• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.206-206
• /
• 2023

지구온난화로 인한 기후변화에 따라 평균강수량과 증발량이 증가하며 강우지역 집중화와 강우강도가 높아질 가능성이 크다. 우리나라의 경우 협소한 국토면적과 높은 인구밀도로 기후변동의 영향이 크기 때문에 한반도에 적합한 유역규모의 수자원 예측과 대응방안을 마련해야 한다. 이를 위한 수자원 관리를 위해서는 유역에서 강수량, 유출량, 증발량 등의 장기적인 자료가 필요하며 경험식, 물리적 강우-유출 모형 등이 사용되었고, 최근들어 연구의 확장성과 비 선형성 등을 고려하기 위해 딥러닝등 인공지능 기술들이 접목되고 있다. 본 연구에서는 ASOS(동해, 태백)와 AWS(삼척, 신기, 도계) 5곳의 관측소에서 2011년~2020년까지의 일 단위 기상관측자료를 수집하고 WAMIS에서 같은 기간의 오십천 하구 일 유출량 자료를 수집 후 5개 관측소를 기준으로Thiessen 면적비를 적용해 기상자료를 구축했으며 Angstrom & Hargreaves 공식으로 잠재증발산량 산정해 3개의 모델에 각각 기상자료(일 강수량, 최고기온, 최대 순간 풍속, 최저기온, 평균풍속, 평균기온), 일 강수량과 잠재증발산량, 일 강수량 - 잠재증발산량을 학습 후 관측 유출량과 비교결과 기상자료(일 강수량, 최고기온, 최대 순간 풍속, 최저기온, 평균풍속, 평균기온)로 학습한 모델성능이 가장 높아 최적 모델로 선정했으며 일, 월, 연 관측유출량 시계열과 비교했다. 또한 같은 학습자료를 사용해 다층 퍼셉트론(Multi Layer Perceptron, MLP) 앙상블 모델을 구축하여 수자원 분야에서의 인공지능 활용성을 평가했다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.420-420
• /
• 2023

하천의 유량 자료는 하천 관리에 필수적인 요소로, 지속적인 유량측정을 위해 국가 유량 측정망을 구성하여 주요 지점을 대상으로 유량 측정을 수행하고 있다. 측정된 유량자료는 일반적으로 수위-유량 관계곡선식을 개발하여 제공되고 있으며, 홍수파와 배수 영향 등으로 인해 수위-유량 관계곡선식에서 발생하는 산포로 인한 신뢰도에 문제가 우려되는 경우에는 실시간의 정확한 유량자료를 제공하기 위해 H-ADCP를 설치하여 지표유속법 기반의 실시간 유량 자료 생산하여 제공하고 있다. 그러나 H-ADCP를 이용한 유량 측정 방법은 장비의 한계로 인해 상대적으로 규모가 작고 수심이 얕은 하천에 적용하기 어려운 문제가 있다. 따라서, 최근에는 자동유량관측소 지점 확대를 위해 비접촉식 유속계를 활용한 자동유량관측소 운영이 점차 고려되고 있다. 이에 따라 비접촉식유속계를 이용한 유량 측정 결과의 검증 및 유지 관리를 위한 소프트웨어가 필요하다. 이에 본 연구에서는 비접촉식유속계 중 전자파를 이용하여 수표면의 표면유속을 측정할 수 있는 장비인 RQ-30의 측정결과를 분석하기 위해 Microsoft Visual Studio(C#) 사용하여 측정결과의 검토 및 자료 관리를 위한 후처리 소프트웨어를 개발하였다. 개발한 소프트웨어는 측정 원시자료를 읽고, 도시하여 측정 결과를 확인할 수 있으며, 머신러닝 기반의 알고리즘을 적용하여 수위 및 유속 시계열 자료에서 발생하는 이상치를 탐색할 수 있도록 개발하였다. 그리고 탐지된 이상치에 대한 보정을 위해 선형보간, LOESS, SuperSmoother를 사용하여 이상치를 보정하여 결과를 도출할 수 있도록 개발하였다. 추후 본 연구를 통해 개발된 프로그램을 활용하여 측정 자료의 유지 관리 효율성을 증대시킬 수 있을 것으로 기대되며, 지속적인 프로그램의 개선을 통해서 실무적으로 활용이 가능할 것으로 판단된다.

