Parameter Estimation of Air-Cooled Furnace for Metal Components using Random Forest

Ikuto Nakatsukasa, Yasuaki Ito, Koji Nakano, Victor Parque

Abstract


The cooling conditions during heat treatment, particularly in the normalizing process, have a critical impact on the final material properties and overall treatment outcomes. These conditions, if not precisely controlled, can lead to significant variations in product quality. Traditionally, industry experts have relied on historical data, trends, and their own experience to estimate the optimal cooling parameters. However, this approach can be subjective and prone to inaccuracies, especially for new or complex components. In this study, we aimed to address this challenge by developing a predictive model using random forest algorithms to estimate the ideal cooling conditions. The model takes into account various factors such as the chemical composition, weight, material type, and geometric shape of the metal components being treated. The results demonstrated that our model achieved a high level of predictive accuracy, with a coefficient of determination (R2) of approximately 0.94, indicating that the model is highly reliable for practical applications in this field.

Keywords


parameter estimation; air-cooled furnace; machine learning; random forest

Full Text:

PDF

Refbacks

  • There are currently no refbacks.