This study suggests a machine learning model for predicting the production quality of free-machining 303-series stainless steel small rolling wire rods according to the manufacturing process's operation condition. The operation condition involves 37 features such as sulfur, manganese, carbon content, rolling time, and rolling temperature. The study procedure includes data preprocessing (integration and refinement), exploratory data analysis, feature selection, machine learning modeling. In the preprocessing stage, missing values and outlier are removed, and variables for the interaction between processes and quality influencing factors identified in existing studies are added. Features are selected by variable importance index of lasso regression, extreme gradient boosting (XGBoost), and random forest models. Finally, logistic regression, support vector machine, random forest, and XGBoost is developed as a classifier to predict good or defective products with new operating condition. The hyper-parameters for each model are optimized using k-fold cross validation. As a result of the experiment, XGBoost showed relatively high predictive performance compared to other models with accuracy of 0.9929, specificity of 0.9372, F1-score of 0.9963 and logarithmic loss of 0.0209. In this study, the quality prediction model is expected to be able to efficiently perform quality management by predicting the production quality of small rolling wire rods in advance.

통영 LNG 기지에서 방류되는 냉배수가 진해만에 미치는 영향을 알아보고, 냉배수의 활용 방안 모색을 위해 총 4개의 냉배수 방류량에 대한 진해만의 환경변화를 1년간(2018년) 모의하였다. 실제 냉배수 방류량인 Case1(10 m3 sec-1)의 모의 결과, 모든 분기에서 냉배 수에 의한 진해만의 환경변화는 매우 미미하게 나타났다. 모의 방류량인 Case2(100 m3 sec-1)의 경우 방류구 반경 5 km 범위에서 1 ~ 3℃의 수온 감소를 보였으며, Case3(1000 m3 sec-1)에서는 방류구 반경 8 km 범위에서 최대 4 ~ 5℃의 수온이 감소하였고 진해만 전 해역에 걸쳐 냉 배수가 확산하는 결과를 보였다. 플랑크톤의 성장 속도는 최대 15% 감소하였으며(11월), 대형조류의 성장 속도는 행암만 부근에서 최대 6 % 증가하는 결과를 보였다. 상기 결과로부터 통영 LNG 기지에서 방류되는 냉배수에 의한 진해만의 환경변화는 미미한 것을 확인하였 다. 또한 Case3 결과로부터 국소지역의 ‘적조 방재’, ‘해조류 성장’을 목적으로 냉배수의 활용이 가능할 것으로 기대된다.

본 연구는 기상청 현업모델(LDAPS)로부터 예측된 서울의 도시열섬 강도와 지상 기온을 AWS 관측과 비교 평가하였다. 관측된 서울의 열섬 강도는 봄과 겨울동안 증가하며 여름동안 감소한다. 열섬 강도의 시간적 변동 경향은 새벽 시간 최대, 오후에 최소를 보인다. 기상청 국지기상예측시스템(LDAPS)으로부터 예측된 열섬 강도는 여름철 과대모의, 겨울철 과소모의 특징을 보인다. 특히 여름철은 주간에 과대 모의 경향이 증가하며, 겨울은 새벽 시간 과소 모의 오차가 크게 나타난다. LDAPS에서 예측된 지면 기온의 오차는 여름철 감소하며 겨울철 증가한다. 겨울철 열섬 강도의 과소 모의는 도시 기온의 과소 모의와 관련되었으며, 여름철 열섬 강도의 과대 모의는 교외 지역 기온의 과소 모의로부터 기인하는것으로 판단된다. 도시 열섬강도 예측성 개선을 위하여 도시효과를 고려하는 도시캐노피모델을 LDAPS와 결합하여 2017년 여름 기간동안 모의하였다. 도시캐노피모델 적용 후 도시의 지면 기온의 오차는 개선되었다. 특히 오전시간 과소모의되는 기온의 오차 개선 효과가 뚜렷하였다. 도시캐노피모델은 여름동안 과대 모의하는 도시열섬강도를 약화시키는 개선 효과를 보였다.

This study was conducted to determine the possibility of estimating the daily mean temperature for a specific location based on the climatic data collected from the nearby Automated Synoptic Observing System (ASOS) and Automated Weather System(AWS) to improve the accuracy of the climate data in forage yield prediction model. To perform this study, the annual mean temperature and monthly mean temperature were checked for normality, correlation with location information (Longitude, Latitude, and Altitude) and multiple regression analysis, respectively. The altitude was found to have a continuous effect on the annual mean temperature and the monthly mean temperature, while the latitude was found to have an effect on the monthly mean temperature excluding June. Longitude affected monthly mean temperature in June, July, August, September, October, and November. Based on the above results and years of experience with climate-related research, the daily mean temperature estimation was determined to be possible using longitude, latitude, and altitude. In this study, it is possible to estimate the daily mean temperature using climate data from all over the country, but in order to improve the accuracy of daily mean temperature, climatic data needs to applied to each city and province.

The objective of this study was to access the effect of climate and soil factors on alfalfa dry matter yield (DMY) by the contribution through constructing the yield prediction model in a general linear model considering climate and soil physical variables. The processes of constructing the yield prediction model for alfalfa was performed in sequence of data collection of alfalfa yield, meteorological and soil, preparation, statistical analysis, and model construction. The alfalfa yield prediction model used a multiple regression analysis to select the climate variables which are quantitative data and a general linear model considering the selected climate variables and soil physical variables which are qualitative data. As a result, the growth degree days(GDD) and growing days(GD), and the clay content(CC) were selected as the climate and soil physical variables that affect alfalfa DMY, respectively. The contributions of climate and soil factors affecting alfalfa DMY were 32% (GDD, 21%, GD 11%) and 63%, respectively. Therefore, this study indicates that the soil factor more contributes to alfalfa DMY than climate factor. However, for examming the correct contribution, the factors such as other climate and soil factors, and the cultivation technology factors which were not treated in this study should be considered as a factor in the model for future study.

Evaluating the quantitative damage to rocks through acoustic emission (AE) has become a research focus. Most studies mainly used one or two AE parameters to evaluate the degree of damage, but several AE parameters have been rarely used. In this study, several data-driven models were employed to reflect the combined features of AE parameters. Through uniaxial compression tests, we obtained mechanical and AE-signal data for five granite specimens. The maximum amplitude, hits, counts, rise time, absolute energy, and initiation frequency expressed as the cumulative value were selected as input parameters. The result showed that gradient boosting (GB) was the best model among the support vector regression methods. When GB was applied to the testing data, the root-mean-square error and R between the predicted and actual values were 0.96 and 0.077, respectively. A parameter analysis was performed to capture the parameter significance. The result showed that cumulative absolute energy was the main parameter for damage prediction. Thus, AE has practical applicability in predicting rock damage without conducting mechanical tests. Based on the results, this study will be useful for monitoring the near-field rock mass of nuclear waste repository.

This ammonia prediction study was performed using the time-series artificial neural network model, Long-short term memory (LSTM), after long-term monitoring of ammonia and environmental factors (ventilation rate (V), temperature (T), humidity (RH)) from a slurry finishing pig farm on mechanical ventilation system. The difference with the actual ammonia concentration was compared through prediction of the last three days of the entire breeding period. As a result of the analysis, the model which had a low correlation (ammonia concentration and humidity) was confirmed to have less error values than the models that did not. In addition, the combination of two or more input values [V, RH] and [T, V, RH] showed the lowest error value. In this study, the sustainability period of the model trained by multivariate input values was analyzed for about two days. In addition, [T, V, RH] showed the highest predictive performance with regard to the actual time of the occurrence of peak concentration compared to other models . These results can be useful in providing highly reliable information to livestock farmers regarding the management of concentrations through artificial neural network-based prediction models.

