Yoshiaki Nakagawa

Recent developments and industrial applications of data-based process monitoring and process control

Abstract Statistical process monitoring and control are now widely accepted in various industries. In recent years, statistical techniques are expected to solve quality-related problems. The issue of how to improve product quality and yield in a brief period of time becomes more critical in many industries where the product life cycle becomes shorter. Examples include steel processes and semiconductor processes. These processes are totally different in appearance, but the problems to solve are highly similar: how to build a reliable model from a limited data, how to analyze the model and optimize operating condition, and how to realize an on-line monitoring and control system and maintain it. In this paper, the problems and solutions are described with our application results in steel facilities.

Quality improvement of steel products by using multivariate data analysis

This paper describes quality improvement methods based on multivariate data analysis and their application to an industrial steel process. The PCA-LDA, which combines principal component analysis and linear discriminant analysis, is used for influential factor analysis of the qualitative quality variables. In addition, Data-Driven Quality Improvement (DDQI) is used to determine the optimal operating condition that can achieve the desired product quality under a given objective function and various constraints. The PCA-LDA and the DDQI can provide useful information to improve product quality. The experimental results show the effectiveness of the proposed methods.