• Journal of Korea Water Resources Association
• /
• v.54 no.10
• /
• pp.795-805
• /
• 2021

In this study, after developing an LSTM-based deep learning model for estimating daily runoff in the Soyang River Dam basin, the accuracy of the model for various combinations of model structure and input data was investigated. A model was built based on the database consisting of average daily precipitation, average daily temperature, average daily wind speed (input up to here), and daily average flow rate (output) during the first 12 years (1997.1.1-2008.12.31). The Nash-Sutcliffe Model Efficiency Coefficient (NSE) and RMSE were examined for validation using the flow discharge data of the later 12 years (2009.1.1-2020.12.31). The combination that showed the highest accuracy was the case in which all possible input data (12 years of daily precipitation, weather temperature, wind speed) were used on the LSTM model structure with 64 hidden units. The NSE and RMSE of the verification period were 0.862 and 76.8 m³/s, respectively. When the number of hidden units of LSTM exceeds 500, the performance degradation of the model due to overfitting begins to appear, and when the number of hidden units exceeds 1000, the overfitting problem becomes prominent. A model with very high performance (NSE=0.8~0.84) could be obtained when only 12 years of daily precipitation was used for model training. A model with reasonably high performance (NSE=0.63-0.85) when only one year of input data was used for model training. In particular, an accurate model (NSE=0.85) could be obtained if the one year of training data contains a wide magnitude of flow events such as extreme flow and droughts as well as normal events. If the training data includes both the normal and extreme flow rates, input data that is longer than 5 years did not significantly improve the model performance.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.51 no.3
• /
• pp.70-82
• /
• 2023

The recent surge of IT and data acquisition is shifting the paradigm in all aspects of life, and these advances are also affecting academic fields. Research topics and methods are being improved through academic exchange and connections. In particular, data-based research methods are employed in various academic fields, including landscape architecture, where continuous research is needed. Therefore, this study aims to investigate the possibility of developing a landscape preference evaluation and prediction model using machine learning, a branch of Artificial Intelligence, reflecting the current situation. To achieve the goal of this study, machine learning techniques were applied to the landscaping field to build a landscape preference evaluation and prediction model to verify the simulation accuracy of the model. For this, wind power facility landscape images, recently attracting attention as a renewable energy source, were selected as the research objects. For analysis, images of the wind power facility landscapes were collected using web crawling techniques, and an analysis dataset was built. Orange version 3.33, a program from the University of Ljubljana was used for machine learning analysis to derive a prediction model with excellent performance. IA model that integrates the evaluation criteria of machine learning and a separate model structure for the evaluation criteria were used to generate a model using kNN, SVM, Random Forest, Logistic Regression, and Neural Network algorithms suitable for machine learning classification models. The performance evaluation of the generated models was conducted to derive the most suitable prediction model. The prediction model derived in this study separately evaluates three evaluation criteria, including classification by type of landscape, classification by distance between landscape and target, and classification by preference, and then synthesizes and predicts results. As a result of the study, a prediction model with a high accuracy of 0.986 for the evaluation criterion according to the type of landscape, 0.973 for the evaluation criterion according to the distance, and 0.952 for the evaluation criterion according to the preference was developed, and it can be seen that the verification process through the evaluation of data prediction results exceeds the required performance value of the model. As an experimental attempt to investigate the possibility of developing a prediction model using machine learning in landscape-related research, this study was able to confirm the possibility of creating a high-performance prediction model by building a data set through the collection and refinement of image data and subsequently utilizing it in landscape-related research fields. Based on the results, implications, and limitations of this study, it is believed that it is possible to develop various types of landscape prediction models, including wind power facility natural, and cultural landscapes. Machine learning techniques can be more useful and valuable in the field of landscape architecture by exploring and applying research methods appropriate to the topic, reducing the time of data classification through the study of a model that classifies images according to landscape types or analyzing the importance of landscape planning factors through the analysis of landscape prediction factors using machine learning.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• 2008.11a
• /
• pp.93-98
• /
• 2008

