Monitoring of continuous steel casting process based on independent component analysis

Multivariate statistical process control (MSPC) has been successfully applied to performance monitoring and fault diagnosis for industrial processes. However, traditional MSPC are based upon the assumption that the separated latent variables must be subject to normal probability distribution, which can not be satisfied for continuous steel casting process. In this paper, independent component analysis ( ICA) is introduced to model non-Gaussian data from continuous steel casting process and improve the monitoring performance of process, which can overcome the need of the data distribution. The basic idea of our approach is to use ICA to extract the essential independent components that drive a process and to combine them with process monitoring techniques. I², I_e² and SPE charts are proposed as monitoring charts. The application results show the advantages of ICA monitoring in comparison to principal component analysis(PCA) monitoring.