• 품질경영학회지
• /
• 제40권2호
• /
• pp.186-198
• /
• 2012

Since Duncan(1956) first proposed an economic design of $\bar{X}$-control charts, most of the succeeding works on economic design of control charts assumed the exponential failure model like Duncan. Hu(1984), however, assumed a more versatile Weibull failure model to develop an economic design of $\bar{X}$-control charts and Banerjee and Rahim(1988) further improved Hu's design by changing the assumption of fixed-length sampling intervals to variable-length ones. In this article we follow the approach of Banerjee and Rahim(1988) but include a pair of warning limits inside the control limits in order to search for a failure without stopping the process when the sample mean falls between warning and control limits. The computational results indicate that the proposed model gives a lower cost than Banerjee and Rahim's model unless the early failure probability of a Weibull distribution is relatively large. The reduction in cost is shown to become larger as the cost of production loss outweighs the cost of searches for a failure.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2020년도 학술발표회
• /
• pp.122-122
• /
• 2020

Urban flood simulation plays a vital role in national flood early warning, prevention and mitigation. In recent studies on 2-dimensional flood modeling, the integrated run-off inundation model is gaining grounds due to its ability to perform in greater computational efficiency. The adaptive quadtree shallow water numerical technique used in this model implements the adaptive mesh refinement (AMR) in this simulation, a procedure in which the grid resolution is refined automatically following the flood flow. The method discounts the necessity to create a whole domain mesh over a complex catchment area, which is one of the most time-consuming steps in flood simulation. This research applies the dynamic grid refinement method in simulating the recent extreme flood events in Metro Manila, Philippines. The rainfall events utilized were during Typhoon Ketsana 2009, and Southwest monsoon surges in 2012 and 2013. In order to much more visualize the urban flooding that incorporates the flow within buildings and high-elevation areas, Digital Surface Model (DSM) resolution of 5m was used in representing the ground elevation. Results were calibrated through the flood point validation data and compared to the present flood hazard maps used for policy making by the national government agency. The accuracy and efficiency of the method provides a strong front in making it commendable to use for early warning and flood inundation analysis for future similar flood events.

• Earthquakes and Structures
• /
• 제18권2호
• /
• pp.163-170
• /
• 2020

Floor acceleration plays a major role in the seismic design of nonstructural components and equipment supported by structures. Large floor acceleration may cause structural damage to or even collapse of buildings. For precision instruments in high-tech factories, even small floor accelerations can cause considerable damage in this study. Six P-wave parameters, namely the peak measurement of acceleration, peak measurement of velocity, peak measurement of displacement, effective predominant period, integral of squared velocity, and cumulative absolute velocity, were estimated from the first 3 s of a vertical ground acceleration time history. Subsequently, a new predictive algorithm was developed, which utilizes the aforementioned parameters with the floor height and fundamental period of the structure as the new inputs of a support vector regression model. Representative earthquakes, which were recorded by the Structure Strong Earthquake Monitoring System of the Central Weather Bureau in Taiwan from 1992 to 2016, were used to construct the support vector regression model for predicting the peak floor acceleration (PFA) of each floor. The results indicated that the accuracy of the predicted PFA, which was defined as a PFA within a one-level difference from the measured PFA on Taiwan's seismic intensity scale, was 96.96%. The proposed system can be integrated into the existing earthquake early warning system to provide complete protection to life and the economy.

