• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.4
• /
• pp.12-22
• /
• 2021

This article suggests the machine learning model, i.e., classifier, for predicting the production quality of free-machining 303-series stainless steel(STS303) small rolling wire rods according to the operating condition of the manufacturing process. For the development of the classifier, manufacturing data for 37 operating variables were collected from the manufacturing execution system(MES) of Company S, and the 12 types of derived variables were generated based on literature review and interviews with field experts. This research was performed with data preprocessing, exploratory data analysis, feature selection, machine learning modeling, and the evaluation of alternative models. In the preprocessing stage, missing values and outliers are removed, and oversampling using SMOTE(Synthetic oversampling technique) to resolve data imbalance. Features are selected by variable importance of LASSO(Least absolute shrinkage and selection operator) regression, extreme gradient boosting(XGBoost), and random forest models. Finally, logistic regression, support vector machine(SVM), random forest, and XGBoost are developed as a classifier to predict the adequate or defective products with new operating conditions. The optimal hyper-parameters for each model are investigated by the grid search and random search methods based on k-fold cross-validation. As a result of the experiment, XGBoost showed relatively high predictive performance compared to other models with an accuracy of 0.9929, specificity of 0.9372, F₁-score of 0.9963, and logarithmic loss of 0.0209. The classifier developed in this study is expected to improve productivity by enabling effective management of the manufacturing process for the STS303 small rolling wire rods.

• Asian Journal of Turfgrass Science
• /
• v.20 no.1
• /
• pp.11-23
• /
• 2006

This study was carried out to propose low types of green speed expectation models for fast putting green management by changing mowing height($4.0{\sim}2.5$ mm) and timing of rolling, dew removal and dew removal+rolling. Ball roll distance data were taken from the creeping bentgrass(Agrostis palustris Huds. 'Penncross') practice green of east course at the Lakeside C.C. in October 18, 2001 and May 25, 2002. Data were subjected to multi-regression analysis using Statistical Package for the Social Science. Among four types of green speed expectation models, the best multiple-regression equation for fast green management was as follows; $Y_4=4.171-0.225{\cdot}X_1-0.038{\cdot}X_2$ (where, $Y_4$ : green speed(m) after single dew removal+single rolling, $X_1$ : mowing height($4.0{\sim}2.5,\;X_2$ : passage of time ($0{\sim}8$ h.)). The equation[single dew removal by using sponge roller $\rightarrow$ single mowing at 3.0 mm height or less $\rightarrow$ single rolling] explained to provide fast green over 3.2 m (Stimpmeter readings required for USGA championship play) until the end of first round. Therefore, this cultural practice system was believed to provide fast putting green condition for professional golf tournament

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.317-317
• /
• 2000

The mathematical set-up model was developed to reduce the mechanical property deviation in annealed and slightly rolled steel strip. The mechanical peculiarities of temper rolling process, such as high friction value and non-circular contact arc, Low reduction and non-negligible entry and exit elastic zones as well as central restricted deformation zone are all taken into account. The deformation of work rolls is calculated with the influence function method and arbitrary contact arc shape is permitted. The strip deformation is modeled by slab method and the entry and exit elastic deformation zones are included. The restricted deformation Bone near the neutral point is also considered. It was revealed that the new model has better accuracy than present regression model by statistical analysis with actual mill rolling data.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.01a
• /
• pp.265-267
• /
• 2020

본 논문에서는 열간압연 공정에 있어 효율적인 제품 생산 스케줄링에 필수적인 제품단위 에너지 사용 모델링 기법을 제안한다. 제안된 모델은 시스템 자원을 효율적 혹은 최소화하여 사용하여 실시간 처리량을 최대화함으로써 생산 예정 리스트로부터의 예측 작업 수행시간을 최소화할 수 있도록 한다. 제안된 기법은 다변량 선형 모델 방식으로 구성됨으로써 인공 지능 혹은 신경망 학습 방식에 비교하여 그 처리 속도가 빠르다는 장점을 가지고 있다. 본 논문에서는 서두에서 대상 응용처인 철강 산업과 열간 압연 공정 및 에너지 스케줄링에 대하여 간략히 언급한 후 본문에서 모델링을 위한 사전 데이터 수집, 모델링 기법을 자세히 설명하고 결론에서 모델의 정확도 성능을 최신 신경망 기법과 비교하여 검증하였다.