Haeundae Doosan We've The Zenith project is adjacent to Suyoung-bay, now it is in the process of excavation and foundation work. The main use of the tower is residence which height is 300m and 80 floor, the highest residential reinforced concrete building through the Orient. It is comprised of 3 high- rised buildings and 1 low-rised building, the basement is 230m wide and 200m length sized mass structure. The lateral resistance system is acted effectively against the lateral load and satisfactorily against the wind vibration by the 4 direction extension of the center core wall($700{\sim}800mm$ thickness) and reinforced concrete column set around the slab. Flat-plate slab system(250mm thickness) is adjusted for the slab system and it enables effective work process and shortening the working term by minimizing the ceiling height and not needing to install perimeter beam and drop panel. The strength and serviceability of the structure is able to be monitored and estimated constantly through the health monitoring system during the construction and after the construction.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.13 no.2
• /
• pp.56-68
• /
• 2011

Wet canopy evaporation ($E_{WC}$) has been recognized as a significant component of total evapotranspiration, especially in forests and therefore it is critical to accurately assess $E_{WC}$ to understand forest hydrological cycle. In this review, I focused on the measurement methods and evaluating the magnitudes of $E_{WC}$ at diverse forest types (e.g., deciduous, coniferous, mixed, and rain forests). I also present the general issues to be considered for $E_{WC}$ measurements. The commonly used measurement methods for $E_{WC}$ include the water balance, energy balance, and the Penman-Monteith (PM) methods. The magnitudes of $E_{WC}$ ranged from 5 to 54% of precipitation based on the literature review, showing a large variation even for a similar forest type possibly related to canopy structure, rainfall intensity, and other meteorological conditions. Therefore, it is difficult to draw a general conclusion on the contribution of $E_{WC}$ to evapotranspiration from a particular forest type. Errors can arise from the measurements of precipitation (due to varying wind effect) and throughfall (due to spatial variability caused by canopy structure) for water balance method, the measurements of sensible heat flux and heat storage for energy balance method, and the estimation of aerodynamic conductance and unaccounted sensible heat advection for the PM method. For a reliable estimation of $E_{WC}$, the combination of ecohydrological and micrometeorological methods is recommended.

• Journal of Bio-Environment Control
• /
• v.24 no.3
• /
• pp.252-257
• /
• 2015

Fruit cracking and vine leaf spot of grapes tend to occur when the plants were directly exposed to rain under outdoor culture. Rain shelter facility can be an alternative method to prevent the cracking and disease of grape, but it also has some limitations in practical usages. We designed rain shelter facility which can completely shut out the rain and ventilate naturally, and it was upgraded to meet the standards of disaster prevention against snow and wind load. The newly developed rain shelter has two-story roof structure, and the $2^{nd}$ floor roof was equipped over $1^{st}$ floor roof at a distance of 40cm. For natural ventilation and water proof, the upper roof protruded about 50cm from the ridge of a $1^{st}$ floor roof. The various tests were carried to examine such as grape quality, brown spot and fruit cracking of Campbell Early under the conventional and the newly developed rain shelter facility which was built about $100{\ss}{\check{S}}$. In comparison of temperature between the conventional and the newly developed rain shelter facility when outside temperature was more than $34^{\circ}C$, the inside temperature was recorded as $40.7^{\circ}C$ and $37.4^{\circ}C$, respectively. There was no significant difference between the two facilities when outside was below $32^{\circ}C$ The quality such as soluble solids and marketable fruit was increased, and fruit cracking of grapes and vine leaf spot also drastically diminished in the newly developed rain shelter.

• Journal of Adhesion and Interface
• /
• v.21 no.2
• /
• pp.65-70
• /
• 2020

Epoxy adhesive was mainly used to combine different composite materials. Epoxy adhesive was a typical thermosetting resin that can be bonded by changing from a linear structure to a three-dimensional network structure by curing reaction of epoxy and hardener. The curing conditions of epoxy adhesive were different with different types of hardener such as mixing ratio, curing time, and temperature. These curing conditions affected to the adhesive property of epoxy adhesive. In industry, it was difficult to proceed the applying epoxy adhesive and combining two parts immediately. The adhesive property decreased by humidity and pre-curing of epoxy adhesive in waiting time between two processes. In this paper, the glass fiber reinforced composite (GFRC) was combined with epoxy adhesive and adhesion property between epoxy adhesive and GFRCs was evaluated using single lap shear test. The different waiting times and humidity conditions were applied to epoxy adhesive in room temperature and adhesive property decreased as the waiting time increased. In small amount of humidity, the adhesive property increased because a small amount of moisture in the surroundings accelerated the curing reaction. In certain amount of humidity, however, the adhesion property decreased.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.7
• /
• pp.135-142
• /
• 2015

