http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Aerodynamic design and performance analysis of multi-MW class wind turbine blade
Bumsuk Kim,Woojune Kim,Sungyoul Bae,Jaehyung Park,Manneung Kim 대한기계학회 2011 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.25 No.8
The rotor blade is an important device that converts kinetic energy of wind into mechanical energy. It affects power performance, efficiency of energy conversion, load and dynamic stability of a wind power generation system. This paper presents an aerodynamic design of 3 MW class blade using BEM and confirms that the design satisfies the initial design target by BEM and CFD analysis. To investigate the effects of radial flow at the inboard region, the result of static BEM analysis was compared with the result of CFD analysis. The result of quantitative comparison among thrust force, power coefficient and mechanical power depending on wind speed change is presented. Furthermore, design reference data such as pressure, streamline, torque and thrust force distribution on the blade surface is presented as well.
A Tool Pack Mechanism for DRM Interoperability
Bumsuk Choi,Youngbae Byun,Jeho Nam,Jinwoo Hong 한국전자통신연구원 2007 ETRI Journal Vol.29 No.4
As the number of digital content service providers increases, a variety of digital rights management (DRM) systems appear without supporting interoperability. The lack of interoperability in DRM systems causes inconvenience to customers, especially when they want to play content through multiple devices manufactured by different vendors. In this letter, we propose a novel method to support interoperability between different DRM systems. The proposed technique aims to build an open framework structure which satisfies DRM vendors’ requirements by enhancing the security of intellectual property management and the protection tools.
CNN based Sound Event Detection Method using NMF Preprocessing in Background Noise Environment
Bumsuk Jang,Sang-Hyun Lee 한국인터넷방송통신학회 2020 Journal of Advanced Smart Convergence Vol.9 No.2
Sound event detection in real-world environments suffers from the interference of non-stationary and time-varying noise. This paper presents an adaptive noise reduction method for sound event detection based on non-negative matrix factorization (NMF). In this paper, we proposed a deep learning model that integrates Convolution Neural Network (CNN) with Non-Negative Matrix Factorization (NMF). To improve the separation quality of the NMF, it includes noise update technique that learns and adapts the characteristics of the current noise in real time. The noise update technique analyzes the sparsity and activity of the noise bias at the present time and decides the update training based on the noise candidate group obtained every frame in the previous noise reduction stage. Noise bias ranks selected as candidates for update training are updated in real time with discrimination NMF training. This NMF was applied to CNN and Hidden Markov Model(HMM) to achieve improvement for performance of sound event detection. Since CNN has a more obvious performance improvement effect, it can be widely used in sound source based CNN algorithm.
Seo, Bumsuk,Lee, Jihye,Lee, Kyung-Do,Hong, Sukyoung,Kang, Sinkyu Elsevier 2019 Field crops research Vol.238 No.-
<P><B>Abstract</B></P> <P>Weather-related risks in crop production are not only crucial for farmers but also for market participants and policymakers since securing food supply is an important issue for society. Although crop growth condition and phenology represent essential information regarding such risks, extensive observations of these variables are virtually non-existent in many parts of the world. In this study, we developed an integrative approach to remotely monitor crop growth at a large scale. For corn and soybeans in Iowa and Illinois in the United States (2003–2015), we monitored crop growth and crop phenology with earth observation data and compared it against the United States Department of Agriculture National Agricultural Statistics Service (NASS) crop statistics. For crop phenology, we calculated three phenology metrics (i.e., start of season, end of season, and peak of season) at the pixel level from the MODIS 16-day Normalized Difference Vegetation Index (NDVI). For growth condition, we used two distinct approaches to acquire crop growth condition indicators: a process-based crop growth modeling and a satellite-NDVI-based method. Based on their pixel-wise historical distributions, we monitored relative growth strength and scaled-up that to the state-level. The estimates were compared with the crop progress and condition data of NASS. For the state-level phenology, the avg. root-mean-square-error (RMSE) of the estimates was 8.6 days for the all three metrics after bias correction. The absolute mean errors for the three metrics were smaller than 2.6 days after bias correction. For the condition, the state-level 10-day estimates showed moderate agreements with the observations (avg. RMSE = 10.02%). Notably, the condition estimates were sensitive to the severe degradation in 2003, 2012, and 2013 for both crops. In 2010, 2011 and 2013, unusually high errors occurred at the very beginning stage of growth (DOY 140–150), which attenuated over time. As the cumulative biomass and NDVI showed little change in comparison to the period mean biomass and NDVI for the spikes, this seems to be an error associated with variations in growth timing. Overall, the model using accumulated NDVI (S5) is preferable due to its performance and methodological simplicity. The proposed approach enables us to monitor crop growth for any given period and place where long-term statistics are available. It can be used to assist crop monitoring at large scales.</P>