로봇을 위한 통합 지식 체계를 구축한다. 둘째, 인식 오류가 발생하는 실제 응용환경에서 지식 인스턴스를 생성하기 위한 강인한 지식 생성 방법을 제시한다. 세째, 사람이나 다른 로봇 등에 의해서 변화 될 수 있는 불완전한 지식하에서 대안적인 행동을 제시함으로써 서비스를 지속 가능케 하는 방법을 제시하고, 마지막으로, 효율적인 지식 추론을 가능케 하는 상황 추론 방법을 제시한다. 통합 로봇 지식 체계의 지식 프레임워크와 강인 지식 생성, 대안 제시 및 상황 추론의 지식 결합 방법을 통하여 지속 가능한 로봇 서비스를 로봇 지능을 구현 가능케 할 수 있음을 실험적인 방법을 통하여 평가를 하였다.; Intelligent service robots operate semi or fully autonomously to perform services useful to the well being of human and equipment. The robotic service which are delivered to satisfy the needs, wants, or aspirations of humans is up-and-coming application to receive attention in various research fields of robotics. To perform service tasks in real environment, service robots are typically complex systems using the knowledge from open, dynamic and unstructured environments. Knowledge represented by semantic network is a key for service robots to successfully complete service tasks in real environments. Semantic knowledge can be represented by network or graph which represents semantic relations between concepts. Additionally, service robots are designed to complete service tasks semi or fully automatically in a service environment. Service robots will need to understand semantic relationships of objects, spaces and contexts in order to assist humans in their everyday lives, and then the robot must carry out its service tasks with its primitive behaviors. For this, a robot needs many kinds of data from low level sensor data to high level symbolic data. High level perceptual tasks such as context awareness, object recognition and navigation are essential for intelligent robots. Also a robot must combine its atomic behaviors to complete the high level service such as delivery service. To offer sustainable robotic services, service robots must accumulate knowledge by using recognition results and choose a action for services intelligently. There are four issues to implement semantic robot intelligence: robot-centered knowledge to integrate from low-level data to high-level knowledge, robust knowledge instantiation and update by using imperfect sensing data such as misidentification of object recognition, selection of alternative actions even with incomplete knowledge , and context reasoning to restrict and pay attention to related domains. In this dissertation, three knowledge management methods and one unified knowledge framework are presented for sustainable robotic service. An ontology-based unified robot knowledge framework that integrates low-level data with high-level semantic knowledge to enable service robot intelligence is proposed as knowledge framework. Unified robot knowledge framework enables a robot to robustly recognize objects and successfully navigate in spite of hidden and partial information, SAAC recursively suggest alternative action instead of domain rules, and RoKI enables robots to detect falsity of object recognition for robust knowledge instantiation, where spatial reasoning, temporal reasoning, movable properties and data confidences are considered. Additionally, SCoRPaK enables a service robot to understand situations which is useful to interact more easily with humans in service task environments.
URI
https://repository.hanyang.ac.kr/handle/20.500.11754/141020http://hanyang.dcollection.net/common/orgView/200000414733
Appears in Collections:
GRADUATE SCHOOL[S](대학원) > ELECTRONICS AND COMPUTER ENGINEERING(전자컴퓨터통신공학과) > Theses (Ph.D.)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE