Multivariate Statistical Process Control Based on Principal Component Analysis: Implementation of Framework in R [chapter]

Ana Cristina Braga, Cláudia Barros, Pedro Delgado, Cristina Martins, Sandra Sousa, J. C. Velosa, Isabel Delgado, Paulo Sampaio
2018 Lecture Notes in Computer Science  
The interest in multivariate statistical process control (MSPC) has increased as the industrial processes have become more complex. This paper presents an industrial process involving a plastic part in which, due to the number of correlated variables, the inversion of the covariance matrix becomes impossible, and the classical MSPC cannot be used to identify physical aspects that explain the causes of variation or to increase the knowledge about the process behaviour. In order to solve this
more » ... lem, a Multivariate Statistical Process Control based on Principal Component Analysis (MSPC-PCA) approach was used and an R code was developed to implement it according some commercial software used for this purpose, namely the ProMV (c) 2016 from ProSensus, Inc. (www.prosensus.ca). Based on used dataset, it was possible to illustrated the principles of MSPC-PCA. This work intends to illustrated the implementation of MSPC-PCA in R step by step, to help the user community of R to be able to perform it.
doi:10.1007/978-3-319-95165-2_26 fatcat:bmsa6ndof5b2nbtpxj7hnvq2si