IIn the context of site response analysis, the use of shear wave velocity (  ) profiles that consider the seismological rock (  ≥ 3,000 m/s) depth is recommended. This study proposes regression analysis and machine learning-based models to predict deep   profiles for a specialized excavated rock site in South Korea. The regression model was developed by modifying mathematical expressions from a previous study and analyzing the correlation between   and model variables to predict deep   beyond 50 m. The machine learning models, designed using tree-based algorithms and a fully connected hierarchical structure, were developed to predict   from 51 m to 300 m at 1 m intervals. These models were validated by comparing them with measured deep   profiles and accurately estimating the trend of deep   variations. The proposed prediction models are expected to improve the accuracy of ground motion predictions for a specialized excavated rock site in Korea.

에스컬레이터는 공공시설과 다중이용시설에서 필수적인 이동 수단으로 사용되며, 특히 고령 인구 증가와 함께 사고 발생률이 꾸준히 증가하고 있다. 이에 따라 공공시설의 안전 관리를 강화하고 중대한 시민 재해를 예방하기 위한 사고 예측 기술의 필요성이 대두되고 있으며, 본 연구는 2010년부터 2022년까지 13년간의 에스컬레이터(무빙워크 포함) 사고 데이터를 활용하여 다중선형회귀분석과 로지스틱 회귀분석을 기반으로 사고 예측 모델을 개발하였다. 다중선형회귀분석을 통해 사고 발생 건수 예측 모델을 구축하였고, 로지스틱 회귀분석을 통해 전도사고의 발생 확률을 분석하여 주요 변수와 영향을 도출하였다. 연구 결과, 이용자 과실과 같은 요인이 사고 발생과 피해 심각성에 가장 큰 영향을 미치는 변수 로 확인되었다. 본 연구에서 제시된 예측 모델은 사고 예방을 위한 체계적인 안전 관리 및 정책 수립에 유용한 자료로 활용될 수 있으며, 공공 및 민간 영역에서의 ESG 활동에도 기여할 수 있을 것이다.

본 연구에서는 박스 구조물의 부재력 예측을 위한 다양한 딥러닝 모델의 정확성을 비교하고자 하였다. 이를 위해 상용 유한 요소 프로그램인 MIDAS를 이용하여 300개의 유한요소모델을 작성하고, 수치해석을 수행하여 딥러닝 모델에 적용하기 위한 학습데이 터를 생성하였다. 또한, 딥러닝 모델의 정확성을 비교하기 위해 MLP, CNN, RNN 및 LSTM과 같은 다양한 신경망 모델과 Adam, SGD, RMSprop 및 Adamax 등 최적화 알고리즘을 교차 적용하여 16개의 딥러닝 모델을 생성하였다. 그 결과 Adam 최적화 알고리즘 이 모든 모델에서 가장 우수한 성능을 보여주었으며, 특히 MLP 모델에서 가장 높은 R2 값을 나타내었다. 이를 통해, 박스 구조물의 부재력 예측을 위한 최적의 딥러닝 모델 구성은 Adam optimizer와 MLP 구조임을 확인하였다.

This study proposes a weight optimization technique based on Mixture Design of Experiments (MD) to overcome the limitations of traditional ensemble learning and achieve optimal predictive performance with minimal experimentation. Traditional ensemble learning combines the predictions of multiple base models through a meta-model to generate a final prediction but has limitations in systematically optimizing the combination of base model performances. In this research, MD is applied to efficiently adjust the weights of each base model, constructing an optimized ensemble model tailored to the characteristics of the data. An evaluation of this technique across various industrial datasets confirms that the optimized ensemble model proposed in this study achieves higher predictive performance than traditional models in terms of F1-Score and accuracy. This method provides a foundation for enhancing real-time analysis and prediction reliability in data-driven decision-making systems across diverse fields such as manufacturing, fraud detection, and medical diagnostics.

