An ANN-GA based optimization of process parameters in wirecut electric discharge machining for aluminium, brass and mild steel

The demand for higher quality customized products, particularly for aerospace applications, has pushed the demand for intricate machining. Wirecut Electrical Discharge Machining (WEDM) has become popular for the precise machining of intricate geometries. The present work aims at improving environmental, quality, and production rate in WEDM by simultaneously optimizing Electricity Consumption (EC), Surface Roughness (SR) and Surface Generation Rate (SGR), respectively as in addition to quality, energy consumption and production rate are two other important parameters for industries to balance sustainable development and revenue generation, respectively. Machining experiments were performed by varying six input parameters – pulse-on time, pulse-off time, spark gap voltage, peak current, wire feed rate, and wire tension – by using Taguchi L 18 orthogonal array on the widely used industrial materials, namely aluminium, brass and mild steel. This paper addresses the multi-objective challenge by using Artificial Neural Network-Genetic Algorithm (ANN-GA). Feature importance and Pareto solutions were utilized to systematically assess and balance trade-offs between conflicting objectives. The obtained results demonstrated high prediction accuracy and robust parameter convergence. This study contributes to provide optimal machining parameters— I p , S v , T o n , T off , W f , W t for aluminium, brass and mild steel to obtain optimum EC, SR, and SGR results. The managerial significance lies in assisting practitioners in implementing optimised strategies, ensuring informed decision-making and optimal outcomes under complex scenarios thereby promoting sustainable and profitable manufacturing practices.