STATISTICAL QUALITY CONTROL OF THE PROCESSES IN MULTIVARIATE AUTOCORRELACIONADOS. AN APPLICATION IN INDUSTRY CUCHILLERA ALBACETE.Author:
ALFARO NAVARRO JOSE LUIS.
Year: 2005.
University:
CASTILLA-LA MANCHA.
Place of defense: FAC.DE CIEN. ECONO.Y EMPRESARIALES.
Place of preparation: FACULTAD CIENCIAS ECONOMICAS Y EMPRESARIALES.
Summary: This research work developed two alternatives to the traditional control charts, a robust alternative that improves behavior in phase I and a new proposal for the process control autocorrelacionados following a lens through multivariate modeling space in the states. So, after exposure of the main control charts classics (from an optical univariate and multivariate), addresses the problems associated with the need for iterative calculations in adjusting the control charts during Stage I, as well as related the failure of the assumption of independence (Phase II). As a contribution, raises the use of robust measures called 'middle truncated' to accelerate the adjustment and estimation of the parameters of the graph control in phase I. It is also proposed the implementation of the modeling space in the states and subsequent use of a graphical control for innovations that collects large variations and short-term temporary and graphic for the states that incorporates changes in the overall performance of the process , minor variations of intensity and longer temporary. Finally, to analyze the value of these methods is its application on the industry cuchillero Albacete, a sector of great importance for the economy Albacete.