The purpose of this study was to develop a more accurate model for predicting the in-situ compressive strength of concrete pavements using Internet-of-Things (IoT)-based sensors and deep-learning techniques. This study aimed to overcome the limitations of traditional methods by accounting for various environmental conditions. Comprehensive environmental and hydration data were collected using IoT sensors to capture variables such as temperature, humidity, wind speed, and curing time. Data preprocessing included the removal of outliers and selection of relevant variables. Various modeling techniques, including regression analysis, classification and regression tree (CART), and artificial neural network (ANN), were applied to predict the heat of hydration and early compressive strength of concrete. The models were evaluated using metrics such as mean absolute error (MAE) to determine their effectiveness. The ANN model demonstrated superior performance, achieving a high prediction accuracy for early-age concrete strength, with an MAE of 0.297 and a predictive accuracy of 99.8%. For heat-of-hydration temperature prediction, the ANN model also outperformed the regression and CART models, exhibiting a lower MAE of 1.395. The analysis highlighted the significant impacts of temperature and curing time on the hydration process and strength development. This study confirmed that AI-based models, particularly ANNs, are highly effective in predicting early-age concrete strength and hydration temperature under varying environmental conditions. The ability of an ANN model to handle non-linear relationships and complex interactions among variables makes it a promising tool for real-time quality control in construction. Future research should explore the integration of additional factors and long-term strength predictions to further enhance the model accuracy.

In this paper, we deal with the design of a model predictive control (MPC) for precise speed servo control of DC motor systems. The proposed controller is designed in the form of optimal control that calculates and outputs the optimized control input under constraints for each sampling. In particular, MPC designs the control inputs in advance for each sampling and predicts the outputs using them. Thus, it shows excellent control performance even in the case of disturbance or model uncertainty. The effectiveness of the proposed controller was demonstrated through computer simulations using MATLAB/Simulink and DC motor experimental system using real time controller. Moreover, the effectiveness of the proposed controller was confirmed by comparing its control performance with PID controller, which was tested under the same experimental condition as the MPC.

This study integrates TabTransformer and CTGAN for predicting job satisfaction among South Korean college graduates. TabTransformer handles complex tabular data relationships with self-attention, while CTGAN generates high-quality synthetic samples. The combined approach achieves an accuracy of 0.85, precision of 0.83, recall of 0.82, F1-score of 0.82, and an AUC of 0.88. Cross-validation confirms the model's robustness and generalizability with a mean accuracy of 0.85 and a standard deviation of 0.008. The integration of TabTransformer and CTGAN enhances predictive accuracy and model generalizability, providing valuable insights for employment policy and research.

기후 변화에 의해 해수면 온도 상승, 태풍의 최고 강도 북상, 태풍 강도 증가가 나타나고 있으며, 미래의 태풍 강도 변화가 더 심화될 것으로 예상하고 있다. 본 논문에서는 기후 변화 시나리오에 의해서 발생할 수 있는 한반도 부근의 태풍 강도를 예측하기 위하여 딥러닝 기반 태풍 강도 예측 모델을 개발하였다. 기후 예측정보를 이용하여 미래 기후 변화 환경장 변화에 따른 태풍의 강도를 예측할 수 있도록 과거 환경장을 학습 자료로 사용하였다. 학습자료는 1980년에서 2022년까지의 태풍 발생 빈도가 높은 6~10월의 기상 및 해양 재분 석 월평균 자료와 Best Track 태풍 241개를 입력자료로 사용하였다. 환경장 변화에 따른 태풍 강도 예측을 위해 자료의 공간적인 특징과 시간적인 특징을 함께 고려하는 딥러닝 모델인 ConvLSTM 기반으로 모델을 개발하였다. 태풍 트랙 시퀀스의 각 이동 경로에 대한 월평균 환경장 자료를 모델에 학습하여 태풍의 중심 기압을 예측하였다. 태풍의 공간적 특성을 반영할 수 있도록 범위를 설정하여 입력자료로 학습하였으며, 5°⨉ 5°의 범위일 때 가장 좋은 결과를 보였다. 몬테카를로 방법을 이용한 민감도 실험을 통해 모델 예측에 가장 큰 영향을 미치는 변수는 SST로 확인되었다.

