Multi-objective machining parameter optimisation of aluminium alloy 6063 by the Taguchi-artificial neural network/genetic algorithm approach

This study investigates the turning of aluminium alloy 6063 to optimise the material removal rate (MRR) and surface roughness (Ra) simultaneously. L27 Taguchi's orthogonal experiments were conducted by incorporating machining parameters of speed (260, 470, 840 rev/min), feed (0.2, 0.3, 0.4 mm/rev) and depth of cut (0.5, 1.0, 1.5). Analysis of variance (ANOVA) and signal-to-noise ratio were applied to determine the optimal control settings and validated by confirmatory tests. The performance characteristics were modelled by second-order regression, artificial neural network (ANN) and genetic algorithm (GA). The results indicate that the optimal conditions for MRR (375 mm3/min) and Ra (1.298 μm) were in agreement with the confirmatory tests. Regression models showed that the optimal points for MRR and Ra can be enhanced by the effect of interactions, but the ANN predicted the experimental data with better accuracy. The GA further elicited a set of optimal solutions for improving machining performance.