• 풍력에너지저널
• /
• 제5권1호
• /
• pp.22-32
• /
• 2014

The structural design of a wind turbine must show the verification of the structural integrity of all load-carrying components. Also, design load calculations shall be performed using appropriate and accurate methods. In this study, advanced numerical approach for the calculation of design loads based on unsteady computational fluid dynamics (CFD) is presented considering extreme design load conditions such as the extreme coherent gust (ECG) and the 50 year extreme operating gust (EOG). Unsteady aerodynamic loads are calculated based on Reynolds average Navier-Stokes (RANS) equations with shear-stress transport k-ω(SST k-ω) turbulent model. A full three-dimensional 5 MW offshore wind-turbine model with rotating blades, hub, nacelle, and tower configuration is practically considered and its aerodynamic interference effect among blades, nacelle, and tower is also accurately considered herein. Calculated blade loads based on unsteady CFD method with respect to blade azimuth angle are compared with those by NREL FAST code and physically investigated in detail.

• 대한토목학회논문집
• /
• 제39권6호
• /
• pp.667-673
• /
• 2019

지구 온난화로 인해 기후변화가 가속화되고 이에 따라 강풍에 대한 재해가 늘어날 것으로 판단된다. 이에 본 연구에서는 시간에 따른 선형 경향성을 고려한 비정상성 빈도해석 모형을 구축하기 위한 방법으로 Bayesian 기법을 적용하였다. 그리고 제주공항 지점의 연 최대풍속자료를 이용하여 극치분포 매개변수들의 사후분포를 추정하고 비정상성 빈도해석을 수행하였다. 재현기간 100년 빈도의 풍속을 추정한 결과를 보면, 경향성이 통계적으로 유의하며 이로 인해 비정상성 빈도해석에 의한 기본풍속이 정상성 빈도해석의 기본풍속보다 크게 추정되고 있다. 이처럼 기상자료의 정상성을 가정한 현재의 빈도해석 절차는 경향성이 존재하는 지역의 경우에 미래의 기본풍속을 과소 추정할 가능성이 크다고 판단된다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권7호
• /
• pp.1705-1720
• /
• 2013

Human face gender recognition requires fast image processing with high accuracy. Existing face gender recognition methods used traditional local features and machine learning methods have shortcomings of low accuracy or slow speed. In this paper, a new framework for face gender recognition to reach fast face gender recognition is proposed, which is based on Local Ternary Pattern (LTP) and Extreme Learning Machine (ELM). LTP is a generalization of Local Binary Pattern (LBP) that is in the presence of monotonic illumination variations on a face image, and has high discriminative power for texture classification. It is also more discriminate and less sensitive to noise in uniform regions. On the other hand, ELM is a new learning algorithm for generalizing single hidden layer feed forward networks without tuning parameters. The main advantages of ELM are the less stringent optimization constraints, faster operations, easy implementation, and usually improved generalization performance. The experimental results on public databases show that, in comparisons with existing algorithms, the proposed method has higher precision and better generalization performance at extremely fast learning speed.

• Wind and Structures
• /
• 제17권2호
• /
• pp.227-244
• /
• 2013

Non-stationary extreme winds such as thunderstorm downbursts are responsible for many structural damages. This research presents a time domain approach for estimating along-wind load effects on tall buildings using multiple wind speed time history samples, which are simulated from evolutionary power spectra density (EPSD) functions of non-stationary wind fluctuations using the method developed by the authors' earlier research. The influence of transient wind loads on various responses including time-varying mean, root-mean-square value and peak factor is also studied. Furthermore, a simplified model is proposed to describe the non-stationary wind fluctuation as a uniformly modulated process with a modulation function following the time-varying mean. Finally, the probabilistic extreme response and peak factor are quantified based on the up-crossing theory of non-stationary process. As compared to the time domain response analysis using limited samples of wind record, usually one sample, the analysis using multiple samples presented in this study will provide more statistical information of responses. The time domain simulation also facilitates consideration of nonlinearities of structural and wind load characteristics over previous frequency domain analysis.

• 드라이브 ㆍ 컨트롤
• /
• 제9권4호
• /
• pp.14-18
• /
• 2012

Two kinds of piezoelectric actuator are applied to the valve for controlling the direction, the flow and the pressure of the fluid. One is a stack type piezoelectric actuator which has very fast response characteristics but very tiny displacement. The other is a bender type piezoelectric actuator which has also fast response characteristics but lower than the stack type one, and has longer displacement than the stack type one. So, the bender type piezoelectric actuator has advantage to apply to the valve for controlling a large amount of flow and fast on-off operating. In this study, the bender type piezoelectric pneumatic valve for color sorter is designed and fabricated. The new type high speed piezo valve with the both side supporting mechanism for high operating frequency and high reliability is discussed for separating the foreign body from the grains. Finally, the performance characteristics of a fabricated valve are analyzed and the frequency characteristics are also discussed for substituting the conventional type solenoid actuator.