• Smart Structures and Systems
• /
• v.13 no.3
• /
• pp.453-471
• /
• 2014

The fault diagnosis of rolling element bearings has drawn considerable research attention in recent years because these fundamental elements frequently suffer failures that could result in unexpected machine breakdowns. Artificial intelligence algorithms such as artificial neural networks (ANNs) and support vector machines (SVMs) have been widely investigated to identify various faults. However, as the useful life of a bearing deteriorates, identifying early bearing faults and evaluating their sizes of development are necessary for timely maintenance actions to prevent accidents. This study proposes a new two-layer structure consisting of support vector regression machines (SVRMs) to recognize bearing fault patterns and track the fault sizes. The statistical parameters used to track the fault evolutions are first extracted to condense original vibration signals into a few compact features. The extracted features are then used to train the proposed two-layer SVRMs structure. Once these parameters of the proposed two-layer SVRMs structure are determined, the features extracted from other vibration signals can be used to predict the unknown bearing health conditions. The effectiveness of the proposed method is validated by experimental datasets collected from a test rig. The results demonstrate that the proposed method is highly accurate in differentiating between fault patterns and determining their fault severities. Further, comparisons are performed to show that the proposed method is better than some existing methods.

• Journal of the Korea Convergence Society
• /
• v.10 no.12
• /
• pp.271-280
• /
• 2019

Smart factory, a critical part of digital transformation, enables data-driven decision making using monitoring, analysis and prediction. Predictive maintenance is a key element of smart factory and the need is increasing. The purpose of this study is to analyze the degradation characteristics of a galvanizing kettle for the steel plating process and to predict the remaining useful life(RUL) for predictive maintenance. Correlation analysis, multiple regression, principal component regression were used for analyzing factors of the process. To identify the trend of degradation, a proposed rolling window was used. It was observed the degradation trend was dependent on environmental temperature as well as production factors. It is expected that the proposed method in this study will be an example to identify the trend of degradation of the facility and enable more consistent predictive maintenance.

• International Commerce and Information Review
• /
• v.16 no.5
• /
• pp.409-433
• /
• 2014

In this paper, I estimate long-run elasticities of US real GDP and real exchange rate between Korean Won and US Dollar on export of Korean goods to US and analyze changes in their trend by using VECM and rolling regression with a fixed window. For the purposes I use the year data from 1990 to 2013 which are selected from UNCTAD, Korea Trade Association(KTA), and Bank of Korea(BOK). The results are that the long-run elasticities of US real GDP vary from 2.849 to 2.938 for the period from 1990 to 2013 depending on the models in VECM and all of them are significant statistically. The elasticities of real exchange rate between Korean WON and US Dollar vary from 0.962 to 0.967 for the same period depending on the models in VECM and all of them are significant statistically. In case of the results through the OLS and the rolling regression, the long-run elasticities of US. real GDP are 3.015 for Basic Model, 2.949 for the modified Model 1, and 2.125 for the modified Model 2 for the period from 1990 to 2013 depending on the models and all of them are significant statistically. The average of long-run elasticities of real US GDP before the global financial crisis of 2008 is greater than that after the global financial crisis of 2008. On the other hand, the long-run elasticities of real exchange rate between Korean WON and US $ are 0.347 for Basic Model, 0.566 for the modified Model 1, and -0.217 for the modified Model 2 for the same period and all of them are significant statistically except for the modified Model 2. The average of long-run elasticities of real exchange rate before the global financial crisis of 2008 is greater than that after the global financial crisis of 2008.

