http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Structural Damage Detection using Deep Convolutional Neural Network and Transfer Learning
Chuncheng Feng,Hua Zhang,Shuang Wang,Yonglong Li,Haoran Wang,Fei Yan 대한토목학회 2019 KSCE JOURNAL OF CIVIL ENGINEERING Vol.23 No.10
During the long-term operation of hydro-junction infrastructure, water flow erosion causes concrete surfaces to crack, resulting in seepage, spalling, and rebar exposure. To ensure infrastructure safety, detecting such damage is critical. We propose a highly accurate damage detection method using a deep convolutional neural network with transfer learning. First, we collectedimages from hydro-junction infrastructure using a high-definition camera. Second, we preprocessed the images using an imageexpansion method. Finally, we modified the structure of Inception-v3 and trained the network using transfer learning to detectdamage. The experiments show that the accuracy of the proposed damage detection method is 96.8%, considerably higher thanthe accuracy of a support vector machine. The results demonstrate that our damage detection method achieves better damagedetection performance.
Application of Amorphous Nanoparticle Fe-B Magnetic Fluid in Wastewater Treatment
Chuncheng Yang,Mengchun Yu,Xiuling Cao,Xiufang Bian 성균관대학교(자연과학캠퍼스) 성균나노과학기술원 2019 NANO Vol.14 No.9
Amorphous magnetic particles demonstrate excellent comprehensive properties and outstanding characteristics for numerous applications. In this report, magnetic crystalline Fe3O4 and amorphous Fe-B nanoparticles were successfully synthesized and introduced to prepare water-based magnetic fluids. The Fe3O4 and Fe-B particles are homogeneous nanoparticles with an average particle size of 12~15 nm. The shape of Fe-B amorphous nanoparticles is regular. The saturation magnetizations of Fe-B and Fe3O4 particles are 74 emu/g and 69 emu/g. The use of crystalline Fe3O4 magnetic fluid and amorphous Fe-B magnetic fluid in advanced treatment of high concentration organic wastewater was presented. The removal rate of chemical oxygen demand by using the amorphous Fe-B magnetic fluid reached 96%, about 16% higher than that by using the Fe3O4 magnetic fluid. Moreover, compared with Fe3O4 magnetic fluid, the treatment results demonstrate that the decolorizing effect by using the amorphous Fe-B magnetic fluid was 20% higher. It has been found that the nano-size Fe-B particles in magnetic fluid with amorphous structure led to high efficiency of wastewater treatment due to the catalytic activity.
Gong, Chuncheng,He, Kuang,Lee, Gun-Do,Chen, Qu,Robertson, Alex W.,Yoon, Euijoon,Hong, Suklyun,Warner, Jamie H. American Chemical Society 2016 ACS NANO Vol.10 No.10
<P>An in situ heating holder inside an aberration-corrected transmission electron microscope (AC-TEM) is used to investigate the real-time atomic level dynamics associated with heterogeneous nucleation and growth of graphene from Au nanoparticle seeds. Heating monolayer graphene to an elevated temperature of 800 degrees C removes the majority of amorphous carbon adsorbates and leaves a clean surface. The aggregation of Au impurity atoms into nanoparticle clusters that are bound to the surface of monolayer graphene causes nucleation of secondary graphene layers from carbon feedstock present within the microscope chamber. This enables the in situ study of heterogeneous nucleation and growth of graphene at the atomic level. We show that the growth mechanism consists of alternating C cluster attachment and indentation filling to maintain a uniform growth front of lowest energy. Back-folding of the graphene growth front is observed, followed by a process that involves flipping back and attaching to the surrounding region. We show how the highly polycrystalline graphene seed evolves with time into a higher order crystalline structure using a combination of AC-TEM and tight-binding molecular dynamics (TBMD) simulations. This helps understand the detailed lowest-energy step-by-step pathways associated with grain boundaries (GB) migration and crystallization processes. We find the motion of the GB is discontinuous and mediated by both bond rotation and atom evaporation, supported by density functional theory calculations and TBMD. These results provide insights into the formation of crystalline seed domains that are generated during bottom-up graphene synthesis.</P>