• 한국농림기상학회지
• /
• 제8권1호
• /
• pp.28-35
• /
• 2006

근년에 자주 나타나고 있는 봄철 과원의 서리피해는 관측된 기온이 비슷한 지역일지라도 개화 혹은 발아 단계의 과원에서 집중되고 있어 효율적인 상해 경보시스템 운영을 위해서는 발아기 혹은 만개기의 정확한 예측이 필요하다. 품종별 모수가 알려져 있는 포도 거봉, Campbell Early를 대상으로 생물계절모형을 적용하여 발아기를 추정하고 최저기온 예상치와 함께 늦서리 위험도 추정방법을 제시하였다. 이 방법은 발아 이후에 최저기온이 영하로 내려가면 상해가 발생한다고 가정하는데, 추정값의 오차범위를 고려한 발아일 이후 일 최저기온이 $-1.5^{\circ}C$ 이하로 떨어지면 경보(Warning), ${\pm}1.5^{\circ}C$ 사이면 주의보(Watch)를 발령한다. 이 방법을 2004년과 2005년 4월 경기 안성의 포도원에 적용하여 결과의 신뢰도를 확인하였다. 같은 방법으로 1971-2000평년의 기후조건에서 예상되는 안성지역의 포도 늦서리피해 위험지역을 30 m의 고해상도 전자기후도로 표현하였다. 안성시 전역을 30 m 격자점으로 표현하면 총 608,585개로 구성되는데, 평년의 포도 상해위험지역 판정결과 거봉은 1,059지역이, Campbell Early는 2,788지역이 주의지대로 예상된다.

• 한국산림과학회지
• /
• 제108권2호
• /
• pp.208-215
• /
• 2019

이 연구는 붕괴모의실험을 통하여 체적함수비센서와 텐시오미터의 반응성을 분석하고, 산사태 조기경보용으로의 활용성을 검토하기 위해 수행되었다. 산림토양과 사질토의 배합비율을 조정한 3개의 토양조건에서 120 mm/h의 인공강우를 적용하여 얕은 깊이에서 빠르게 진행되는 붕괴형태를 실험적으로 모의하고, 그 과정에서의 두 센서의 반응을 분석하였다. 그 결과, 모든 실험조건에서 체적함수비센서 및 텐시오미터의 계측값은 각각 30~37%, -3~-5 kPa으로 수렴된 이후에 붕괴가 발생하였다. 실험결과를 토대로 토층 최하부에 설치된 체적함수비센서의 계측값을 활용하여 조기경보 발생시점의 범위를 논의하였으나, 이를 일반화하여 명확한 시점으로 규정할 수는 없었다. 두 센서를 실용적인 차원에서 산사태 조기경보용으로 활용하기 위해서는 다양한 조건에서의 추가적인 실험 및 검증이 필요할 것으로 생각되었다.

• 산업공학
• /
• 제18권3호
• /
• pp.268-276
• /
• 2005

In recent years, the money Laundering crimes are increasing by means of foreign exchange transactions. Our study proposes four scoring models to provide early warning of the laundering in foreign exchange transactions for both inward and outward remittances: logistic regression model, decision tree, neural network, and ensemble model which combines the three models. In terms of accuracy of test data, decision tree model is selected for the inward remittance and an ensemble model for the outward remittance. From our study results, the accumulated number of transaction turns out to be the most important predictor variable. The proposed scoring models deal with the transaction level and is expected to help the bank teller to detect the laundering related transactions in the early stage.

• 지질공학
• /
• 제18권2호
• /
• pp.191-196
• /
• 2008

이 연구는 집중호우시 산사태의 발생가능성과 발생시간을 사전에 예측하기 위한 노력의 일환으로 기존에 개발된 RTI 경보모델을 우리나라에 적용 분석한 사례이다. RTI(Rainfall Triggering Index)는 강우강도(I) 유효 누적강우량($R_t$)의 곱으로 정의되는 것으로서, 강우기간 동안 특정 시간(t)에서 산사태가 발생할 가능성을 평가하는데 사용된다. RTI의 상부임계값($RTI_{UC}$)과 하부임계값($RTI_{LC}$) 과거 산사태 발생시 강우자료 분석을 통해 각 지역별로 설정할 수 있으며, 강우강도가 상부임계값을 초과할 때 실제 산사태가 발생하는 것으로 이해할 수 있다. 이러한 분석은 궁극적으로 향후 집중호우가 내릴 경우 특정지역의 산사태 발생가능성은 물론 산사태 발생시기를 예상할 수 있으며, 이를 토대로 사전에 산사태 발생경보를 발령하는데 중요한 근거로 활용될 수 있다. 이와 같은 이론을 우리나라에 적용하기 위해 2006년 7월 13일부터 7월 19일까지 강원도 인제군 일대에 내린 강우자료와 산사태 발생과의 관계를 분석한 결과, 실제 산사태가 발생한 7월16일 오전 11시경을 기준으로 23시간, 11시간, 9시간 전에 강우강도가 RTI의 상부임계값을 초과하였다. 이를 토대로 이와 같은 세 차례에 걸친 산사태 경보의 발령이 필요하였음을 알 수 있었다.

