저널 : JICRS(Journal of Institute of Control, Robotics and Systems), Vol.23, No.8, pp.650-660, 2017. 8.
· 논문제목 : “Generating Robot Task Plans from Action Ontology”
· 저자 : 이석준, 김인철
· 요약 : This paper presents an action ontology for service robots and the automated task planning algorithm based on this ontology. While an action ontology provides the conceptual model for robot actions and tasks, task plans represent the control model for behavior execution in a dynamic environment. The action ontology proposed in this paper defines not only the class hierarchy of actions, but also useful properties of individual actions like input parameters, preconditions, durative conditions, effects, and partial orderings of sub-actions. In order to generate task plans from action ontology, our task planning algorithm includes the model translation process which translates the conceptual model in OWL into the dynamic control model in JPL. Experimental studies are performed to check the applicability of the proposed action ontology and the task planning algorithm.