• The Korean Journal of Applied Statistics
• /
• v.35 no.2
• /
• pp.229-250
• /
• 2022

Since the publication of Markowitz's (1952) mean-variance portfolio model, research on portfolio optimization has been conducted in many fields. The existing mean-variance portfolio model forms a nonlinear convex problem. Applying Dantzig's linear programming method, it was converted to a linear form, which can effectively reduce the algorithm computation time. In this paper, we proposed a Dantzig perturbation portfolio model that can reduce management costs and transaction costs by constructing a portfolio with stable and small (sparse) assets. The average return and risk were adjusted according to the purpose by applying a perturbation method in which a certain part is invested in the existing benchmark and the rest is invested in the assets proposed as a portfolio optimization model. For a covariance estimation, we proposed a Gaussian kernel weight covariance that considers time-dependent weights by reflecting time-series data characteristics. The performance of the proposed model was evaluated by comparing it with the benchmark portfolio with 5 real data sets. Empirical results show that the proposed portfolios provide higher expected returns or lower risks than the benchmark. Further, sparse and stable asset selection was obtained in the proposed portfolios.

• Journal of Korea Water Resources Association
• /
• v.50 no.7
• /
• pp.489-502
• /
• 2017

The objective of this study is to propose and evaluate the BAYES-ESP, which is a dam inflow prediction method based on Ensemble Streamflow Prediction method (ESP) and Bayesian theory. ABCD rainfall-runoff model was used to predict monthly dam inflow. Monthly meteorological data collected from KMA, MOLIT and K-water and dam inflow data collected from K-water were used for the model calibration and verification. To estimate the performance of ABCD model, ESP and BAYES-ESP method, time series analysis and skill score (SS) during 1986~2015 were used. In time series analysis monthly ESP dam inflow prediction values were nearly similar for every years, particularly less accurate in wet and dry years. The proposed BAYES-ESP improved the performance of ESP, especially in wet year. The SS was used for quantitative analysis of monthly mean of observed dam inflows, predicted values from ESP and BAYES-ESP. The results indicated that the SS values of ESP were relatively high in January, February and March but negative values in the other months. It also showed that the BAYES-ESP improved ESP when the values from ESP and observation have a relatively apparent linear relationship. We concluded that the existing ESP method has a limitation to predict dam inflow in Korea due to the seasonality of precipitation pattern and the proposed BAYES-ESP is meaningful for improving dam inflow prediction accuracy of ESP.

• Journal of Digital Convergence
• /
• v.18 no.8
• /
• pp.169-177
• /
• 2020

The purpose of this study is to forecast the sales of convenience store flagship products with nonlinear characteristics and time series characteristics. According to the results, the sales of 'Ice Cup' began to increase from March, reached the highest value in summer, especially July and August, and then decreased, revealing a seasonal pattern. Cigarettes showed a seasonal pattern of higher sales in summer and lower sales in winter and was predicted to decrease in sales in the future. This study provides an academic implication in that it focused on the top-selling products that affected an increase in financial performance in a specific convenience store, a method that has been hardly adopted by the existing studies.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.4B
• /
• pp.399-412
• /
• 2010

The goal of this research is to apply the neural networks models for the temporal disaggregation of the yearly pan evaporation (PE) data, Republic of Korea. The neural networks models consist of multilayer perceptron neural networks model (MLP-NNM) and generalized regression neural networks model (GRNNM), respectively. And, for the performances evaluation of the neural networks models, they are composed of training and test performances, respectively. The three types of data such as the historic, the generated, and the mixed data are used for the training performance. The only historic data, however, is used for the testing performance. From this research, we evaluate the application of MLP-NNM and GRNNM for the temporal disaggregation of nonlinear time series data. We should, furthermore, construct the credible monthly PE data from the temporal disaggregation of the yearly PE data, and can suggest the available data for the evaluation of irrigation and drainage networks system.

