A Study on the Prediction of Characteristics of Molding Sand Using Machine Learning and Data Preprocessing Techniques
| العنوان: | A Study on the Prediction of Characteristics of Molding Sand Using Machine Learning and Data Preprocessing Techniques |
|---|---|
| المؤلفون: | Jeong-Min Lee, Moon-Jo Kim, Kyeong-Hwan Choe, DongEung Kim |
| المصدر: | Korean Journal of Metals and Materials. 61:18-27 |
| بيانات النشر: | The Korean Institute of Metals and Materials, 2023. |
| سنة النشر: | 2023 |
| مصطلحات موضوعية: | Modeling and Simulation, Metals and Alloys, Surfaces, Coatings and Films, Electronic, Optical and Magnetic Materials |
| الوصف: | The main components of molding sand used in sand casting are sand, clay and water. The composition of the molding sand has a great influence on the properties of the casting. In order to obtain high-quality castings, it is important to manage the components that affect the properties of the molding sand such as especially green compression strength and compactability. In this work, green compression strength and compactability are predicted through a machine learning technique using the processing data obtained from a foundry and the properties of molding sand as the input variables. Through the correlation analysis between the input variables and the response variable, we investigated the relationship between the processing data and the properties of the molding sand. The possibility of predicting the characteristics of molding sand with high accuracy was confirmed using a model created through data preprocessing with the real foundry data. If the composition of the molding sand is adjusted in the foundry using the generated model, it is expected that higher quality castings can be produced and the productivity can be increased. |
| تدمد: | 2288-8241 1738-8228 |
| DOI: | 10.3365/kjmm.2023.61.1.18 |
| URL الوصول: | https://explore.openaire.eu/search/publication?articleId=doi_________::8545dae41bad7a2960386f3e5dbbda0f https://doi.org/10.3365/kjmm.2023.61.1.18 |
| Rights: | OPEN |
| رقم الانضمام: | edsair.doi...........8545dae41bad7a2960386f3e5dbbda0f |
| قاعدة البيانات: | OpenAIRE |
| تدمد: | 22888241 17388228 |
|---|---|
| DOI: | 10.3365/kjmm.2023.61.1.18 |