• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.1107-1108
• /
• 2011

본 논문에서는 전력용 변압기 온도상승을 예측하기 위해 CFD 상용 프로세서인 Fluent를 이용하였다. 온도상승의 원인이 되는 전력손실은 자계 상용 프로세서인 Maxwell을 이용하였으며, 자계해석에 의해 얻은 전력손실을 유체역학과 열전달을 동시에 고려한 열유동해석의 열원으로 적용하였다. 해석의 정확도를 향상시키기 위해 변압기 권선의 형상을 실제형상과 유사하게 모델링하였으며, 해석결과의 타당성을 검증하기 위해 온도 상승 시험을 통해 얻은 측정값과 비교하였다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.2
• /
• pp.83-94
• /
• 2019

Recently, the importance of Total Life Cycle System Management (TLCSM) and LIFE-CYCLE COSTS management is increasing in the development of weapon systems. In cost management, cost forecasting is important from the initial development stage, but it is difficult to predict the total life cycle cost at the development stage. In this study, we propose efficient management cost calculation and management at the development stage of the weapon system by comparison analysis between the PRICE-HL model and NemoSIM to calculate the maintenance cost under the CAIV concept. Based on the study results, further in-depth analyzes of the PRICE-HL model and NemoSIM input values / results are performed. In addition, we provide a more accurate method of calculating the cost of maintaining and operating the weapon system and a plan to utilize the result of NemoSIM in the ILS element development.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.16 no.6
• /
• pp.32-40
• /
• 2012

For predicting eigenfrequency and initial growth rate of combustion instabilities in lean premixed gas turbine combustor, linear thermoacoustic analysis model was developed in the current paper. A model combustor was selected for the model validation, which has well-defined inlet and outlet conditions and a relatively simple geometry, compared to the combustor in the previous works. Analytical linear equations for thermoacoustic waves were derived for a given combustion system. It was found that the prediction results showed a good agreement with the measurements, even though there was underestimation for instability frequencies. This underestimation was more obvious for a longer flame (i.e. wider temperature distribution) than for a shorter flame.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2013.04a
• /
• pp.795-795
• /
• 2013

본 논문에서는 에어컨 실외기의 내부 소음원에 의한 방사소음의 수준을 예측하기 위한 해석시스템을 구축하였다. 에어컨 실외기를 구성하는 여러 구성품을 내부 모델과 외부 모델로 구분하여 모델링하고, 주요한 소음원에 대한 최소한의 해석을 통하여 에어컨 실외기의 방사소음 수준을 예측하였다. 에어컨 실외기의 주요한 소음원 데이터를 실험을 통하여 계측하고, 이를 소음원 입력값으로 하는 진동해석 및 방사소음해석을 실시하여 구성품별 기여도를 예측 및 내부 구성품에 의한 소음원 데이터를 도출하였다. 해석을 통해 도출된 소음원 데이터를 이용하여 구조-음향 연성해석을 실시하여 방사소음수준을 예측하고, 실험을 통한 계측결과와 비교하였다.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.7
• /
• pp.11-18
• /
• 2023

In this paper, we evaluate deep learning time series forecasting models. Recent studies show that those models perform better than the traditional prediction model such as ARIMA. Among them, recurrent neural networks to store previous information in the hidden layer are one of the prediction models. In order to solve the gradient vanishing problem in the network, LSTM is used with small memory inside the recurrent neural network along with BI-LSTM in which the hidden layer is added in the reverse direction of the data flow. In this paper, we compared the performance of Informer by comparing with other models (LSTM, BI-LSTM, and Transformer) for real Nitrogen dioxide (NO2) data. In order to evaluate the accuracy of each method, mean square root error and mean absolute error between the real value and the predicted value were obtained. Consequently, Informer has improved prediction accuracy compared with other methods.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2009.05a
• /
• pp.1265-1271
• /
• 2009

본 연구에서는 강우의 시 공간적 분포의 불규칙한 변동성을 고려한 강우-유출예측모형을 위해 인공신경망(Artificial Neural Networks: ANNs)의 기법의 일종인 자기조직화(Self Organizing Map: SOM) 이론과 역전파 학습 알고리즘(Back Propagation Algorithm: BPA) 이론을 복합적으로 이용하였다. 기존의 인공신경망 연구에서 야기된 저 갈수기의 유출량에 대한 과대평가, 홍수기의 유출량에 대한 과소평가, 예측값이 연속적으로 선행 유출량을 나타내는 Persistence 현상을 해결하기 위하여 패턴분류 성능을 지닌 SOM 이론을 예측모형의 전처리 과정으로 이용하였다. 먼저, 본 연구에서 제안한 방법은 SOM에 의해 강우-유출 관계를 분류하고, SOM에 의한 분류에 따라 각각의 모형을 구성한다. 개별적으로 구축된 모형은 유출량의 예측을 위해 각각의 양상에 따라 분류된 자료를 이용한다. 결과적으로 본 연구에서 제안한 방법은 과거의 인공신경망의 일반적인 적용에 의한 결과보다 더 나은 예측능력을 보여주었으며, 더불어 유출량의 과소 및 과대추정과 Persistence 현상과 같은 문제점이 나타나지 않았다. 또한 강우량 및 유출량의 범위에 제한을 받지 않는 강우-유출예측 모형의 개발 및 홍수기로부터 갈수기까지의 보다 넓은 범위의 유출량의 예측에 기여할 것으로 기대된다.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.14 no.4
• /
• pp.189-195
• /
• 2012

