A modified principal component analysis-based utility theory approach for optimization of correlated responses of EDM process

Abstract

Electrical discharge machining (EDM) is one of the most extensively used  non-traditional machining processes having multiple performance characteristics, some of which are usually correlated. So, ideally, use of principal component analysis (PCA)-based approaches that take into account the possible correlations between the responses are suitable for optimization of EDM process. A recently reported study reveals that PCA-based proportion of quality loss reduction (PQLR) method results in the best optimization performance among the four considered PCA-based approaches for EDM process. This paper presents a modified  PCA-based utility theory (UT) approach for optimization of correlated responses. The reported experimental data on EDM processes in literature are analyzed using the modified PCA-based UT approach and PCA-based PQLR method. Comparison of the predicted performance measures at the optimal process conditions derived based on these two PCA-based approaches reveal that the modified PCA-based UT approach leads to better overall optimization performance. So, it can be the most promising approach for optimizing EDM processes.

Keywords: EDM; Correlated responses; Optimization; Principal component analysis; Proportion of quality loss reduction; Utility theory.