http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
탄소섬유적층 각도에 따른 CFRP 평판의 실내소음 특성
방성일(Sungil Bang),김연수(Yeun Soo Kim),백지선(Jiseon Back),이상권(Sang-Kwon Lee) 한국소음진동공학회 2018 한국소음진동공학회 논문집 Vol.28 No.5
Recently, the researche on the weight lightening of the vehicle body for increasing the fuel efficiency of automobile have been actively carried out. One way to reduce the weight is to change the material applied to the car body. In recent years, CFRP laminated plates, which has lower density and similar strength to conventional metals, has been increasingly applied to the car body. However, since the CFRP laminated plate has different sound and vibration characteristics compared with conventional metal plate, the research on the sound radiation characteristics is required for the successful application. This paper studies the effect of sound and vibration characteristics of CFRP plates according to the fiber lamination angle on interior noise of the enclosure made of CFRP laminated plate. The purpose of this paper is numerically to predict the interior noise of the enclosure caused by the vibration of the CFRP plates with different laminated angles of carbon fiber. The numerical method is validated by theoretical and experimental method.
소음 신호를 이용한 딥러닝 이용 파워 드라이빙 시스템의 건전성 감시
김선원(Seon-Won Kim),안강현(Kanghyeon An),백지선(Jiseon Back),이상권(Sang-Kwon Lee),이창호(Changho Lee),김풍길(Pungil Kim) 한국소음진동공학회 2021 한국소음진동공학회 논문집 Vol.31 No.1
The power driving system (PDS) comprises parts such as the chain, sprocket, gear, bearing, and rotating shaft. The purpose of this study is to develop a condition-monitoring device that diagnoses component defects early by using a convolutional neural network to prevent complete damage due to component defects. For this study, eight types of defects are artificially manufactured in various parts and assembled to build a PDS. A convolutional neural network is developed to classify and diagnose the eight types of defects. A feature for faults is successfully extracted, and fault classification is achieved with 90 % accuracy.