본 연구는 기온 상승에 따른 개별 콘크리트 슬래브의 팽창과 그로 인한 Pavemnent Growth 및 Blow-up 현상을 분석하고 예측하기 위 해 수행되었다. 기온이 상승함에 따라 슬래브는 팽창하며, 콘크리트 슬래브들의 팽창량은 팽창 줄눈 사이에 존재하는 모든 수축 줄눈 이 닫히게 될 정도로 발생하게 되고 그 결과 모든 슬래브들이 접촉하게 된다. 온도의 추가적인 증가로 슬래브가 계속해서 팽창하게 되면 팽창 줄눈의 수축 허용 폭을 초과하는 경우 일체화된 슬래브 내에서 압축 응력이 발생하게 되며, 이러한 현상을 "Pavement Growth"라 정의된다. 이로 인해 콘크리트 포장은 팽창하면서 파손이나 균열에서 좌굴 및 파괴와 같은 압력 관련 문제를 일으킬 수 있 다. 이는 교량 및 도로 내 접근 구조물과 같은 인접 구조물에도 손상을 줄 수 있다. 그러나 현재 사용 가능한 이론적 해결책이나 Pavement Growth 평가 방법과 Blow-up 예측에 관한 연구는 매우 제한적이다. 따라서 본 연구에서는 콘크리트 포장의 팽창을 예측하기 위해 Pavement Growth 및 Blow-up 분석 모델인 PGBA(Pavement Growth and Blow-up Analysis) Model을 개발하였다. 이 모델은 기후 조 건, 포장 구조, 재료, 팽창 줄눈 등의 요인을 고려하였다. 본 모델은 일체화된 슬래브가 팽창하여 팽창 줄눈의 수축 허용 폭을 초과하 는 시기를 결정한다. 슬래브와 기층 사이의 Frictional Darg 및 슬래브의 End Restraint으로 인해 발생하는 압축 응력을 계산 할 수 있는 것이다. 또한 Geometric Imperfection의 변화에 따른 Blow-up Stress를 검토하기 위해 Large-scale Blow-up Test를 진행하였으며, 측정된 결과를 Blow-up 발생 임계값으로 사용하였다. 일체화된 슬래브 내부에 발생하는 연도별 압축 응력을 예측하고 Blow-up Stress와 비교 하여 압축 응력이 Blow-up Stress를 초과하는 시점을 Blow-up 발생 시기로 선정하였다.

Dynamic responses of nuclear power plant structure subjected to earthquake loads should be carefully investigated for safety. Because nuclear power plant structure are usually constructed by material of reinforced concrete, the aging deterioration of R.C. have no small effect on structural behavior of nuclear power plant structure. Therefore, aging deterioration of R.C. nuclear power plant structure should be considered for exact prediction of seismic responses of the structure. In this study, a machine learning model for seismic response prediction of nuclear power plant structure was developed by considering aging deterioration. The OPR-1000 was selected as an example structure for numerical simulation. The OPR-1000 was originally designated as the Korean Standard Nuclear Power Plant (KSNP), and was re-designated as the OPR-1000 in 2005 for foreign sales. 500 artificial ground motions were generated based on site characteristics of Korea. Elastic modulus, damping ratio, poisson’s ratio and density were selected to consider material property variation due to aging deterioration. Six machine learning algorithms such as, Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Artificial Neural Networks (ANN), eXtreme Gradient Boosting (XGBoost), were used t o construct seispic response prediction model. 13 intensity measures and 4 material properties were used input parameters of the training database. Performance evaluation was performed using metrics like root mean square error, mean square error, mean absolute error, and coefficient of determination. The optimization of hyperparameters was achieved through k-fold cross-validation and grid search techniques. The analysis results show that neural networks present good prediction performance considering aging deterioration.

