• 한국지구과학회지
• /
• 제38권1호
• /
• pp.11-23
• /
• 2017

풍력 자원의 단기 예측 가능성은 풍력 발전 단지의 경제적 타당성을 평가하는 중요한 요소이다. 본 연구에서는 풍력 자원의 단기 예측 가능성을 향상시키는 방법의 하나로 베이지안 칼만 필터를 후처리 과정으로 적용하였다. 이때 추정된 모델과 관측 데이터의 상관관계를 평가하기 위하여 일정 시간 동안 베이지안 칼만 훈련 기간이 요구된다. 본 연구는 여러 훈련 기간에 따라 예측 특성을 정량적으로 분석하였다. 태백 지역에서는 3일 단기 베이지안 칼만 훈련으로 기온과 풍속을 예측하는 것이 다른 훈련 기간을 적용할 때보다 우수한 예측 성능을 보였다. 반면 이어도는 6일 이상의 베이지안 칼만 필터의 훈련 기간을 적용한 경우 가장 좋은 예측 성능을 나타낸다. WRF 예측 성능이 떨어지는 사례에서 베이지안 칼만 필터의 예측 성능향상이 뚜렷하게 나타나며, 반대로 WRF 예측이 정확한 지점에서는 필터적용에 따른 성능향상 정도가 약한 경향을 가진다.

• 한국항행학회논문지
• /
• 제23권6호
• /
• pp.597-604
• /
• 2019

인공신경망은 뇌의 뉴런들에서 상호 작용과 경험을 통해 학습해 나가는 것을 모사해 만든 알고리즘으로, 데이터의 특성이 반영된 학습을 통하여 정확한 결과를 산출하는데 사용할 수 있는 방법이다. 본 연구에서 기상 역학 모델에서 예측된 풍속 값의 개선을 위하여 심층신경망을 이용한 모델을 제시하였다. 연구에서 제시한 심층신경망을 이용한 풍속 예측 개선 모델은 기상 역학 모델의 예측 값을 재 보정하는 모델을 구축하고 이에 대한 검증과 시험 과정 후 별도의 데이터를 통한 예측의 정확도를 높일 수 있는 것을 확인하였다. 풍속 예측의 개선을 위하여 예측 시간, 온도, 기압, 습도, 대기상태변수, 풍속 등과 같은 일반적 기상 현상 자료의 예측 값을 활용한 심층신경망을 구축하였고, 전체 데이터 중 일부 데이터는 모델의 적정성 확인용 데이터로 구분하여, 모델 구축 및 학습에 사용하지 않고 별도의 정확도를 확인하여 연구에서 제시한 방법의 적합성을 확인하였다.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2005년도 추계 학술대회논문집
• /
• pp.275-279
• /
• 2005

The vortex convection and induced flow field behind the KARI 3m x 4m LSWT gust generator was computed by using Computational Fluid Dynamics. For the accurate simulation of vortex convection, inviscid, laminar, Spalart-Allmars k-e and k-w turbulence models were tested with the NAL gust generator configuration and Spalart-Allmaras turbulence model was selected for the prediction of induced flow field behind the KARI LSWT gust generator. The wind tunnel test was also carried out at KARI LSWT and the results were compared with CFD prediction.

• International Journal of Advanced Culture Technology
• /
• 제11권4호
• /
• pp.393-405
• /
• 2023

Accurate prediction of wind speed and power is vital for enhancing the efficiency of wind energy systems. Numerous solutions have been implemented to date, demonstrating their potential to improve forecasting. Among these, deep learning is perceived as a revolutionary approach in the field. However, despite their effectiveness, the noise present in the collected data remains a significant challenge. This noise has the potential to diminish the performance of these algorithms, leading to inaccurate predictions. In response to this, this study explores a novel feature engineering approach. This approach involves altering the data input shape in both Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) and Autoregressive models for various forecasting horizons. The results reveal substantial enhancements in model resilience against noise resulting from step increases in data. The approach could achieve an impressive 83% accuracy in predicting unseen data up to the 24th steps. Furthermore, this method consistently provides high accuracy for short, mid, and long-term forecasts, outperforming the performance of individual models. These findings pave the way for further research on noise reduction strategies at different forecasting horizons through shape-wise feature engineering.

• 한국추진공학회지
• /
• 제24권5호
• /
• pp.56-64
• /
• 2020

29.7% 축소 NASA 표준 연구 모형에 대하여 소형 저속 풍동에서 시험을 수행하였다. 풍동시험 모형은 NASA Langley 연구소와 AIAA 항력추정 워크숍 위원회의 지침에 따라 알루미늄으로 제작하였다. 풍동의 능력 한계 등으로 상대적으로 낮은 레이놀즈수인 0.3 × 10⁶ 에서 정적 힘과 모멘트를 측정하였다. 세 가지 타입(핀 스팅, 블레이드 스팅, 벨리 스팅) 모형지지부에 대하여 피칭모멘트를 비교하였다. 보정된 벨리 스팅과 핀 스팅에 의한 피칭모멘트 보정량은 비슷하였으며 블레이드 스팅에 의한 피칭모멘트 값이 가장 적었다.

