Empirical Bayes Monitoring for Univariate and Multivariate Processes and Other Techniques

Autor: Manuel Alonso Rodríguez Morachis, Eduardo Rafael Poblano-Ojinaga, Luz Elena Terrazas Mata, Manuel Iván Rodríguez, Manuel Arnoldo Rodriguez Medina
Rok vydání: 2021
Předmět:
Zdroj: Techniques, Tools and Methodologies Applied to Quality Assurance in Manufacturing ISBN: 9783030693138
DOI: 10.1007/978-3-030-69314-5_4
Popis: The purpose of this document is to use the basic concepts given in the celebrated Kalman filter, which can be derived using a Bayesian approach. Such an approach is implemented through Bayesian empirical monitoring for process control. The purpose of the Kalman filter is to obtain real-time estimates of the results of critical process variables, subject to process variations and noise conditions, such as environmental and measurement conditions. This document analyzes both the univariate and multivariate case of Bayesian empirical monitoring. The first application is analyzed to data taken from a molding process of a critical quality characteristic of an automotive sensor. For the multivariate case, measures of the characteristics to be controlled of a molded part were taken. The three-dimensional behavior was analyzed, first, by means of Bayesian empirical monitoring and then illustrating tests of multivariate normality and the management of Hotelling's square T.
Databáze: OpenAIRE