This study aims to develop a regression model using data from the Ammunition Stockpile Reliability Program (ASRP) to predict the shelf life of 81mm mortar high-explosive shells. Ammunition is a single-use item that is discarded after use, and its quality is managed through sampling inspections. In particular, shelf life is closely related to the performance of the propellant. This research seeks to predict the shelf life of ammunition using a regression model. The experiment was conducted using 107 ASRP data points. The dependent variable was 'Storage Period', while the independent variables were 'Mean Ammunition Velocity,' 'Standard Deviation of Mean Ammunition Velocity,' and 'Stabilizer'. The explanatory power of the regression model was an R-squared value of 0.662. The results indicated that it takes approximately 55 years for the storage grade to change from A to C and about 62 years to change from C to D. The proposed model enhances the reliability of ammunition management, prevents unnecessary disposal, and contributes to the efficient use of defense resources. However, the model's explanatory power is somewhat limited due to the small dataset. Future research is expected to improve the model with additional data collection. Expanding the research to other types of ammunition may further aid in improving the military's ammunition management system.

본 연구는 연안해양 수치모델에 활용되는 LDAPS 강우예보 자료의 시공간적 오차와 한계점을 분석하고 자료의 신뢰성을 검증 하였다. LDAPS 강우자료의 검증은 진해만 주변 우량계 3개소를 기준으로 2020년의 강우를 비교하였으며 우량계와 LDAPS의 비교 결과, LDAPS 강우자료는 장기적인 강우의 경향은 대체로 잘 재현하였으나 단기적으로는 큰 차이를 보였다. 정량적인 강우량 오차는 연간 197.5mm였으며, 특히 하계는 285.4mm로 나타나 계절적으로 강우변동이 큰 시기일수록 누적 강우량의 차이가 증가하였다. 강우 발생 시점 의 경우 약 8시간의 시간 지연을 나타내어 LDPAS 강우자료의 시간적 오차가 연안해양환경 예측 시 정확도를 크게 감소시킬 수 있는 것 으로 나타났다. 연안의 강우를 정확히 반영하지 못하는 LDAPS 강우자료를 무분별하게 사용할 경우 연안역에서 오염물질 확산 또는 극한 강우로 인한 연안환경 변화 예측에 심각한 문제를 발생시킬 수 있으며 LDAPS 강우자료의 적절한 활용을 위해서는 검증과 추가적인 개선 을 통한 정확도 향상이 필요하다.

본 연구는 돼지 간 거리(PD), 돈사 내 상대 습도(RRH), 돈사 내 이산화탄소(RCO2) 세 가지 변수를 사용하여, 네 개의 데이터 세트를 구성하고, 이를 다중 선형 회귀(MLR), 서포트 벡터 회귀(SVR) 및 랜덤 포레스트 회귀(RFR) 세 가지 모델 기계학습(ML)에 적용하여, 돈사 내 온도(RT)를 예측하고자 한다. 2022년 10월 5일부터 11월 19일까지 실험을 진행하였다. Hik-vision 2D카메라를 사용하여, 돈사 내 영상을 기록하였다. 이후 ArcMap 프로그램을 사용하여, 돈사 내 영상에서 추출한 이미지 안 돼지의 PD를 계산하였다. 축산환경관리시스템(LEMS) 센서를 사용하여, RT, RRH 및 RCO2를 측정하였다. 연구 결과 각 변수 간 상관분석 시 RT와 PD 간의 강한 양의 상관관계가 나타났다(r > 0.75). 네 가지 데이터 세트 중 데이터 세트 3을 사용한 ML 모델이 높은 정확도가 나타났으며, 세 가지 회귀 모델 중에서 RFR 모델이 가장 우수한 성능을 보였다.