• Journal of the Korean Society for Marine Environment & Energy
• /
• v.12 no.2
• /
• pp.61-67
• /
• 2009

In this paper time series wave data which contain a freak wave is investigated. Various wave characteristics are compared between wave data with a freak wave and without. Among 24 hour wave data measured in the Yura Sea, two adjacent 30 min wave data with and without a freak wave are examined intensively. It is seen that the highest waves do not have the longest wave period. The wave period of the longest period waves is a little longer than the average wave period and much shorter than the significant wave period. Although the sea state is quite high, the Rayleigh distribution fits well to the probability of wave height. The characteristics of the wave spectra do not change much, but the nonlinearity increases for the wave data with a freak wave. The significant wave height without a freak wave is larger than that with a freak wave. Hence, the higher significant wave height does not always increase the probability of the occurrence of the freak waves.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.3
• /
• pp.419-425
• /
• 2010

Kernel machine learning is gaining a lot of popularities in analyzing large or high dimensional nonlinear data. We use this technique to estimate a GARCH model for predicting the conditional volatility of stock market returns. GARCH models are usually estimated using maximum likelihood (ML) procedures, assuming that the data are normally distributed. In this paper, we show that GARCH models can be estimated using kernel machine learning and that kernel machine has a higher predicting ability than ML methods and support vector machine, when estimating volatility of financial time series data with fat tail.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.5
• /
• pp.487-496
• /
• 2000

In this paper, the Multi-FNN(Fuzzy-Neural Networks) model is identified and optimized using HCM(Hard C-Means) clustering method and genetic algorithms. The proposed Multi-FNN is based on Yamakawa's FNN and uses simplified inference as fuzzy inference method and error back propagation algorithm as learning rules. We use a HCM clustering and Genetic Algorithms(GAs) to identify both the structure and the parameters of a Multi-FNN model. Here, HCM clustering method, which is carried out for the process data preprocessing of system modeling, is utilized to determine the structure of Multi-FNN according to the divisions of input-output space using I/O process data. Also, the parameters of Multi-FNN model such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. A aggregate performance index with a weighting factor is used to achieve a sound balance between approximation and generalization abilities of the model. The aggregate performance index stands for an aggregate objective function with a weighting factor to consider a mutual balance and dependency between approximation and predictive abilities. According to the selection and adjustment of a weighting factor of this aggregate abjective function which depends on the number of data and a certain degree of nonlinearity, we show that it is available and effective to design an optimal Multi-FNN model. To evaluate the performance of the proposed model, we use the time series data for gas furnace and the numerical data of nonlinear function.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.4
• /
• pp.579-586
• /
• 2018

Short-term prediction of travel speed has been widely studied using data-driven non-parametric techniques. There is, however, a lack of research on the prediction aimed at urban areas due to their complex dynamics stemming from traffic signals and intersections. The purpose of this study is to develop a hybrid approach combining ensemble empirical mode decomposition (EEMD) and artificial neural network (ANN) for predicting urban travel speed. The EEMD decomposes the time-series data of travel speed into intrinsic mode functions (IMFs) and residue. The decomposed IMFs represent local characteristics of time-scale components and they are predicted using an ANN, respectively. The IMFs can be predicted more accurately than their original travel speed since they mitigate the complexity of the original data such as non-linearity, non-stationarity, and oscillation. The predicted IMFs are summed up to represent the predicted travel speed. To evaluate the proposed method, the travel speed data from the dedicated short range communication (DSRC) in Daegu City are used. Performance evaluations are conducted targeting on the links that are particularly hard to predict. The results show the developed model has the mean absolute error rate of 10.41% in the normal condition and 25.35% in the break down for the 15-min-ahead prediction, respectively, and it outperforms the simple ANN model. The developed model contributes to the provision of the reliable traffic information in urban transportation management systems.