On the application of independent component analysis to the problem of process monitoring through control charts

• Grant number: 11/14025-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2011
• End date: December 31, 2012
• Field of knowledge: Engineering - Electrical Engineering
• Principal Investigator: Leonardo Tomazeli Duarte
• Grantee: Guilherme Dean Pelegrina
• Host Institution: Faculdade de Ciências Aplicadas (FCA). Universidade Estadual de Campinas (UNICAMP). Limeira , SP, Brazil
• Associated scholarship(s): 11/23546-3 - Methods based on independent component analysis for monitoring industrial processes, BE.EP.IC

Abstract

Since its origin, which dates back to the 1920s, the tools of statistical process control (SPC) have been extensively applied to quality monitoring and improvement in industrial processes. Among the methods of SPC is the control chart, which was originally proposed by Shewhart, and whose goal is to verify, by analyzing a set of data, if the observed process variability stems from special causes, such as operator error or equipment failures. In many industrial processes, the number of parameters to be monitored is high, and usually such parameters are correlated. In these situations, it is interesting to make use of multivariate control charts, which basically aims at exploring in a joint fashion all the information brought from the control charts associated with each parameter. Recently, some studies showed that multivariate control charts based on a signal processing methodology known as independent component analysis (ICA) can provide better results in tasks such as fault diagnosis and classification of abnormal patterns when compared to established approaches. These results motivate the present research project, whose goal is to conduct a study on the applicability of the ICA in the analysis of multivariate control charts. In particular, we shall analyze the main results of the literature, in order to verify the viability of such proposals, and to obtain possible improvements. (AU)