• Journal of the Korean Society of Food Culture
• /
• v.19 no.6
• /
• pp.630-639
• /
• 2004

Korean households' expenditures on foodservices are on the steady increase. This paper aims to examine the foodservice expenditures of salary and wage earners's households by income decile group. This is analysed through comparing foodservice expenditures with private education expenditures because households' expenditures are likely to be weighted in favor of eating-out rather than private education. We also model the consumption function in terms of income and price, examining the responsiveness of private education demand and eating-out demand to changes in income and price using econometric methods such as regression, rolling regression and impulse response. This paper show that foodservice demand increases more than the private education does in the long-run. The result indicates that households are likely to evaluate the desire for foodservice more important than private education contrary to our expectations in the long-run. The impulse response analysis, however, suggests that households tend to increase private education expenditures rather than eating-out expenditures in the short-run.

• Environmental and Resource Economics Review
• /
• v.22 no.2
• /
• pp.251-279
• /
• 2013

Prices of oil products such as gasoline and diesel are deregulated since 1997 while electricity price is still controlled by government. This difference may explain recent discrepancy in the patterns of demand for oil products and electricity - constant increase in electricity consumption and stagnant demand for oil. To verify it empirically, we estimate price and income (production) elasticity of demand across time by using a rolling regression with 10 year-window based on monthly data for 1981-2011. Estimation results show that the sensitivity to price in demand for gasoline and diesel has increased since mid-90s while the elasticity of demand for electricity has become smaller. Second, income (production) elasticities of demand have shown no significant changes for both oil products and electricity. Third, cross-price elasticity was found meaningful only for gasoline before mid 1990s and for diesel after then.

• Journal of The Korean Society of Inherited Metabolic disease
• /
• v.11 no.1
• /
• pp.88-98
• /
• 2011

Lesch-Nyhan syndrome is a X-linked recessive disorder caused by a deficiency of the enzyme hypoxanthine-guanidine phosphoribosyltransferase (HPRT), enzyme to recycle purines. Case history: born induced vaginal delivery at 40 weeks complicated by premature membrane ruputure, body weight 2.820 gm. He showed failure to thrive showing severe protein aversion like milk products and pink daper. Developmental delay revealing rolling over at 10.5 month, followed by regression. Seizure at 2 months, His poor oral feeding was lifelong problem. Weak crying, spastic, choreoathetoid movement. Self mutilating behavior noted and diagnosed at age 3 years. No family history of consanguinity and neurological disorders. Method: Laboratory test, physical exam, imaging study and molecular. Clinical follow up Treat ment with allopurinol. Result: uric acid 10.5 mg/dL (N 3.5-7.9), APRT 151.1uM/ min/ml pro(25.7-101), HPRT 7.6 (N 233.5-701) and c.151C>T hemizygote (p,Arg51X). Abdominal sonogram showed staghorn calculi in both kidneys, brain MRI brain atrophy. Clinical follow up showed, seizure at 2 mo, developmental delay (head control and, rolling over at at 11mo, pointing body part at 2 yr 7 mo, eye hand coordination at 2 y 11mo,creeping at 3 y 7 mo, speaking words at 6 y 6 mo ),and developmental regression at 3 yr of age. Sleeping problem including insomnia and severe constipation. Self mutilating behavior (lip bite) started at 2.5 yr, neurologic sx including intermittent upward gaze accompanied by swallowing difficulty at 3 y 7 mo grand mal seizure at 4.5 yr and spastic extremity and trunchal hypotonia and choleoathetoid movement and ataxia at 6.5 yr. Scoliosis with severe spasticity at 9 yr 9 mo. Acute life threatening episode with irregular breathing at 9 yr and 9 mo, Emaciation and nephrolithiasis and recurrent pneumonia. Died suddenly at 10 yr 3 mo. Conclusion: life long feeding problem, chronic gut motility dysfunction, sleeping difficulty and progressing neurologic deterioration and nephrolithiasis despite normal serum uric acid maintence by allopurinol treatment.