Bayesian approach to measurement scheme analysis in multistation machining systems
Different measurement schemes in multistation machining systems carry different amounts of information about the root causes of dimensional machining errors. The choice of a measurement strategy in a multistation machining system is therefore crucial for subsequent successful identification of the machining error root causes. Recent advances in the linear state-space modelling of dimensional errors in multistation machining processes facilitate a formal and systematic characterization of measurement schemes. In this paper, the stream-of-variation methodology is employed to characterize various measurement schemes quantitatively in multistation machining systems using the Bayesian approach in statistics. Application of these methods is demonstrated in the characterization of measurement schemes in the machining process used for machining of an automotive cylinder head.