• 대한토목학회논문집
• /
• 제28권2A호
• /
• pp.197-205
• /
• 2008

본 연구에서는 광양-묘도 지역의 평균풍속을 추정하기 위하여, 일별 최대 풍속과 해당 방향에 대한 결합분포확률의 통계학적 해석에 극한값 확률분포 모델이 사용되었다. 이를 위하여, 교량 가설지점 인근의 기상관측소에 대한 일별 최대풍속 및 해당풍향의 데이터로부터 각각의 관측소에 대한 일별 최대기록의 빈도를 조사하였으며, 16방위 및 전방위에 대한 년 최대풍속의 표본을 추출하였다. 이러한 풍속기록은 Gumbel 및 Weibull 분포모델에 적용하였으며, 모멘트방법 및 최소제곱법 등을 통해 모수를 추정하였다. 또한, PPCC 검사를 통해 분포모델 및 모수의 적합 여부를 검사하였다. 적합 여부가 판단된 모수로부터, 해당 관측소별로 데이터의 표본 크기 및 교량 가설지점으로부터의 거리에 대한 요소를 고려하여 16방위 및 전방위에 대한 년 최대풍속을 추정하였다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권2호
• /
• pp.125-130
• /
• 2016

The proposal of this study is a fast version of the conventional extreme learning machine (ELM), called pseudoinverse matrix decomposition based incremental ELM (PDI-ELM). One of the main problems in ELM is to determine the number of hidden nodes. In this study, the number of hidden nodes is automatically determined. The proposed model is an incremental version of ELM which adds neurons with the goal of minimization the error of the ELM network. To speed up the model the information of pseudoinverse from previous step is taken into account in the current iteration. To show the ability of the PDI-ELM, it is applied to few benchmark classification datasets in the University of California Irvine (UCI) repository. Compared to ELM learner and two other versions of incremental ELM, the proposed PDI-ELM is faster.

• 한국해안해양공학회지
• /
• 제10권3호
• /
• pp.120-124
• /
• 1998

본 연구는 인공항 속에 떠 있는 원자로를 토네이도로부터 보호하기 시작되었다. 연구결과에 의하면 토네이도가 유발할 수 있는 파고는 8.2 m이다. 본 연구에 있어 바람의 마찰 전단력에 의하여 발생하는 파는 작다고 가정하여 무시하였다.

• Journal of the Korean Data and Information Science Society
• /
• 제28권4호
• /
• pp.797-810
• /
• 2017

한반도는 매년 태풍의 위험에 노출되어 있다. 태풍은 강풍과 강우가 동반되는 열대성 저기압으로 사회 경제적으로 막대한 피해를 유발한다. 현재의 자연재해 경고 시스템은 풍속과 강우를 구분하여 위험을 감지토록 설계되어 강풍과 폭우를 동반한 태풍의 위험을 경고하는데 한계점이 존재한다. 코플라모형은 확률변수들 사이의 복잡한 의존성 구조를 파악하기 위해 단변량분포의 집합을 다변량분포로 연결하는 모형으로 강우, 홍수, 가뭄 등의 분야에서 활발하게 연구되고 있다. 본 연구에서는 한반도에서 태풍에 가장 많이 노출된 도시인 부산과 제주도의 기상 관측소 (ASOS)에서 수집된 1904년 4월 9일부터 2015년 12월 31일까지 일강수량 (precipitation), 일최대풍속 (maximum wind speed) 자료를 이용하였다. 각 변수의 주변부확률을 추정하기 위해 두꺼운 꼬리 분포인 로그정규분포, 감마분포, 와이블분포를 고려하였다. 주변부 확률분포의 적합성검정은 Kolmogorov-Smirnov와 Cramervon-Mises, Anderson-Darling 검정통계량을 이용하였다. 코플라모형을 위해 순위를 기반으로 한 유사자료 (pseudo observation)를 생성하여 두 변수 간 의존성을 추정하였다. 강풍과 폭우의 의존성을 설명하기 위한 코플라모형으로 타원형, 나선형, 극단치 코플라모형이 고려되었다. 코플라모형의 적합성은 Cramer-von-Mises로 검정하였고, 교차검증을 통해 최적모형을 선택하였다. 연구결과 일강우량과 풍속의 주변부 확률분포로 대부분 로그정규분포가 적합하였다. 부산의 일평균풍속에 따른 일강우량은 t 코플라, 일최대풍속에 따른 일강우량은 Clayton 코플라가 최적모형으로 선정되었다. 제주도의 일최대풍속에 따른 일강우량은 정규코플라, 일강우량에 따른 일평균풍속은 Frank 코플라, 일강우량에 따른 일최대풍속은 Husler-Reiss 코플라가 최적모형으로 선택되었다.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제6권2호
• /
• pp.442-458
• /
• 2014

Wave induced vibrations increase the fatigue and extreme loading, but this is normally neglected in design. The industry view on this is changing. Wave induced vibrations are often divided into springing and whipping, and their relative contribution to fatigue and extreme loading varies depending on ship design. When it comes to displacement vessels, the contribution from whipping on fatigue and extreme loading is particularly high for certain container vessels. A large modern design container vessel with high bow flare angle and high service speed has been considered. The container vessel was equipped with a hull monitoring system from a recognized supplier of HMON systems. The vessel has been operating between Asia and Europe for a few years and valuable data has been collected. Also model tests have been carried out of this vessel to investigate fatigue and extreme loading, but model tests are often limited to head seas. For the full scale measurements, the correlation between stress data and wind data has been investigated. The wave and vibration damage are shown versus heading and Beaufort strength to indicate general trends. The wind data has also been compared to North Atlantic design environment. Even though it has been shown that the encountered wind data has been much less severe than in North Atlantic, the extreme loading defined by IACS URS11 is significantly exceeded when whipping is included. If whipping may contribute to collapse, then proper seamanship may be useful in order to limit the extreme loading. The vibration damage is also observed to be high from head to beam seas, and even present in stern seas, but fatigue damage in general is low on this East Asia to Europe trade.