http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이완기 ( Wan-gi Lee ),조범연 ( Beom-yeon Cho ),이한세 ( Han-se Lee ),이강주 ( Kang-ju Lee ),김형훈 ( Hyung-hoon Kim ),심현민 ( Hyeon-min Shim ) 한국정보처리학회 2022 한국정보처리학회 학술대회논문집 Vol.29 No.2
화재로 발생하는 산업시설의 인명ㆍ재산 피해를 줄이고 기존 소방 설비의 단점을 보완하는 소방 로봇을 제안한다. 소방 로봇은 무인 시스템으로 설계되었으며 6개의 핵심 기능인화재 감지, 화재 진압, 현장 이동, 화재 알림, 소방서 신고, 현장 모니터링으로 구성된다. 로봇의 구성은 구동부, 제어부, 소화부로 이루어져 있으며, 각 구성 중 일부를 선정하고 테스트 통하여 화재 진압에 유효함을 증명하였다.
Hertz-Mindlin 접촉 모델이 적용된 음해적 입자법을 통한 다물체 충돌해석
이완기(Wan Ki Lee),김기덕(Gi Deok Kim),변제윤(Je Yun Byun),주영석(Young Suk Joo) 대한기계학회 2021 대한기계학회 춘추학술대회 Vol.2021 No.11
Existing Smoothed particle hydrodynamics(SPH) structural analysis model has two problems: tensile instability resulting in particle clumping and the use of non-physical artificial stress. In this study, to solve the above two issues, the structural analysis is implemented with total Lagrangian based corotated SPH, and the Hertz-Mindlin contact model is applied as the interface condition for collision of multi-body. For numerical validation of structural analysis model, dynamic behavior simulation of cantilever beam is performed. Transient analysis of deflection at the cantilever beam tip is conducted, and maximum deflection and frequency are calculated and compared with finite element method(FEM) results and analytical solution. For numerical validation of collision model, transient analysis for reaction force between two elastic rings is performed. The stress distribution and deformation shape at specific time steps are compared with the FEM results.
Multi-GPU 기반 SPH에 대한 동적 부하분산 전략
김기덕(Gi-Deok Kim),이완기(Wan-Ki Lee),변제윤(Je-Yun Byun),주영석(Young-Seok Joo) 대한기계학회 2021 대한기계학회 춘추학술대회 Vol.2021 No.11
Graphics processing unit(GPU) has been developed as much research has been conducted on smoothed particle hydrodynamics(SPH). Therefore, many researchers developed a variety of particle methods based on GPU as well as studies on multi-GPU instead of single-GPU are conducted. One of most widely adopted load balancing algorithms is the fixed analysis area division. The fixed region load balancing algorithm has the advantage that the implementation of load balancing is not complicated, since the particle distribution is not necessary to be considered. However, in the case of a dynamic flow analysis as dam-break, there is a drawback that particles are not distributed evenly into each device. To solve this issue, this paper presents a breakthrough that allocates the number of particles evenly to each GPU. Thus, it can be stated that the proposed load balancing algorithm has a higher applicability than the fixed analysis area division algorithms.