Uncertainties on CMMs by Applying a Model of Corrections Based on a Global Metric, Monte Carlo and Neural Network Methods

AbstractThis paper is aimed at increasing knowledge about uncertainties of measurements on Coordinate Measuring Machines (CMMs) by applying a model of corrections based on combining a Global Metric procedure, a Monte Carlo statistical analysis and a Neural Network method that links and commands both items. In fact, this paper analyses general criteria for using a variant of Monte Carlo method (π-Monte Carlo) not only in propagating uncertainties, but also as methodology using numerical simulation for collecting necessary information for such propagation. Viewpoint presented here provides a model that stands out with contributions to uncertainty of CMM in use, making it a task-oriented calibration alternative methodology applied to CMM with special guidance to industry