PURPOSES : The skid resistance between tires and the pavement surface is an important factor that directly affects driving safety and must be considered when evaluating the road performance. In especially wet conditions, the skid resistance of the pavement surface decreases considerably, increasing the risk of accidents. Moreover, poor drainage can lead to hydroplaning. This study aimed to develop a prediction equation for the roughness coefficient—that is, an index of frictional resistance at the interface of the water flow and surface material—to estimate the thickness of the water film in advance to prevent human and material damage. METHODS : The roughness coefficient can be changed depending on the surface material and can be calculated using Manning's theory. Here, the water level (h), which is included in the cross-sectional area and wetted perimeter calculations, can be used to calculate the roughness coefficient by using the water film thickness measurements generated after simulating specific rainfall conditions. In this study, the pavement slope, drainage path length, and mean texture depth for each concrete surface type (non-tined, and tined surfaces with 25-mm and 16-mm spacings) were used as variables. A water film thickness scale was manufactured and used to measure the water film thickness by placing it vertically on top of the pavement surface along the length of the scale protrusion. Based on the measured water film thickness, the roughness coefficient could be back-calculated by applying Manning's formula. A regression analysis was then performed to develop a prediction equation for the roughness coefficient based on the water film thickness data using the water film thickness, mean texture depth, pavement slope, and drainage path length as independent variables. RESULTS : To calculate the roughness coefficient, the results of the water film thickness measurements using rainfall simulations demonstrated that the water film thickness increased as the rainfall intensity increased under N/T, T25, and T16 conditions. Moreover, the water film thickness decreased owing to the linear increase in drainage capacity as the mean texture depth and pavement slope increased, and the shorter the drainage path length, the faster the drainage, resulting in a low water film thickness. Based on the measured water film thickness data, the roughness coefficient was calculated, and it was evident that the roughness coefficient decreased as the rainfall intensity increased. Moreover, the higher the pavement slope and the shorter the drainage path length, the faster the drainage reduced the water film thickness and increased the roughness coefficient (which is an indicator of the friction resistance). It was also evident that as the mean texture depth increased, the drainage capacity increased, which also reduced the roughness coefficient. CONCLUSIONS : As the roughness coefficient of the concrete road surface changes based on the environmental factors, road geometry, and pavement surface characteristics, we developed a prediction equation for the concrete pavement roughness coefficient that considered these factors. To validate the proposed prediction equation, a sensitivity analysis was conducted using the water film thickness prediction equation from previous studies. Existing models have limitations on the impact of the pavement type and rainfall intensity and can be biased toward underestimation; in contrast, the proposed model demonstrated a high correlation between the calculated and measured values. The water film thickness was calculated based on the road design standards in Korea—in the order of normal, caution, and danger scenarios—by using the proposed concrete pavement roughness coefficient prediction model under rainy weather conditions. Specifically, because the normal and caution stages occur before the manifestation of hydroplaning, it should be possible to prevent damage before it leads to the danger stage if it is predicted and managed in advance.

In this study, the magnetocaloric effect and transition temperature of bulk metallic glass, an amorphous material, were predicted through machine learning based on the composition features. From the Python module ‘Matminer’, 174 compositional features were obtained, and prediction performance was compared while reducing the composition features to prevent overfitting. After optimization using RandomForest, an ensemble model, changes in prediction performance were analyzed according to the number of compositional features. The R2 score was used as a performance metric in the regression prediction, and the best prediction performance was found using only 90 features predicting transition temperature, and 20 features predicting magnetocaloric effects. The most important feature when predicting magnetocaloric effects was the ‘Fe’ compositional ratio. The feature importance method provided by ‘scikit-learn’ was applied to sort compositional features. The feature importance method was found to be appropriate by comparing the prediction performance of the Fe-contained dataset with the full dataset.

