http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
ZTA 제조시 알루미나 입자크기가 치밀화 거동에 미치는 영향
채지훈,조범래,Chae, Jihoon,Cho, Bumrae 한국재료학회 2013 한국재료학회지 Vol.23 No.4
In order to increase the toughness of ZTA(zirconia toughened alumina) ceramics, the present study focused on rearrangement and densification of particles according to the particle size of the parent material. When rough alumina was used for production of ZTA, densification behavior was observed in the specimen sintered at a temperature over $1550^{\circ}C$. However, it was found that the densification behavior was occurred in the specimen sintered at $1450^{\circ}C$ when fine alumina powder was used. High relative density exceeding 98% was obtained when fine alumina powder was mixed with 15 wt% of 3Y-TZP and sintered at $1600^{\circ}C$. Also, a hardness of 1820.2 Hv was obtained when a specimen containing 10 wt% of 3Y-TZP was sintered at $1600^{\circ}C$. In the case of 3Y-TZP containing rough alumina powder that had been sintered the hardness value was around 1720.3 Hv. It was predicted that an improved toughening effect in ZTA could be achieved by using finer alumina powder as the parent material.
SDN에서 강화학습을 통한 효과적인 멀티캐스트 라우팅 트리 생성 방법
채지훈(Jihun Chae),이병대(Byoung-Dai Lee),김남기(Namgi Kim) 한국정보기술학회 2020 한국정보기술학회논문지 Vol.18 No.10
Along with the development of artificial intelligence technology, researches that apply reinforcement learning to routing problems in the network field are emerging. However, the basic reinforcement learning method assumes a fixed environment, so performance is limited in variable network environment that varies over time. Therefore, we proposes a deep reinforcement learning-based multicast routing tree construction method that can overcome these limitations and reflect the variable network environment in SDN. To evaluate the method proposed, experiments were performed to compare performance in various network topology. As a result, It was found that the deep reinforcement learning agent learned by proposed method in various network topology produced optimal close multicast routing tree than deep reinforcement learning agent learned in fixed network topology.