• Korean Journal of Remote Sensing
• /
• v.32 no.2
• /
• pp.155-169
• /
• 2016

In this study, we developed a multi-sensor blending short-term rainfall forecasting technique using radar and satellite data during extreme rainfall occurrences in Busan and Gyeongnam region in August 2014. The Tropical Z-R relationship ($Z=32R^{1.65}$) has applied as a optimal radar Z-R relation, which is confirmed that the accuracy is improved during 20mm/h heavy rainfall. In addition, the multi-sensor blending technique has applied using radar and COMS (Communication, Ocean and Meteorological Satellite) data for quantitative precipitation estimation. The very-short-term rainfall forecasting performance was improved in 60 mm/h or more of the strong heavy rainfall events by multi-sensor blending. AWS (Automatic Weather System) and MAPLE data were used for verification of rainfall prediction accuracy. The results have ensured about 50% or more in accuracy of heavy rainfall prediction for 1-hour before rainfall prediction, which are correlations of 10-minute lead time have 0.80 to 0.53, and root mean square errors have 3.99 mm/h to 6.43 mm/h. Through this study, utilizing of multi-sensor blending techniques using radar and satellite data are possible to provide that would be more reliable very-short-term rainfall forecasting data. Further we need ongoing case studies and prediction and estimation of quantitative precipitation by multi-sensor blending is required as well as improving the satellite rainfall estimation algorithm.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.64 no.1
• /
• pp.128-135
• /
• 2015

In this study, we develop the very short-term precipitation forecasting model as well as classifier based on polynomial radial basis function neural networks by using AWS(Automatic Weather Station) and KLAPS(Korea Local Analysis and Prediction System) meteorological data. The polynomial-based radial basis function neural networks is designed to realize precipitation forecasting model as well as classifier. The structure of the proposed RBFNNs consists of three modules such as condition, conclusion, and inference phase. The input space of the condition phase is divided by using Fuzzy C-means(FCM) and the local area of the conclusion phase is represented as four types of polynomial functions. The coefficients of connection weights are estimated by weighted least square estimation(WLSE) for modeling as well as least square estimation(LSE) method for classifier. The final output of the inference phase is obtained through fuzzy inference method. The essential parameters of the proposed model and classifier such ad input variable, polynomial order type, the number of rules, and fuzzification coefficient are optimized by means of Particle Swarm Optimization(PSO) and Differential Evolution(DE). The performance of the proposed precipitation forecasting system is evaluated by using KLAPS meteorological data.

• Journal of Korea Society of Industrial Information Systems
• /
• v.25 no.2
• /
• pp.57-72
• /
• 2020

The price of iron ore has continued to fluctuate with high demand and supply from many countries and companies. In this business environment, forecasting the price of iron ore has become important. This study developed the machine learning model forecasting the price of iron ore a one month after the trading events. The forecasting model used distributed lag model and deep learning models such as MLP (Multi-layer perceptron), RNN (Recurrent neural network) and LSTM (Long short-term memory). According to the results of comparing individual models through metrics, LSTM showed the lowest predictive error. Also, as a result of comparing the models using the ensemble technique, the distributed lag and LSTM ensemble model showed the lowest prediction.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.1
• /
• pp.61-74
• /
• 2019

In this study, the SPI (Standardized Precipitation Index) of meteorological drought and SDI (Streamflow Drought Index) of hydrological drought for 1, 3, 6, 9, and 12 months duration were estimated to analyse the characteristics of drought using rainfall and dam inflow data for Chungju dam ($6,661.8km^2$) with 31 years (1986-2016) and Boryeong dam ($163.6km^2$) watershed with 19 years (1998-2016) respectively. Using the estimated SPI and SDI, the drought forecasting was conducted using seasonal autoregressive integrated moving average (SARIMA) model for the 5 durations. For 2016 drought, the SARIMA had a good results for 3 and 6 months. For the 3 months SARIMA forecasting of SPI and SDI, the correlation coefficient of SPI3, SPI6, SPI12, SDI1, and SDI6 at Chungju Dam showed 0.960, 0.990, 0.999, 0.868, and 0.846, respectively. Also, for same duration forecasting of SPI and SDI at Boryeong Dam, the correlation coefficient of SPI3, SPI6, SDI3, SDI6, and SDI12 showed 0.999, 0.994, 0.999, 0.880, and 0.992, respectively. The SARIMA model showed the possibility to provide the future short-term SPI meteorological drought and the resulting SDI hydrological drought.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.47 no.2
• /
• pp.155-167
• /
• 2024