• 한국환경과학회지
• /
• 제21권4호
• /
• pp.517-533
• /
• 2012

This study calculated wind speed at the height of 10 m using a disaster prediction model(Florida Public Hurricane Loss Model, FPHLM) that was developed and used in the United States. Using its distributions, a usable information of surface wind was produced for the purpose of disaster prevention when the typhoon attack. The advanced research version of the WRF (Weather Research and Forecasting) was used in this study, and two domains focusing on South Korea were determined through two-way nesting. A horizontal time series and vertical profile analysis were carried out to examine whether the model provided a resonable simulation, and the meteorological factors, including potential temperature, generally showed the similar distribution with observational data. We determined through comparison of observations that data taken at 700 hPa and used as input data to calculate wind speed at the height of 10 m for the actual terrain was suitable for the simulation. Using these results, the wind speed at the height of 10 m for the actual terrain was calculated and its distributions were shown. Thus, a stronger wind occurred in coastal areas compared to inland areas showing that coastal areas are more vulnerable to strong winds.

• 한국대기환경학회지
• /
• 제13권1호
• /
• pp.65-77
• /
• 1997

A theoretical and numerical investigation on the boundary-layer flow over a two- or three-dimensional hill is presented. The numerical model is based on the finite volume method with boundary-fitted coordinates. The k-$\varepsilon$ turbulence model with modified wall function and the low-Reynolds-number model are employed. The hypothesis of Reynolds number independency for the atmospheric boundary-layer flow over aerodynamically rough terrain is confirmed by the numerical simulation. Comparisons of the mean velocity profiles and surface pressure distributions between the numerical predictions and the wind-tunnel experiments on the flow over a hill show good agreement. The linear theory provides generally good prediction of speed-up characteristics for the gentle-sloped hills. The flow separation occurs in the hill slope of 0.5 and the measured reattachment points are compared with the numerical prediction. It is found that the k- $\varepsilon$ turbulence model is reasonably accurate in predicting the attached flow, while the low- Reynolds-number model is more suitable to simulate the separated flows.ows.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1994년도 Proceedings of the Korea Automatic Control Conference, 9th (KACC) ; Taejeon, Korea; 17-20 Oct. 1994
• /
• pp.202-206
• /
• 1994

Mathematical model of the pollution rate of substation insulators is constructed, taking the model parameters as wind speed, wind direction, typhoon conditions and rainfall in an hourly basis. The main feature of model construction is to distinguish the effect of each parameter by separately analyzing the positive and negative pollution causing factors. Model parameters for the insulators of Karatsu substation, Saga, Japan were estimated and model validation was done using the actual data, in which the pollution deposits on the insulators were measured using pilot insulator and 'salt meter'. The proposed model of the pollution rate [mg/cm$^{2}$/hr] enables the identification of the effective parameters and prediction of the pollution rate so that it helps for the automatic decision making for insulator cleaning or the model can be used as a tool for the substation engineers to make precautionary measures.

• Wind and Structures
• /
• 제5권2_3_4호
• /
• pp.151-164
• /
• 2002

Predictions of the pedestrian level wind speeds for the downtown area of Auckland that have been obtained by wind tunnel and computational fluid dynamic (CFD) modelling are presented. The wind tunnel method involves the observation of erosion patterns as the wind speed is progressively increased. The computational solutions are mean flow calculations, which were obtained by using the finite volume code PHOENICS and the $k-{\varepsilon}$ turbulence model. The results for a variety of wind directions are compared, and it is observed that while the patterns are similar there are noticeable differences. A possible explanation for these differences arises because the tunnel prediction technique is sensitivity to gust wind speeds while the CFD method predicts mean wind speeds. It is shown that in many cases the computational model indicates high mean wind speeds near the corner of a building while the erosion patterns are consistent with eddies being shed from the edge of the building and swept downstream.

• 한국전산유체공학회지
• /
• 제13권2호
• /
• pp.55-61
• /
• 2008

The Reynolds-Averaged Navier-Stokes(RANS) analysis of the 3-D steady flow around the NREL Phase VI horizontal axis wind turbine(HAWT) rotor was performed. The CFD analysis results were compared with experimental data at several different wind speeds. The present CFD model shows good agreements with the experiments both at low wind speed which formed well-attache flow mostly on the upper surface of the blade, and at high wind speed which blade surface flow completely separated. However, some discrepancy occurs at the relatively high wind speeds where mixed attached and separated flow formed on the suction surface of the blade. It seems that the discrepancy is related to the onset of stall phenomena and consequently separation prediction capability of the current turbulence model. It is also found that strong span-wise flow occurs in stalled area due to the centrifugal force generated by rotation of the turbine rotor and it prevents abrupt reduction of normal force for higher wind speed than the designed value.