Existing reinforced concrete (RC) building frames constructed before the seismic design was applied have seismically deficient structural details, and buildings with such structural details show brittle behavior that is destroyed early due to low shear performance. Various reinforcement systems, such as fiber-reinforced polymer (FRP) jacketing systems, are being studied to reinforce the seismically deficient RC frames. Due to the step-by-step modeling and interpretation process, existing seismic performance assessment and reinforcement design of buildings consume an enormous amount of workforce and time. Various machine learning (ML) models were developed using input and output datasets for seismic loads and reinforcement details built through the finite element (FE) model developed in previous studies to overcome these shortcomings. To assess the performance of the seismic performance prediction models developed in this study, the mean squared error (MSE), R-square (R2), and residual of each model were compared. Overall, the applied ML was found to rapidly and effectively predict the seismic performance of buildings according to changes in load and reinforcement details without overfitting. In addition, the best-fit model for each seismic performance class was selected by analyzing the performance by class of the ML models.

This study aims to predict return-to-work outcomes for workers injured in industrial accidents using a TabNet-RUSBoost hybrid model. The study analyzed data from 1,383 workers who had completed recuperation. Key predictors identified include length of recuperation, disability grade, occupation activity, self-efficacy, and socioeconomic status. The model effectively addresses class imbalance and demonstrates superior predictive performance. These findings underscore the importance of a holistic approach, incorporating both medical and psychosocial factors.

콩과 같은 밭작물은 주로 토양으로부터 수분을 공급받으며 토양 수분 조건에 따라 생육 반응이 민감하게 반응한다. 작물의 생육과 재배 지역의 토양 조건, 기상 등에 따라 적정 토양 수분을 유지하는 것은 작물 생산량의 증가를 위해 중요하다. 따라서, 본 연구에서는 머신러닝 기법을 이용하여 토양 수분 함량 예측 모델을 개발하였다. 깊이에 따른 토양 수분과 외기, 강수량 등 기상 변수와의 상관 관계를 구명하고, 깊이별 토양 수분예측을 위한 부분최소제곱회귀(PLSR) 모델을 알고리즘을 개발하였다. 콩 재배포장의 10cm, 20cm, 30cm 깊이의 토양수분은 FDR 방식의 센서로 측정하였 고, 콩 작물 주변 환경인자(재배환경의 기온, 상대습도, 풍속, 일사량, 일조시간)는 주변의 기상관측소에서 측정된 데이터를 이용하였다. 이를 이용하여 깊이별 미래의 토양수분함량 예측 모델을 개발한 결과, 10cm와 20cm깊이에서 주요 인자는 현재 토양수분함량과 기온이었으며, 30cm 깊이에서의 주요 인자는 현재 토양수분함량과 기온, 풍속으로 나타났다. 토양 깊이가 깊어짐에 따라 토양수분함량 예측 정확도가 향상되었으며, 이는 표면에 가까울수록 토양수분함량이 변화가 크기 때문으로 예상된다. 또한 미래의 토양 수분함량예측시 1시간 후 예측 정확도가 가장 우수하였으며, 이때의 Rv 2와 RMSEV가 10cm 깊이에서 0.993와 1.069%, 20cm 깊이에서 0.994와 0.821% 였으며, 30cm 깊이에서 0.999와 0.149% 였다. 본 연구 결과는 콩 생육환경 진단을 위해 재배 포장의 토양수분함량을 토양층별로 미래의 토양수분함량도 예측이 가능함을 보여준다.

Machine learning is widely applied to various engineering fields. In structural engineering area, machine learning is generally used to predict structural responses of building structures. The aging deterioration of reinforced concrete structure affects its structural behavior. Therefore, the aging deterioration of R.C. structure should be consider to exactly predict seismic responses of the structure. In this study, the machine learning based seismic response prediction model was developed. To this end, four machine learning algorithms were employed and prediction performance of each algorithm was compared. A 3-story coupled shear wall structure was selected as an example structure for numerical simulation. Artificial ground motions were generated based on domestic site characteristics. Elastic modulus, damping ratio and density were changed to considering concrete degradation due to chloride penetration and carbonation, etc. Various intensity measures were used input parameters of the training database. Performance evaluation was performed using metrics like root mean square error, mean square error, mean absolute error, and coefficient of determination. The optimization of hyperparameters was achieved through k-fold cross-validation and grid search techniques. The analysis results show that neural networks and extreme gradient boosting algorithms present good prediction performance.