Human-powered self-generating devices have been attractive with its operation characteristic independent from outer environment such as weather condition and wind speed. However, conventional self-generators have low electric power output due to their weakly-coupled electromagnetic structure. More importantly, rotary crank motion which is usually adopted by conventional self-generator to generate electricity requires specific skeletal muscles to maintain large torque circular motion and consequently, causes fatigue on those muscles before it can generate enough amount of electricity for any practical application. Without improvement in electric power output and usability, the human-powered self-generator could not be used in everyday life. This study aims to develop a human-powered self-generator which realized a strong electromagnetic coupling in a closed-loop tubular structure (hula-hoop shape) for easy and steady long-term driving as well as larger electric output. The performance and usability of the developed human-powered generator is verified through experimental comparison with a commercial one. Additionally, human workload which is a key element of a human-powered generator but not often considered elsewhere, is estimated based on metabolic energy expenditure measured respiratory gas analyzer. Further study will focus on output and portability enhancement, which can contribute to the continuous power supply of mobile equipments.

• Journal of the Korea Concrete Institute
• /
• v.24 no.5
• /
• pp.553-558
• /
• 2012

Airborne chlorides are transported to inland by sea wind to be attached to seashore concrete structure surface then penetrated into concrete structure members. Since the surface attached chloride amount are dependent on the amount of airborne chlorides, the prediction of distribution of airborne chlorides is important information in preventing chloride corrosion problems in seashore concrete structures. The prediction of surface chloride amount from airborne chlorides environment is extremely difficult than concrete directly in contact with seawater. In addition, their penetrating tendency is different from that of concrete immersed in seawater. In this study, properties of surface and penetrated chlorides under airborne chlorides environment are investigated. Concrete specimens were manufactured and exposed to marine environment for 3 years. The specimens were analyzed at the time durations of 1, 2, and 3 years to check surface chloride amount to penetrated chloride depth. The results revealed that there were certain differences according to surface roughness of concrete and with and without washing effect due to rainfalls. The evaluation results showed that penetrated chlorides depend on amount of airborne chlorides and duration of exposure. In addition, a notable tendency of having deeper chloride penetration and higher chloride content in concrete members under long-term exposure was observed.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.17 no.2
• /
• pp.83-91
• /
• 1984

The formation and structure of tidal front in the eastern part of the Yellow Sea were studied based on the oceanographic data compiled during the periods of $1982{\sim}1983$ and $1966{\sim}1970$. Well-defined fronts occurring in the Yellow Sea in summer mark the boundary between the stratified and vertically mixed regimes. The occurrence of vertically mixed regimes may be interpreted in terms of available turbulent kinematic energy of tidal currents. The tidal frontal regions were determined by horizontal gradients of temperature, salinity and dissolved oxygen, and were verified by water colour and transparency. In summer the tidal fronts were found at depths of $15{\sim}25m$ at about 20 miles from the shore. Potential energy of vortical stratification in the tidal frontal region was 10 $Joule/m^3$. The stratification parameter in the frontal region computed from the numerical tidal model was $S_p=1.0.$ Tidal front is formed in regions with $S_p=1-1.5,$ if surface heat flux are constant. Waters in the stratified region have the layer structures of wind-mixed surface layer, thermocline and tidal-mixed bottom layer. In the vertically mixed region, however, sea water is nearly homogeneous. in winter no distinctive tidal front was seen.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.1
• /
• pp.1-7
• /
• 2017

In general, the large loads which are applied from explosion, impact, earthquake and wind at a short time caused the materials of structures to large deformations, rotations and strains locally. If such phenomena will be analyzed, hydrocodes which can be considered fluid-structure interaction under computational continuum mechanics are inevitably needed. Also, the explosion mechanism is so complicated, it is reasonable that the behaviors of structure are predicted through explosion analyses and experiment at the same time. But, unfortunately, it is true that explosion experiments are limited to huge cost, large experiment facilities and safety problems. Therefore, in this study, it is shown that the results of explosion analyses using the AUTODYN are agreed with those of existing explosion experiments for reinforced concrete slabs within reasonable error limits. And the explosion damage of the same reinforced concrete slab are assessed for quite different reinforcement arrangement spacings, concrete cover depths, and vertical reinforcements. From the explosion analyses, it is known that the more the ratio of slab thickness to reinforcement arrangement spacing is increased, and small-diameter reinforcements are used than large-diameter reinforcements on the same reinforcement ratio, and vertical reinforcements are used, the more the anti-knock capacities are improved.