본 연구는 화재진압 및 피난활동을 지원하는 딥러닝 기반의 알고리즘 개발에 관한 기초 연구로 선박 화재 시 연기감지기가 작동하기 전에 검출된 연기 데이터를 분석 및 활용하여 원격지까지 연기가 확산 되기 전에 연기 확산거리를 예측하는 것이 목적이다. 다음과 같은 절차에 따라 제안 알고리즘을 검토하였다. 첫 번째 단계로, 딥러닝 기반 객체 검출 알고리즘인 YOLO(You Only Look Once)모델에 화재시뮬레이션을 통하여 얻은 연기 영상을 적용하여 학습을 진행하였다. 학습된 YOLO모델의 mAP(mean Average Precision)은 98.71%로 측정되었으며, 9 FPS(Frames Per Second)의 처리 속도로 연기를 검출하였다. 두 번째 단계로 YOLO로부터 연기 형상이 추출된 경계 상자의 좌표값을 통해 연기 확산거리를 추정하였으며 이를 시계열 예측 알고리즘인 LSTM(Long Short-Term Memory)에 적용하여 학습을 진행하였다. 그 결과, 화재시뮬레이션으로부터 얻은 Fast 화재의 연기영상에서 경계 상자의 좌표값으로부터 추정한 화재발생~30초까지의 연기 확산거리 데이터를 LSTM 학습모델에 입력하여 31초~90초까지의 연기 확산거리 데이터를 예측하였다. 그리고 추정한 연기 확산거리와 예측한 연기 확산거리의 평균제곱근 오차는 2.74로 나타났다.

The increased turbidity in rivers during flood events has various effects on water environmental management, including drinking water supply systems. Thus, prediction of turbid water is essential for water environmental management. Recently, various advanced machine learning algorithms have been increasingly used in water environmental management. Ensemble machine learning algorithms such as random forest (RF) and gradient boosting decision tree (GBDT) are some of the most popular machine learning algorithms used for water environmental management, along with deep learning algorithms such as recurrent neural networks. In this study GBDT, an ensemble machine learning algorithm, and gated recurrent unit (GRU), a recurrent neural networks algorithm, are used for model development to predict turbidity in a river. The observation frequencies of input data used for the model were 2, 4, 8, 24, 48, 120 and 168 h. The root-mean-square error-observations standard deviation ratio (RSR) of GRU and GBDT ranges between 0.182~0.766 and 0.400~0.683, respectively. Both models show similar prediction accuracy with RSR of 0.682 for GRU and 0.683 for GBDT. The GRU shows better prediction accuracy when the observation frequency is relatively short (i.e., 2, 4, and 8 h) where GBDT shows better prediction accuracy when the observation frequency is relatively long (i.e. 48, 120, 160 h). The results suggest that the characteristics of input data should be considered to develop an appropriate model to predict turbidity.

The sunspot area is a critical physical quantity for assessing the solar activity level; forecasts of the sunspot area are of great importance for studies of the solar activity and space weather. We developed an innovative hybrid model prediction method by integrating the complementary ensemble empirical mode decomposition (CEEMD) and extreme learning machine (ELM). The time series is first decomposed into intrinsic mode functions (IMFs) with different frequencies by CEEMD; these IMFs can be divided into three groups, a high-frequency group, a low-frequency group, and a trend group. The ELM forecasting models are established to forecast the three groups separately. The final forecast results are obtained by summing up the forecast values of each group. The proposed hybrid model is applied to the smoothed monthly mean sunspot area archived at NASA's Marshall Space Flight Center (MSFC). We find a mean absolute percentage error (MAPE) and a root mean square error (RMSE) of 1.80% and 9.75, respectively, which indicates that: (1) for the CEEMD-ELM model, the predicted sunspot area is in good agreement with the observed one; (2) the proposed model outperforms previous approaches in terms of prediction accuracy and operational efficiency.

In this study, in order of to reflect the mold deformation in the injection molding process to design of mold, the mold deformation was analyzed by performing flow and structural analysis. The 5 inch LGP(light guide plate) mold, platen and tie bar were modeled and applied to the analysis. The result of melt pressure from flow analysis was extracted for use as boundary conditions acting on the mold surface in the structural analysis. In order to evaluate the accuracy of simulation analysis results, injection molding was performed under the process conditions of simulation. As a results, the mold deformation during injection molding tends to be similar that of injection pressure, and it is confirmed that it shows the behavior and properties of melt resins. Compared with the simulation and experiment, the error of the maximum mold deformation in the injection phase was 4.20%.