• 한국컴퓨터정보학회논문지
• /
• 제28권12호
• /
• pp.67-77
• /
• 2023

본 연구는 스마트팜 환경에서 진행된 혁신적인 연구로, 딥러닝을 기반으로 한 질병 및 해충 탐지 모델을 개발하고, 이를 지능형 사물인터넷(IoT) 플랫폼에 적용하여 디지털 농업 환경 구현의 새로운 가능성을 탐색하였다. 연구의 핵심은 Pseudo-Labeling, RegNet, EfficientNet 등 최신 ImageNet 모델과 전처리 방식을 통합하여, 복잡한 농업 환경에서 다양한 질병과 해충을 높은 정확도로 탐지하는 것이었다. 이를 위해 앙상블 학습 기법을 적용하여 모델의 정확도와 안정성을 극대화했으며, 평균 정밀도(mAP), 정밀도, 재현율, 정확도, 박스 손실 등의 다양한 성능 지표를 통해 모델을 평가하였다. 또한, SHAP 프레임워크를 활용하여 모델의 예측 기준에 대한 깊은 이해를 도모하였고, 이를 통해 모델의 결정 과정을 보다 투명하게 만들었다. 이러한 분석은 모델이 어떻게 다양한 변수들을 고려하여 질병 및 해충을 탐지하는지에 대한 중요한 통찰력을 제공하였다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2023년도 학술발표회
• /
• pp.95-95
• /
• 2023

Common hydrological problems of developing countries include poor data management, insufficient measuring devices and ungauged watersheds, leading to small or unreliable data availability. This has greatly affected the adoption of artificial intelligence techniques for flood risk mitigation and damage control in several developing countries. While climate datasets have recorded resounding applications, but they exhibit more uncertainties than ground-based measurements. To encourage AI adoption in developing countries with small ground-based dataset, we propose data augmentation for regression tasks and compare performance evaluation of different AI models with and without data augmentation. More focus is placed on simple models that offer lesser computational cost and higher accuracy than deeper models that train longer and consume computer resources, which may be insufficient in developing countries. To implement this approach, we modelled and predicted streamflow data of the Asa River Watershed located in Ilorin, Kwara State Nigeria. Results revealed that adequate hyperparameter tuning and proper model selection improve streamflow prediction on small water dataset. This approach can be implemented in data-scarce regions to ensure timely flood intervention and early warning systems are adopted in developing countries.

• 상하수도학회지
• /
• 제34권1호
• /
• pp.35-43
• /
• 2020

Turbidity has various effects on the water quality and ecosystem of a river. High turbidity during floods increases the operation cost of a drinking water supply system. Thus, the management of turbidity is essential for providing safe water to the public. There have been various efforts to estimate turbidity in river systems for proper management and early warning of high turbidity in the water supply process. Advanced data analysis technology using machine learning has been increasingly used in water quality management processes. Artificial neural networks(ANNs) is one of the first algorithms applied, where the overfitting of a model to observed data and vanishing gradient in the backpropagation process limit the wide application of ANNs in practice. In recent years, deep learning, which overcomes the limitations of ANNs, has been applied in water quality management. LSTM(Long-Short Term Memory) is one of novel deep learning algorithms that is widely used in the analysis of time series data. In this study, LSTM is used for the prediction of high turbidity(>30 NTU) in a river from the relationship of turbidity to discharge, which enables early warning of high turbidity in a drinking water supply system. The model showed 0.98, 0.99, 0.98 and 0.99 for precision, recall, F1-score and accuracy respectively, for the prediction of high turbidity in a river with 2 hour frequency data. The sensitivity of the model to the observation intervals of data is also compared with time periods of 2 hour, 8 hour, 1 day and 2 days. The model shows higher precision with shorter observation intervals, which underscores the importance of collecting high frequency data for better management of water resources in the future.