To predict the blooming date of 'Yumyeong' peach trees, the models for flower bud developmental rate (DVR) were constructed. The DVRs were calculated from the demanded times at controlled air temperatures. The branches of 'Yumyeong' peach trees were incubated at three different temperatures of 9.7, 15.2, and $18.9^{\circ}C$. The DVRs were also constructed with blooming dates and air temperatures in the field, collected from 1979 to 2008 at the experimental orchard of National Institute of Horticultural and Herbal Science, Suwon, Korea. All the DVRs increased linearly or exponentially with air temperature. The DVR equations evaluated under controlled air temperatures were y=0.0018x+0.0051 and y=$0.0125e^{0.0603x}$. The DVR equations under field conditions were calculated as y=0.0039x-0.0112 and y=$0.0062e^{0.1512x}$. These DVR equations offer developmental indices and predict the date for blooming with air temperature data. These DVR equations were validated against the blooming data observed in the field. When the blooming dates were calculated with exponential DVR equations and daily air temperature data, the root mean squared errors between the observed and predicted dates were around 2 days. These results suggest that the DVR models are useful to predict the blooming date of 'Yumyeong' peach trees.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.42 no.10
• /
• pp.1690-1700
• /
• 2013

The objective of this study is to develop and validate predictive growth models for Bacillus cereus (diarrhea type) vegetative cells, spores and Staphylococcus aureus in preprocessed Namul (bracken and Chwinamul) and root vegetables (bellflower and burdock). For validation of model performance, growth data for S. aureus in preprocessed vegetables were collected at independent temperatures (18 and $30^{\circ}C$) not used in the model development. In addition, model performance of B. cereus (diarrhea type) in preprocessed vegetables was validated with an emetic type of B. cereus strain. In primary models, the specific growth rate (SGR) of the B. cereus spores was faster than that of the B. cereus vegetative cells, regardless of the kinds of vegetables at 24 and $35^{\circ}C$, while lag time (LT) of the B. cereus spores was longer than that of the B. cereus vegetative cells, except for burdock. The growth of B. cereus and S. aureus was not observed in bracken at temperatures lower than 13 and $8^{\circ}C$, respectively. The LT models for B. cereus (diarrhea type) in this study were suitable in predicting the growth of B. cereus (emetic type) on burdock and Chwinamul. On the other hand, SGR models for B. cereus (diarrhea type) were suitable for predicting the growth of B. cereus (emetic type) on all preprocessed vegetables. The developed models can be used to predict the risk of B. cereus and S. aureus in preprocessed Namul and root vegetables at the retail markets.

• Proceedings of the KIPE Conference
• /
• 2010.07a
• /
• pp.464-465
• /
• 2010

EMI 감쇄성능의 정확한 예측을 위해서는 EMI 필터에 사용되는 소자에 대한 명확한 공통 및 차동 모드 임피던스 모델 정보가 필요하다. 하지만 기존의 전도성 EMI 감쇄성능 예측 방식은 이러한 모델의 부재로 인해 고주파수에서 예측 값과 실험 결과에 큰 오차가 발생하는 문제점이 있다. 이를 해결하기 위해 본 논문에서는 일반적으로 사용되는 EMI 필터의 소자를 전도성 전파 규제 범위에서 모델링하고 이를 이용하여 공통 및 차동모드 임피던스로 다시 모델링한다. 실험 결과 EMI 감쇄성능을 1MHz 이하의 영역에서만 예측할 수 있었던 기존 방식과 비교해 제안 방식은 10MHz 영역까지 예측할 수 있는 장점이 있다. 최종적으로 임피던스 분석기를 이용한 측정 결과와 모의실험 결과를 제시하여 제안 방식의 타당성 및 유용성을 검증한다.

• Journal of the Korean Geotechnical Society
• /
• v.26 no.1
• /
• pp.55-62
• /
• 2010

Dynamic soil resistances were simulated by modified Ramberg-Osgood model in order to predict penetration rate of sheet pile installed by vibratory pile driver. Various factors which characterize modified Ramberg-Osgood model were determined considering the shapes of dynamic soil resistance curves obtained from field test and standard penetration value (N value) was used as parameter that relates field test results to the suggested model. Penetration rates calculated by analytical model were smaller than those of field test and penetration times were vice versa. Therefore, predicted penetration rate and penetration time by analytical model are more conservative than those of filed test.