Recently, many studies have been conducted to improve quality by applying machine learning models to semiconductor manufacturing process data. However, in the semiconductor manufacturing process, the ratio of good products is much higher than that of defective products, so the problem of data imbalance is serious in terms of machine learning. In addition, since the number of features of data used in machine learning is very large, it is very important to perform machine learning by extracting only important features from among them to increase accuracy and utilization. This study proposes an anomaly detection methodology that can learn excellently despite data imbalance and high-dimensional characteristics of semiconductor process data. The anomaly detection methodology applies the LIME algorithm after applying the SMOTE method and the RFECV method. The proposed methodology analyzes the classification result of the anomaly classification model, detects the cause of the anomaly, and derives a semiconductor process requiring action. The proposed methodology confirmed applicability and feasibility through application of cases.

Ball stud parts are manufactured by a cold forging process, and fastening with other parts is secured through a head part cutting process. In order to improve process quality, stabilization of the forging quality of the head is given priority. To this end, in this study, a predictive model was developed for the purpose of improving forging quality. The prediction accuracy of the model based on 450 data sets acquired from the manufacturing site was low. As a result of gradually multiplying the data set based on FE simulation, it was expected that it would be possible to develop a predictive model with an accuracy of about 95%. It is essential to build automated labeling of forging load and dimensional data at manufacturing sites, and to apply a refinement algorithm for filtering data sets. Finally, in order to optimize the ball stud manufacturing process, it is necessary to develop a quality prediction model linked to the forging and cutting processes.

Due to environmental pollution, regulations on existing petroleum-based fuels are increasing day by day. LNG is in the spotlight as an eco-friendly fuel that does not emit NOx or SOx, but its boiling point is -163°C, so it needs to be handled with care. Materials that can be used at the above temperature are defined by IMO through the IGC Code. Among them, 9% nickel steel has great advantages in yield strength and tensile strength under cryogenic conditions, but it is difficult to use in arc welding such as FCAW for various reasons. This study is a study to apply fiber laser welding to solve this problem. As a previous study, this study conducted a study to find a welding heat source. After performing bead on plate welding, the optimal heat source was derived by analyzing the shape of the bead and adjusting the parameters of the heat source model. In this case, by applying the multi-island genetic algorithm, which is a global optimization algorithm, not the intuition of the researcher, accurate results could be derived in a wide range.

The government announced that it would ask the contractors not only the supplier but also the contractors to take the same responsibilities if they did not observe industrial accident safety measures from 2019. The semiconductor manufacturing process belongs to the representative disaster industry group in which the facility is directly located inside a closed space called clean room. According to previous studies, the semiconductor industry group used checklists for safety management of their suppliers. This study has developed a model for assessing suppliers by constructing a quantitative checklist item through the risk assessment methodology, laws and regulations. The evaluation model of the supplier set up through this study becomes the safety management standard in the semiconductor industry .Furthermore, it is applied to the partner companies in the operation of ISO 14001, 45001, I would like to apply it as a measure of performance management for CSR (Corporate Social Responsibility).

The manufacturing companies under Make-To-Order (MTO) production environment face highly variable requirements of the customers. It makes them difficult to establish preemptive production strategy through inventory management and demand forecasting. Therefore, the ability to establish an optimal production schedule that incorporates the various requirements of the customers is emphasized as the key success factor. In this study, we suggest a process of designing the simulation model for establishing production schedule and apply this model to the case of a flat glass processing company. The flat glass manufacturing industry is under MTO production environment. Academic research of flat glass industry is focused on minimizing the waste in the cutting process. In addition, in the practical view, the flat glass manufacturing companies tend to establish the production schedule based on the intuition of production manager and it results in failure of meeting the due date. Based on these findings, the case study aims to present the process of drawing up a production schedule through simulation modeling. The actual data of Korean flat glass processing company were used to make a monthly production schedule. To do this, five scenarios based on dispatching rules are considered and each scenario is evaluated by three key performance indicators for delivery compliance. We used B2MML (Business To Manufacturing Markup Language) schema for integrating manufacturing systems and simulations are carried out by using SIMIO simulation software. The results provide the basis for determining a suitable production schedule from the production manager's perspective.

본 논문은 개인의 비합리적인 이타주의 행동을 설명하는 규범활성화모델을 활용하여 공정무역제품 구매행동을 촉진하는 개인적, 규범적, 문화적 요인을 통합적으로 다루고 있다. 모델 관련, 예기된 감정(예기된 자부심과 예기된 죄책감)과 두 가지 상충하는 개인가치(박애주의 가치와 권력 가치)를 공정무역 관련 개인규범 활성화 선행요인으 로 개념화 하였다. 그리고 활성화된 개인규범이 공정무역제품 구매행동에 영향을 주는 것으로 보았고 특히 이들 간의 관계를 문화 클러스터(유교, 라틴 유럽)가 조절하는 것으로 개념화 하였다. 구조방정식을 통해 얻은 실증결 과는 예기된 자부심이 개인규범에 미치는 영향력이 예기된 죄책감 보다 크고, 상충되는 두 가지 개인가치 중 이 타적 가치인 박애주의 가치만이 개인규범에 유의한 영향을 미치는 것으로 나타났다. 그리고 활성화된 개인규범은 공정무역제품구매 행동에 유의한 영향을 주는 것으로 조사되었다. 문화 클러스터의 조절효과는 다집단비교 구조 방정식을 통해 분석하였다. 검증결과, 해당 경로에 대한 영향력이 라틴 유럽 클러스터 보다 유교 클러스터에서 더 강한 것으로 나타났다. 이러한 결과는 유교 클러스터가 라틴 유럽 클러스터에 비해 직접적인 비용지불에 따른 자기희생 정도가 더 크고, 사회 구성원 기대에 상응하는 도의적 의무감을 강하게 느끼기 때문에 공정무역제품 구 매 행위가 더 강하게 나타난 것으로 해석 가능하다. 본 논문은 규범활성화모델을 활용하여 공정무역제품 구매행 동을 비교 문화적 관점에서 접근하여 윤리적 소비자의 의사결정과정을 실증적으로 구명했다는데 의의가 있다.