Spatially Dependent Lattice Deformations for Dislocations at the Edges of Graphene
Gong, Chuncheng,He, Kuang,Robertson, Alex W.,Yoon, Euijoon,Lee, Gun-Do,Warner, Jamie H. American Chemical Society 2015 ACS NANO Vol.9 No.1
<P>We show that dislocations located at the edge of graphene cause different lattice deformations to those located in the bulk lattice. When a dislocation is located near an edge, a decrease in the rippling and increase of the in-plane rotation occurs relative to the dislocations in the bulk. The increased in-plane rotation near the edge causes bond rotations at the edge of graphene to reduce the overall strain in the system. Dislocations were highly stable and remained fixed in their position even when located within a few lattice spacings from the edge of graphene. We study this behavior at the atomic level using aberration-corrected transmission electron microscopy. These results show detailed information about the behavior of dislocations in 2D materials and the strain properties that result.</P><P><B>Graphic Abstract</B> <IMG SRC='http://pubs.acs.org/appl/literatum/publisher/achs/journals/content/ancac3/2015/ancac3.2015.9.issue-1/nn505996c/production/images/medium/nn-2014-05996c_0009.gif'></P><P><A href='http://pubs.acs.org/doi/suppl/10.1021/nn505996c'>ACS Electronic Supporting Info</A></P>
Shuang Du,Chuncheng Zuo 보안공학연구지원센터 2015 International Journal of u- and e- Service, Scienc Vol.8 No.3
Compared with the traditional fuel vehicle, the pure electric vehicle has excellent characteristics in the emissions and energy use. But its driving ranges are much shorter than the traditional fuel vehicle. It has become a bottleneck problem for the development of electric vehicle. It is difficult to establish the accurate model of driving ranges in the actual working condition. Its main reason is that the influence factors of electric vehicle driving ranges and driving ranges have a non-linear relationship. BP neural network can map the complex non-linear relationship and has strong non-linear fitting ability. Compared the driving ranges of fuzzy control with fuzzy PI control in pure electric vehicle with dual-energy storage system, the results of simulation experiment indicate that the fuzzy PI control can extend the driving rages of electric vehicle. The maximum value of the average error is 2.66% in BP neural network prediction.
Research on Crack Segmentation Method of Hydro-Junction Project Based on Target Detection Network
Jie Pang,Hua Zhang,Chuncheng Feng,Linjing Li 대한토목학회 2020 KSCE JOURNAL OF CIVIL ENGINEERING Vol.24 No.9
The defect detection is an important task for maintaining the hydro-junction project. A two-stage crack defect segmentation method based on target detection network is proposed to solve the problem of severe brightness imbalance and large noise in dam surface images. In the first stage, to improve the ability to locate crack areas, Inception Resnet V2 is used as feature extraction network to help Faster-RCNN extract more effective deep features, and the brightness, contrast of image is randomly adjusted before training. In the second segmentation stage, the crack areas are segmented at pixel-level using K-means. The experimental results on the self-made crack image dataset show that the location accuracy (AP) of the crack areas can be improved by 1.9%, reaching 96.8%, compared with other segmentation networks that do not locate crack areas, the intersection over union for segmentation of cracks (Iou) of the final segmentation results is at least 9.4% higher, reaching 52.7%. This method can provide effective technical support for inspection work of hydro-junction project.
유아 교육용 어플리케이션 동향분석 및 발전방향에 관한 연구
ZHANG YUANYUAN,KUO CHUNCHENG,여등승(YU DENGSHENG) 한국IT서비스학회 2018 한국IT서비스학회 학술대회 논문집 Vol.2018 No.-
2015년까지 교육용 어플리케이션의 규모는 애플 스토어 중 게임 다음으로 두 번째를 기록하며 약 15억 위안을 넘어 약 70,000여개의 어플리케이션이 차지하고 있다. 교육용 어플리케이션에서 인기 다운로드의 80%이상이 유아 대상이다. 본 논문에서는 중국에 유아 교육용 어플리케이션을 대상으로 살펴보고 개발자를 위한 유아 교육용 어플리케이션의 발전방향을 논의하는데 그 목적이다. 연구결과, MAU 상위 50개의 유아 교육용 어플리케이션의 내용은 언어를 바탕으로 나타났으며, 인앱 구매는 32%인 것으로 나타났다. 또는 유아 교육용 어플리케이션의 내용과 평가, 내용과 사용자, 내용과 지불 여부, 사용자와 평가 사이 유의한 것으로 나타났다.