태풍 내습 시 신속하고 정확한 해일고 예측은, 연안재해 대응에 필수적인 요소이다. 이러한 해일고의 예측을 위해서 기존에는 태풍예측정보를 수치모델에 적용하여 예측자료를 생산하는 것이 대부분이였다. 이러한 방법은 대용량의 컴퓨팅 자원과 시간이 소요된다는 단점이 있다. 최근에는 인공지능 기반으로 신속하게 예측자료를 생산하는 연구가 다양한 분야에서 진행되고 있으며, 본 연구에서는 인공지능 기반 해일고 예측을 수행하였다. 인공지능 적용을 위해서는 많은 수의 학습자료가 필요하게 되며, 기왕 발생태풍은 개수가 한정되어 있어 본 연구에서는 TCRM(Tropical Cyclone Risk Model)을 통하여 합성태풍을 생성하고, 이를 폭풍해일 모델에 적용하여 해일고 자료를 생성한 후, 학습자료로 활용하였다. 인공지능으로 예측한 해일고와 실제 발생 태풍에 대한 비교 결과, RMSE(Root Mean Square Error)는 0.09 ~ 0.30 m, CC(Correlation Coefficient)는 0.65 ~ 0.94, 최대 해일고의 ARE(Absolute Relative Error)는 1.0 ~ 52.5 %로 분석되었다. 특정 태풍/지점에 서는 다소 오차가 크게 나타나고 있으나, 향후 학습자료의 최적화 등을 통하여 정확도를 개선할 수 있을 것으로 기대된다.

유리섬유강화 모르타르 관을 구성하는 보강섬유는 직교이방성 부재로 간주되며 재료의 성질은 서로 직각을 이루는 두 개의 축을 기준으로 정의된다. 유리섬유 모르타르 관의 구조적 거동 해석을 수행하기 위해서 길이방향과 원주방향의 재료의 역학적 성질, 즉 탄성계수, 전단탄성계수, 포아송비 등이 필요하며 각각의 성질들은 실험을 통해 결정하였다. 이 실험으로부터 구한 각각의 역학적 성질을 적용하여 간소화된 유한요소해석방법을 제안하기 위해 적층판 이론으로부터 유리섬유강화 모르타르 관의 탄성계수를 계산하고, 계산된 탄성계수를 적용하여 유한요소 해석을 수행하였다. 또한, 유한요소해석과 편평시험을 통해 구한 하중-변위 관계를 비교하였으며 ASTM D2412에서 제시하고 하고 있는 관의 강성 값을 유한요소해석과 실험을 통해 예측하여 비교하였다.

Postal logistics organizations are characterized as having high labor intensity and short response times. These characteristics, along with rapid change in mail volume, make load scheduling a fundamental concern. Load analysis of major postal infrastructures such as post offices, sorting centers, exchange centers, and delivery stations is required for optimal postal logistics operation. In particular, the performance of mail traffic forecasting is essential for optimizing the resource operation by accurate load analysis. This paper addresses a traffic forecast problem of postal parcel that arises at delivery stations of Korea Post. The main purpose of this paper is to describe a method for predicting short-term traffic of postal parcel based on self-similarity analysis and to introduce an application of the traffic prediction model to postal logistics system. The proposed scheme develops multiple regression models by the clusters resulted from feature engineering and individual models for delivery stations to reinforce prediction accuracy. The experiment with data supplied by main postal delivery stations shows the advantage in terms of prediction performance. Comparing with other technique, experimental results show that the proposed method improves the accuracy up to 45.8%.

There has been increasing interest in UHPC (Ultra-High Performance Concrete) materials in recent years. Owing to the superior mechanical properties and durability, the UHPC has been widely used for the design of various types of structures. In this paper, machine learning based compressive strength prediction methods of the UHPC are proposed. Various regression-based machine learning models were built to train dataset. For train and validation, 110 data samples collected from the literatures were used. Because the proportion between the compressive strength and its composition is a highly nonlinear, more advanced regression models are demanded to obtain better results. The complex relationship between mixture proportion and concrete compressive strength can be predicted by using the selected regression method.