Pulp and paper industry produces large volumes of wastewater and residual sludge waste, resulting in many issues in relation to wastewater treatment and sludge disposal. Contaminants in pulp and paper wastewater include effluent solids, sediments, chemical oxygen demand (COD), and biological oxygen demand (BOD), which should be treated by wastewater treatment processes such as coagulation and biological treatment. However, few works have been attempted to predict the treatment efficiency of pulp and paper wastewater. Accordingly, this study presented empirical models based on experimental data in laboratory-scale coagulation tests and compared them with statistical models such as artificial neural network (ANN). Results showed that the water quality parameters such as turbidity, suspended solids, COD, and UVA can be predicted using either linear or expoential regression models. Nevertheless, the accuracies for turbidity and UVA predictions were relatively lower than those for SS and COD. On the other hand, ANN showed higher accuracies than the emprical models for all water parameters. However, it seems that two kinds of models should be used together to provide more accurate information on the treatment efficiency of pulp and paper wastewater.

최근 국내외에서는 수질안정성 향상 및 부지면적 저감을 위해 막여과 공정도입이 활발한 추세이며 특히, 정수처리 분야에서는 정밀여과(Microfiltration) 및 한외여과(Ultrafiltration) 공정이 많이 적용되고 있다. 막여과 공정의 경제성 향상을 위해서는 세정 시점의 예측 및 세정 주기 연장이 매우 중요한 요소이다. 따라서, 본 연구에서는 인공신경망(Artificial neural network)을 활용하여 UF 공정차압(Transmembrane pressure) 예측 모델을 개발하고자 한다. 입력변수로는 유입수 온도, pH, 탁도 등의 일평균값을 이용하였다.

역삼투 해수담수화 공정에서 막 오염은 생산수량 감소 및 공정의 에너지 소비량 증가를 야기한다. 막간 차압 증가, 생산수량 감소 외에 막 저항 값의 증가는 막 오염 정도를 판단하는 수치로 사용이 가능하다. 특히 막 저항 값 기반의 세정은 막 오염 제어를 통해 역삼투 해수담수화 공정에서 막의 성능 유지 시 사용 가능하다. 이에 본 연구에서는 해수 수질 인자 및 공정 운전 인자에 기반하여 막 저항 값을 예측하는 알고리즘을 제안한다. 알고리즘은 해수담수화 플랜트의 운전 데이터에 기반하여 인자들과 막 저항 값 사이의 관계를 학습하고 검증과정을 거쳐 막 오염 발생 시점을 사전에 예측하는 방식으로 개발되었다. 예측 정확도를 분석하고 개발된 알고리즘의 수정을 통해 예측 정확도 향상을 위한 연구를 진행하였다.

This study provides a comparison and analysis of the predicted damages related to hazardous chemical substances used in “A” solar cell manufacturing process. In order to predict potential damages, different accident scenarios were established using the ALOHA model and the KOSHA guideline. This study evaluates chemical spills and leaks from cylinder and pipeline. Maximum distance of chemical movement, based on an initial concentration of 150 ppm, was estimated as up to 258 m in summer and 251 m in winter. The impacts of the leakage of chemicals such as ammonia, were dependent on the initial concentration of the chemical leaked, the atmospheric stability and temperature, and the wind speed. All of those however, were affected by air humidity.

본 연구에서는 멸균처리공정이 없는 돈육 포장공정을 대상으로 작업장에서 직접 분리한 야생균주인 Salmonellaspp. KSC101를 작업장의 온도와 시간을 주요 변수로 하여, 이들 현장에서의 Salmonella spp. KSC101의 성장 특성을 파악하고, 이를 수학적으로 예측할 수 있는 모델을 개발하였다. 돈육포장공장 현장을 반영하여 온도는 0, 5, 10,15, 20oC로, 시간은 0, 1, 2, 3시간으로 하였으며, 0oC와5oC에서는 성장이 발생하지 않았으며, 10oC, 15oC, 20oC에서는 약간의 성장이 있었으나 증가수준은 평균 0.34 logCFU/g정도였고, 20oC에서 성장율이 더 높았으나 15oC와는 통계적으로는 유의하지 않았다(p < 0.05). 하지만 PMP와 비교시 야생균주인 Salmonella spp. KSC101의 성장이더 빠른 것으로 나타났다. 이들 실험결과를 바탕으로 1차모델은 Gompertz 4 parameter식을, 2차 모델은 Exponentialdecay식을 이용하여 성장예측모델을 개발하였으며, R2값은0.99이상으로 나타났다. 개발된 모델에 대한 검증으로RMSE를 이용하였으며, 값이 0.103으로 양(+)의 방향으로약간 초과 예측하는 것으로 나타났으나, 최종적으로 실험값과 예측값이 ± 0.5 log cfu/g 내에서 일치하고 있어, 본연구에서 개발된 모델은 추후 냉장돈육 포장공정에서 위생관리기준 설정에 대한 과학적 근거자료로 활용할 수 있을 것이다.