An adaptive neuro-fuzzy and NSGA-II-based hybrid approach for modelling and multi-objective optimization of WEDM quality characteristics during machining titanium alloy

In this paper, an intelligent approach of adaptive neuro-fuzzy inference system (ANFIS) and non-dominated sorting genetic algorithm-II (NSGA-II) was delineated to establish model and optimize the wire electrical discharge machining (WEDM) process. The WEDM experiments were designed utilizing Taguchi L18 mixed orthogonal array for machining of Ti-6Al-4V titanium alloy. The ANFIS model was delineated to explain the influence of input machining characteristics, viz. peak current (IP), pulse on-time (T-on), pulse off-time (T-off) and wire feed (WF) on output response of material removal rate (MRR) and wire wear ratio (WWR). The proximity of results with confirmation experimental results revealed the effectiveness of the developed ANFIS model in prediction of output quality characteristics for the chosen input machining factors. The artificial neural network (ANN) and NSGA-II were integrated and applied for multi-objective optimization in determining optimal WEDM machining process conditions. The optimal results obtained with NSGA-II for multi-objective optimization of input control variables are within tolerance limits and realized an improvement in MRR and WWR (maximum absolute percentage errors as 6.784% and 7.589%) with optimal machining characteristics.