• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.131-131
• /
• 2021

Deep learning models, especially those based on long short-term memory (LSTM), have presented their superiority in addressing time series data issues recently. This study aims to comprehensively evaluate the performance of deep learning models that belong to the supervised learning category in streamflow prediction. Therefore, six deep learning models-standard LSTM, standard gated recurrent unit (GRU), stacked LSTM, bidirectional LSTM (BiLSTM), feed-forward neural network (FFNN), and convolutional neural network (CNN) models-were of interest in this study. The Red River system, one of the largest river basins in Vietnam, was adopted as a case study. In addition, deep learning models were designed to forecast flowrate for one- and two-day ahead at Son Tay hydrological station on the Red River using a series of observed flowrate data at seven hydrological stations on three major river branches of the Red River system-Thao River, Da River, and Lo River-as the input data for training, validation, and testing. The comparison results have indicated that the four LSTM-based models exhibit significantly better performance and maintain stability than the FFNN and CNN models. Moreover, LSTM-based models may reach impressive predictions even in the presence of upstream reservoirs and dams. In the case of the stacked LSTM and BiLSTM models, the complexity of these models is not accompanied by performance improvement because their respective performance is not higher than the two standard models (LSTM and GRU). As a result, we realized that in the context of hydrological forecasting problems, simple architectural models such as LSTM and GRU (with one hidden layer) are sufficient to produce highly reliable forecasts while minimizing computation time because of the sequential data nature.

• 한국신재생에너지학회:학술대회논문집
• /
• 2009.06a
• /
• pp.69-72
• /
• 2009

In this paper, the Stand-alone underground water pumping system was developed that is consist of Submersible Pump (AC type), Photovoltaic Array and Power converter by the application of solar energy. And also wish to introduce system that is possible to supply of drinking water or water for agriculture using solar energy at desertification area or a Off-grid area, interior etc. and operation test results. This system can use in deep tube well of 200m range with common Submersible Pump and maximized to the quantity of pumping through M.P.P.T control. Also system availability raised through apply various driving mode.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2007.09a
• /
• pp.611-622
• /
• 2007

In the design and construction of the coastal/offshore structures, it is very important to evaluate the geotechnical characteristics of marine soils, which support the structures. Although the offshore site investigation is much more difficult than onshore, safe and precise jobs have not been accomplished in Korea because of the insufficiency of the test equipments especially for the site with deep water depth. The main objective of this study is to develop a new type of marine cone penetration testing(CPT) system, which can be utilized to even deep sea and high depth of soil layer. The system is one of seabed types and employs the conventional cone, which shows more reliable results than miniature cone. The most important parts of the marine CPT including continuous rod system, cone penetration system with wheel drive, automatic cone rod assembly/dissembly system etc., were designed and manufactured. Some tests to verify the developed marine CPT system were performed at both onshore and offshore sites as well as mechanical test in laboratory. The test results show the consistent and promising performance of the new equipment, and thereafter the system would be applicable to various sites with practical/economical advantages.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.10a
• /
• pp.663-664
• /
• 2016

Temperature difference power generation using sea water is a method repeatedly closed liquefaction and gasification by using the ammonia (refrigerant) of the deep sea water and surface water with a temperature difference between turning the turbine. The larger the temperature difference between the nature of the temperature characteristic energy generation development, the better. This is the story that the surface waters of the deep-water temperature difference is large. But the winter is not large temperature difference between surface water and deep water has lowered energy efficiency. And desalination technologies accounted for 97% of the earth, but we can not eat the technology to convert sea water into fresh water, fresh water produced by the desalination technology that is available for various industries such as irrigation, drinking water in the vessel.In this paper, LNG transport vessels, based on the LNG transport ship to the temperature difference power generation using cold energy of thermal energy and LNG marine diesel engines, which use the existing order to improve the temperature of the surface waters of the season that is the current problem we propose that a complex development of desalination and desalination of seawater freezing research into hybrid research and utilizing the cold energy of the engine.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.38 no.1
• /
• pp.65-69
• /
• 2005

We have developed a new method of manufacturing salts by horizontal spray drying technique, using the concentrated deep ocean water after desalination processes. We studied the chemical characteristics of the spray-dried salts. Sodium content in the spray salts is $28.4\%$, which is $10-30\%$ lower than that of bay salts, bamboo salts and boiling salts $(32-38.2\%)$. However, the contents of magnesium, potassium and calcium of the spray salts are 2.5 times, 3 times and 4.5 times higher relative to those of bay salts, respectively. On the one hand, sulfur content in spray salts is 14 times lower than those of bay salts, which is caused by their volatilization during spray and vaporization of the concentrated seawater. Enrichment factors of Mg (0.8), K (0.9) and Ca (1.0) in the spray salts are relatively higher than those in bay salt (0.2-0.3), bamboo salt (0.15-0.4) and boiled salt (0.4-0.7), respectively. On the contrary, enrichment factor of sulfur in spray salts is observed to be 0.07, which is considerably lower than those in other salts (0.3-0.7). It means that the minerals like Mg, K and Ca can be well conserved from seawater to salts through spray drying techniques, while volatile elements like sulfur, lead, mercury and organic compounds can be easily removed from seawater via spray and heating processes.

