The Application of Independent Component Analysis in Process Monitoring

A method of using Independent Component Analysis (ICA) to monitor process is proposed in this paper. The concept, model and algorithm of ICA are introduced first, and then ICA applications in the multivariable statistical process monitoring are discussed. A simulation example is used to demonstrate this approach. The results show that the independent component analysis can be considered as a new effective tool providing for statistical process monitoring and fault diagnosis.