Distribution and logistics industries contribute some of the biggest GDP(gross domestic product) in South Korea and the number of related companies are quarter of the total number of industries in the country. The number of retail tech companies are quickly increased due to the acceleration of the online and untact shopping trend. Furthermore, major distribution and logistics companies try to achieve integrated data management with the fulfillment process. In contrast, small and medium distribution companies still lack of the capacity and ability to develop digital innovation and smartization. Therefore, in this paper, a deep learning-based demand forecasting & recommendation model is proposed to improve business competitiveness. The proposed model is developed based on real sales transaction data to predict future demand for each product. The proposed model consists of six deep learning models, which are MLP(multi-layers perception), CNN(convolution neural network), RNN(recurrent neural network), LSTM(long short term memory), Conv1D-BiLSTM(convolution-long short term memory) for demand forecasting and collaborative filtering for the recommendation. Each model provides the best prediction result for each product and recommendation model can recommend best sales product among companies own sales list as well as competitor's item list. The proposed demand forecasting model is expected to improve the competitiveness of the small and medium-sized distribution and logistics industry.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.59 no.11
• /
• pp.1956-1963
• /
• 2010

Studies on the development of various energy management programs and real-time bidirectional information infrastructures have been actively conducted to promote the reduction of power demands and CO2 emissions effectively. In the conventional energy management programs, the demand response program that can transition or transfer the power use spontaneously for power prices and other signals has been largely used throughout the inside and outside of the country. For measuring the effect of such demand response program, it is necessary to exactly estimate short-term loads. In this study, the power consumption patterns in both individual and group consumers were analyzed to estimate the exact short-term loads, and the relationship between the actual power consumption and seasonal factors was also analyzed.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.07a
• /
• pp.617-618
• /
• 2020

우리나라의 지리적인 여건상 대륙과 연결되지 않기 때문에 해상운송에 절대적으로 의존하고 있다. 해상운송에 있어 항만시설의 확보가 필요하며 대외무역의존도가 높은 우리나라의 경우 더욱 중요한 역할을 한다. 항만시설은 장기적인 항만수요예측을 통해 대규모 인프라투자를 결정하며 단기적인 예측은 항만운영의 효율성을 개선하고 항만의 경쟁력을 제고하는데 기여하므로 예측의 정확성을 높이기 위해 많은 노력이 필요하다. 본 논문에서는 딥러닝 모델 중에 하나인 LSTM(Long Short Term Memory)을 적용하여 우리나라 주요항만의 컨테이너 물동량 단기예측을 수행하여 선행연구들에서 주류를 이뤘던 ARIMA류의 시계열모델과 비교하여 예측성능을 평가할 것이다. 본 논문은 학문적으로 항만수요예측에 관한 새로운 예측모델을 제시하였다는 측면에서 의미가 있으며 실무적으로 항만수요예측에 대한 정확성을 개선하여 항만투자의사결정에 과학적인 근거로서 활용이 가능할 것으로 기대된다.

• Proceedings of the KIEE Conference
• /
• 2005.11b
• /
• pp.148-150
• /
• 2005

Since the SMP(System Marginal Price) is a vital factor to the market participants who intend to maximize the their profit and to the ISO(Independent System Operator) who wish to operate the electricity market in a stable sense, the short-term marginal price forecasting should be performed correctly. This paper presents a methodology of a day-ahead SMP forecasting using ARIMA(Autoregressive Integrated Moving Average) based on the Time Series. And also we suggested a correction algorithm to minimize the forecasting error in order to improve efficiency and accuracy of the SMP forecasting. To show the efficiency and effectiveness of the proposed method, the numerical studies have been performed using Historical data of SMP in 2004 published by KPX(Korea Power Exchange).

• IE interfaces
• /
• v.18 no.3
• /
• pp.333-342
• /
• 2005

In this paper, we propose an efficient forecasting methodology of the mid and long-term frequency demand in Korea. The methodology consists of the following three steps: classification of basic service group, calculation of effective traffic, and frequency forecasting. Based on the previous studies, we classify the services into wide area mobile, short range radio, fixed wireless access and digital video broadcasting in the step of the classification of basic service group. For the calculation of effective traffic, we use the measures of erlang and bps. The step of the calculation of effective traffic classifies the user and basic application, and evaluates the effective traffic. Finally, in the step of frequency forecasting, different methodology will be proposed for each service group and its applications are presented.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.55 no.11
• /
• pp.485-491
• /
• 2006

In this paper, we study the problem of model selection for Support Vector Machine(SVM) predictor for short-term load forecasting. The model selection amounts to tuning SVM parameters, such as the cost coefficient C and kernel parameters and so on, in order to maximize the prediction performance of SVM. We propose that Cross-Validation method can be used as a model selection algorithm for SVM-based load forecasting technique. Through the various experiments on several data sets, we found that the difference between the prediction error of SVM using Cross-Validation and that of ideal SVM is less than 5%. This shows that SVM parameters for load forecasting can be efficiently tuned by using Cross-Validation.