• Journal of Biosystems Engineering
• /
• 제42권3호
• /
• pp.190-198
• /
• 2017

Purpose: A thin-layer drying equation was developed to analyze the drying processes of soybeans (white and black beans) and investigate drying conditions by verifying the suitability of existing grain drying equations. Methods: The drying rates of domestic soybeans were measured in a drying experiment using air at a constant temperature and humidity. The drying rate of soybeans was measured at two temperatures, 50 and $60^{\circ}C$, and three relative humidities, 30, 40 and 50%. Experimental constants were determined for the selected thin layer drying models (Lewis, Page, Thompson, and moisture diffusion models), which are widely used for predicting the moisture contents of grains, and the suitability of these models was compared. The suitability of each of the four drying equations was verified using their predicted values for white beans as well as the determination coefficient ($R^2$) and the root mean square error (RMSE) of the experiment results. Results: It was found that the Thompson model was the most suitable for white beans with a $R^2$ of 0.97 or greater and RMSE of 0.0508 or less. The Thompson model was also found to be the most suitable for black beans, with a $R^2$ of 0.97 or greater and an RMSE of 0.0308 or less. Conclusions: The Thompson model was the most appropriate prediction drying model for white and black beans. Empirical constants for the Thompson model were developed in accordance with the conditions of drying temperature and relative humidity.

• Journal of Radiation Protection and Research
• /
• 제41권3호
• /
• pp.237-244
• /
• 2016

Background: Korea Atomic Energy Research Institute is developing a fission product transport module for predicting the behavior of radioactive materials in the primary cooling system of a nuclear power plant as a separate module, which will be connected to a severe accident analysis code, Core Meltdown Progression Accident Simulation Software (COMPASS). Materials and Methods: This fission product transport (COMPASS-FP) module consists of a fission product release model, an aerosol generation model, and an aerosol transport model. In the fission product release model there are three submodels based on empirical correlations, and they are used to simulate the fission product gases release from the reactor core. In the aerosol generation model, the mass conservation law and Raoult's law are applied to the mixture of vapors and droplets of the fission products in a specified control volume to find the generation of the aerosol droplet. In the aerosol transport model, empirical correlations available from the open literature are used to simulate the aerosol removal processes owing to the gravitational settling, inertia impaction, diffusiophoresis, and thermophoresis. Results and Discussion: The COMPASS-FP module was validated against Aerosol Behavior Code Validation and Evaluation (ABCOVE-5) test performed by Hanford Engineering Development Laboratory for comparing the prediction and test data. The comparison results assuming a non-spherical aerosol shape for the suspended aerosol mass concentration showed a good agreement with an error range of about ${\pm}6%$. Conclusion: It was found that the COMPASS-FP module produced the reasonable results of the fission product gases release, the aerosol generation, and the gravitational settling in the aerosol removal processes for ABCOVE-5. However, more validation for other aerosol removal models needs to be performed.

• 생물정신의학
• /
• 제21권4호
• /
• pp.141-150
• /
• 2014

Objectives The purpose of this study is to evaluate the effects of attachment security, social support and health-related burden in the prediction of psychological distress and the mediation effects of social support and health-related burden in relationship between attachment security and psychological distress. Methods Finally, 161 patients were included for the analysis. Chi-square test and independent samples t-test were used for comparing differences between depressive/anxious group and non-depressive/non-anxious group. For evaluating the relationship among attachment security, social support, psychological distress and health-related burden, structural equation modeling analysis were performed. Results 40.7% and 32.0% of the patients have significant depressive symptoms and anxiety symptoms, respectively. In the analysis for testing the differences between groups who have psychological distress and who have not, there were no significant differences of sociodemographic factors and medical characteristics between groups, except for association between depressive symptoms and type of surgery (p = 0.01). Contrary to sociodemographic and medical characteristics, there were significant differences of health-related burden and two coping resources (attachment security and social support) between groups (all p < 0.01), except for the support from medical team in between anxious group and non-anxious group (p = 0.20). In the structural equation model analysis (Model fit : chi-square/df ratio = 0.8, root mean square error of approximation = 0.000, comparative fit index = 1.000, non-normed fit index =0.991), attachment security and social support emerged as an important predictor of psychopathology. Conclusions Attachment security and social support are important factors affecting the psychological distress. We suggest that individual attachment style and the social support state must be considered to approach the newly diagnosed breast cancer patients with psychological distress.

