Análise multivariada de capabilidade de processos baseada em características da qualidade com distintos graus de importância

Abstract

Capability analysis is an essential tool for evaluating the performance of a process, ensuring it meets customer requirements. When dealing with a manufacturing process involving multiple correlated variables, it is necessary to employ multivariate methods for analyzing quality characteristics. Various proposals for multivariate process capability indices (MPCIs) have been discussed in the literature, based on distinct approaches such as the proportion of nonconforming items or principal component analysis (PCA). In this context, a new method for capability analysis with multiple responses of varying degrees of importance, called Weighted Arithmetic Mean with Prioritization (WAMP), has been proposed. The results showed that WAMP is more robust than the methods Simple Arithmetic Mean (SAM) and Simple Geometric Mean (SGM), although the methods Weighted Geometric Mean (WGM) and Weighted Arithmetic Mean (WAM) also exhibited good behavior according to acceptance criteria. The proposed method was applied in two cases to assess its effectiveness. The first case involved applying the method to data collected from a turning process of ABNT 52100 steel. The second case involved using simulated data. When assigning a higher weight to a single original variable, the MPCIs tend to be similar to the univariate value of the most heavily weighted variable. When equal weights are assigned to two original variables, the MPCIs fall between the MPCs of the most heavily weighted variables. Additionally, it was observed that the performance of the MPCIs improves with increased correlation. In summary, WAMP proved to be a useful and efficient tool for capability analysis in systems with multiple correlated responses of varying degrees of importance.