A Hybrid Methodology of ANN-NSGA-II for Optimization of the Process Parameters of Slotted-Electrical Discharge Abrasive Grinding Process

Abstract

Slotted-Electrical Discharge Abrasive Grinding (S-EDAG) is a new innovative hybrid machining process, which is developed by comprising the Electrical Discharge Grinding (EDG) and Abrasive Grinding (AG) where both the processes occur alternatively. The selection of the suitable combination of process parameters is much difficult due to non-linear relation between input and output. The aim of this study is to apply a hybrid approach of Artificial Neural Network (ANN) and Non-Dominated Sorting Genetic Algorithm (NSGA-II) to optimize the process parameters of S-EDAG process during machining of Al/SiC/B4C composite. ANN architecture with back propagation algorithm has been used for modeling, which is trained and tested with experimental data. The experimental data have been collected after exhaustive experimentation considering the effect of current, pulse duration, pulse interval, wheel RPM and abrasive grit number on material removal rate (MRR) and average surface roughness (Ra). The developed ANN model has been coupled with NSGA-II to generate the population and evaluate the objective functions during optimization. The results have been shown that hybrid approach of ANN-NSGA-II provides set of optimal solution that helps to manufacturer to select the appropriate combination of input parameters and reduce the product manufacturing cost.