• 한국정보기술학회논문지
• /
• 제16권12호
• /
• pp.127-137
• /
• 2018

가축의 분만은 농가의 재산을 늘릴 수 있는 중요한 수단이므로 이를 관리하는 것은 농업 경영에 필수적인 항목이다. 특히 축우는 다른 가축에 비해 단가가 높고, 생산성 측면에서 농가의 소득과 밀접히 연관되어 있으며 축우의 42%는 밤에 분만이 이루어지고 있어 정확한 분만 예측은 더 중요하다고 할 수 있다. 그리하여 본 논문에서는 경구 투여용 센서를 통해 반추위 내의 심부 체온을 외부 환경의 간섭 없이 안정적으로 실시간 측정하고 이를 딥러닝에 적용함으로써 분만 시점을 예측하는 방법을 제안 하였고, 실제 축우를 대상으로 실험을 수행한 결과 실제 분만 시간 대비 평균 3시간 40분의 오차만 보여 기존 분만 예측 방법보다 정확하게 분만일을 예측하는 것을 확인하였다. 제안하는 방법을 통해 축우의 분만을 정확하게 예측하여 난산의 위험 없이 성공적으로 분만 하도록 도움을 줌으로써 농가의 경제적 피해를 절감할 수 있을 것으로 기대한다.

• Structural Engineering and Mechanics
• /
• 제78권4호
• /
• pp.485-496
• /
• 2021

So far analytical methods on collapse assessment of three-dimensional (3-D) steel frames have mainly focused on a single-column-removal scenario. However, the collapse of the Federal Building in the US due to car bomb explosion indicated that the loss of multiple columns may occur in the real structures, wherein the structures are more vulnerable to collapse. Meanwhile, the General Services Administration (GSA) in the US suggested that the removal of side columns of the structure has a great possibility to cause collapse. Therefore, this paper analytically deals with the robustness of 3-D steel frames in a two-side-column-removal (TSCR) scenario. Analytical method is first proposed to determine the collapse resistance of the frame during this column-removal procedure. The reliability of the analytical method is verified by the finite element results. Moreover, a design-based methodology is proposed to quickly assess the robustness of the frame due to a TSCR scenario. It is found the analytical method can reasonably predict the resistance-displacement relationship of the frame in the TSCR scenario, with an error generally less than 10%. The parametric numerical analyses suggest that the slab thickness mainly affects the plastic bearing capacity of the frame. The rebar diameter mainly affects the capacity of the frame at large displacement. However, the steel beam section height affects both the plastic and ultimate bearing capacity of the frame. A case study on a six-storey steel frame shows that the design-based methodology provides a conservative prediction on the robustness of the frame.

• 한국지형학회지
• /
• 제23권1호
• /
• pp.87-101
• /
• 2016

This study was conducted to develop a hydrological model of catchment water balance which is able to estimate irrigation return flows, so BROOK90-K (Kongju National University) was developed as a result of the study. BROOK90-K consists of three main modules. The first module was designed to simulate water balance for reservoir and its catchment. The second and third module was designed to simulate hydrological processes in rice paddy fields located on lower watershed and lower watershed excluding rice paddy fields. The models consider behavior of floodgate manager for estimating the storage of reservoir, and modules for water balance in lower watershed reflects agricultural factors, such as irrigation period and, complex sources of water supply, as well as irrigation methods. In this study, the models were applied on Guryangcheon stream watershed. R2, Nash-Sutcliffe efficiency (NS), NS-log1p, and root mean square error between simulated and observed discharge were 0.79, 0.79, 0.69, and 4.27 mm/d respectively in the model calibration period (2001~2003). Furthermore, the model efficiencies were 0.91, 0.91, 0.73, and 2.38 mm/d respectively over the model validation period (2004~2006). In the future, the developed BROOK90-K is expected to be utilized for various modeling studies, such as the prediction of water demand, water quality environment analysis, and the development of algorithms for effective management of reservoir.

• 한국전기전자재료학회논문지
• /
• 제35권2호
• /
• pp.134-142
• /
• 2022

기존의 시행착오를 거쳐 소재를 개발하는 방법은 조금씩 한계를 보이고 있는데, 왜냐하면 산업과 기술이 고도화되고 기능성 소재가 가져야 하는 특성은 복잡해지면서 그 요구치가 높아지고 있기 때문이다. 이를 극복하기 위해 데이터 기반의 인공신경망으로 복잡한 소재 공간을 빠르게 탐색하여 소재 개발을 가속화하고자 하는 연구들이 진행되고 있다. 특히 결정그래프 합성곱 인공신경망은 결정 소재의 구조에 따른 특성을 학습하는 인공신경망으로 소재의 특성(생성 에너지, 밴드갭, 부피 탄성 계수 등)을 양자역학 기반의 제일원리 계산보다 빠르게 예측한다. 본 논문에서는 46,629개의 결정구조 데이터와 그 생성 에너지를 공공데이터베이스에서 불러와 결정그래프 합성곱 인공신경망 모델을 학습시키고 이를 특성 예측에 적용해 보는 예제를 설명한다. 이를 통해 간단한 프로그래밍 지식으로 소재 특성 예측 모델을 재현해 보고 원하는 데이터 셋과 연구 분야에 적용할 수 있을 것으로 기대된다. 인공지능 모델의 개발은 앞으로 더 복잡한 특성을 가져야만 하는 소재의 개발을 위해 넓은 범위의 소재를 탐색해야만 하는 과정을 획기적으로 단축시켜 소재 개발의 가속화를 촉진시킬 것으로 생각된다.