• Journal of the Korean Society of Industry Convergence
• /
• v.24 no.3
• /
• pp.253-260
• /
• 2021

The implementation of subsea separation and liquid boosting is becoming a common development scheme for the exploration of deep water fields. Subsea separation is an attractive and economic solution to develop deep offshore fields producing fluid without hydrate or wax. A subsea separator can avoid or simplifying costly surface platforms of floating vessels, as well as being an efficient tool to enhance hydrocarbon production. Subsea separation system should be reliable to ensure successful operation in a wide range of 3-phase flow regime. In this study, multiphase flow characteristics inside in-line type subsea separation system are investigated for the design of subsea separation system.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.39 no.1
• /
• pp.64-74
• /
• 1997

A CELTHYM(CEll-based Long-term HYdrologic Model), a pre-processor and a post-processor that can be integrated with geographic information system(GIS) were developed to predict the stream flow from the small agricultural watershed on the daily basis. The CELTHYM calculates the direct runoff from a grid using SCS curve number method and then sum up all of cells with respect to a sub-catchment area belonged to a stream grid and integrated to an outlet. Base flow of a watershed outlet was computed by integrating of the base flow of each stream grid that was averaged the sub-catchment deep-percolation and calculated with the release rate. Two kind of water budget equation were used to compute the water balance in a grid that was classified into not paddy field and paddy field. One of the two equation is a soil water balance equation to account the soil moisture of the upland, forest and excluding paddy field grid. The other is a paddy water balance equation for the paddy field, calculating the ponding depth, the effective rainfall, the deep percolation and the evapotranspiration.

• The KSFM Journal of Fluid Machinery
• /
• v.14 no.5
• /
• pp.69-74
• /
• 2011

This paper deals with the detail rotordynamic analysis for the vertical rotor system as development of vertical sea water lift pump for FPSO deep well. In a vertical rotor system, since linearized stiffness and damping coefficients of fluid film bearing are no longer be valid, hence the transient response analysis considering a fluid film force for every journal position in the bearing needs to be required. In this study, the transient response analysis of the proposed vertical pump rotor system was carried out in dry-run and wet-run conditions, respectively. The results show that orbital vibration responses of the rotor system remain stable at rated speed and thereby operating reliability of the vertical rotor system is confirmed. To overcome complexity of calculation pr ocedure and time consuming calculation of transient analysis, the calculating technique of steady-state response analysis is also proposed. The results of steady-state response obtained by applying the proposed technique to the rotor system are good agreement with the reference results, that is, transient responses.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.24 no.5
• /
• pp.343-351
• /
• 2012

In this study, shallow-water design waves are calculated for the return period of 10, 20, 30, and 50 years, based on the extreme value analysis of the wave measurement data at Gangneung beach. These values are compared with the results of SWAN simulation with the boundary condition of the deep-water design waves of the corresponding return periods at the Gangneung sea area provided by the Fisheries Agency (FA, 1988) and Korea Ocean Research & Development Institute (KORDI, 2005). It is found that the shallow-water wave heights at Gangneung beach calculated by the deep-water design waves were significantly less than the observation data. As the return period becomes higher, the significant wave heights obtained by the extreme value analysis becomes higher than those computed by SWAN with the deep-water design waves of the corresponding return periods. KORDI computed the hindcast wave data from January 2004 to August 2008 by WAM with a finer-grid mesh system than those of previous studies. Comparisons of the wave hindcast results with the wave observation show that the reproducibility of the winter-season storm wave was considerably improved compared to the hindcast data from 1979 to 2003. Hereafter, it is necessary to carry out hindcast wave data for the years before 2004 using WAM with the finer-grid mesh system and to supplement the deep-water design wave.

• Journal of Korea Water Resources Association
• /
• v.54 no.3
• /
• pp.157-166
• /
• 2021

Forecasting dam inflow based on high reliability is required for efficient dam operation. In this study, deep learning technique, which is one of the data-driven methods and has been used in many fields of research, was manipulated to predict the dam inflow. The Long Short-Term Memory deep learning with Sequence-to-Sequence model (LSTM-s2s), which provides high performance in predicting time-series data, was applied for forecasting inflow of Soyang River dam. Various statistical metrics or evaluation indicators, including correlation coefficient (CC), Nash-Sutcliffe efficiency coefficient (NSE), percent bias (PBIAS), and error in peak value (PE), were used to evaluate the predictive performance of the model. The result of this study presented that the LSTM-s2s model showed high accuracy in the prediction of dam inflow and also provided good performance for runoff event based runoff prediction. It was found that the deep learning based approach could be used for efficient dam operation for water resource management during wet and dry seasons.