• Geomechanics and Engineering
• /
• 제31권6호
• /
• pp.545-556
• /
• 2022

Evaluation and optimization of tunnel wall convergence (TWC) plays a vital role in preventing potential problems during tunnel construction and utilization stage. When convergence occurs at a high rate, it can lead to significant problems such as reducing the advance rate and safety, which in turn increases operating costs. In order to design an effective solution, it is important to accurately predict the degree of TWC; this can reduce the level of concern and have a positive effect on the design. With the development of soft computing methods, the use of deep learning algorithms and neural networks in tunnel construction has expanded in recent years. The current study aims to employ the long-short-term memory (LSTM) deep neural network predictor model to predict the TWC, based on 550 data points of observed parameters developed by collecting required data from different tunnelling projects. Among the data collected during the pre-construction and construction phases of the project, 80% is randomly used to train the model and the rest is used to test the model. Several loss functions including root mean square error (RMSE) and coefficient of determination (R²) were used to assess the performance and precision of the applied method. The results of the proposed models indicate an acceptable and reliable accuracy. In fact, the results show that the predicted values are in good agreement with the observed actual data. The proposed model can be considered for use in similar ground and tunneling conditions. It is important to note that this work has the potential to reduce the tunneling uncertainties significantly and make deep learning a valuable tool for planning tunnels.

• 한국지진공학회논문집
• /
• 제27권5호
• /
• pp.221-230
• /
• 2023

This study proposes a methodology for assessing seismic liquefaction hazard by implementing high-resolution three-dimensional (3D) ground models with high-density/high-precision site investigation data acquired in an area of interest, which would be linked to geotechnical numerical analysis tools. It is possible to estimate the vulnerability of earthquake-induced geotechnical phenomena (ground motion amplification, liquefaction, landslide, etc.) and their triggering complex disasters across an area for urban development with several stages of high-density datasets. In this study, the spatial-ground models for city development were built with a 3D high-precision grid of 5 m × 5 m × 1 m by applying geostatistic methods. Finally, after comparing each prediction error, the geotechnical model from the Gaussian sequential simulation is selected to assess earthquake-induced geotechnical hazards. In particular, with seven independent input earthquake motions, liquefaction analysis with finite element analyses and hazard mappings with LPI and LSN are performed reliably based on the spatial geotechnical models in the study area. Furthermore, various phenomena and parameters, including settlement in the city planning area, are assessed in terms of geotechnical vulnerability also based on the high-resolution spatial-ground modeling. This case study on the high-precision 3D ground model-based zonations in the area of interest verifies the usefulness in assessing spatially earthquake-induced hazards and geotechnical vulnerability and their decision-making support.

• Computers and Concrete
• /
• 제32권3호
• /
• pp.327-337
• /
• 2023

Concrete is the most widely used building material, with various types including high- and ultra-high-strength, reinforced, normal, and lightweight concretes. However, accurately predicting concrete properties is challenging due to the geotechnical design code's requirement for specific characteristics. To overcome this issue, researchers have turned to new technologies like machine learning to develop proper methodologies for concrete specification. In this study, we propose a highly accurate deep learning-based predictive model to investigate the compressive strength (UCS) of lightweight concrete with natural aggregates (pumice). Our model was implemented on a database containing 249 experimental records and revealed that water, cement, water-cement ratio, fine-coarse aggregate, aggregate substitution rate, fine aggregate replacement, and superplasticizer are the most influential covariates on UCS. To validate our model, we trained and tested it on random subsets of the database, and its performance was evaluated using a confusion matrix and receiver operating characteristic (ROC) overall accuracy. The proposed model was compared with widely known machine learning methods such as MLP, SVM, and DT classifiers to assess its capability. In addition, the model was tested on 25 laboratory UCS tests to evaluate its predictability. Our findings showed that the proposed model achieved the highest accuracy (accuracy=0.97, precision=0.97) and the lowest error rate with a high learning rate (R2=0.914), as confirmed by ROC (AUC=0.971), which is higher than other classifiers. Therefore, the proposed method demonstrates a high level of performance and capability for UCS predictions.