Usage of artificial neural network and quantum-inspired evolutionary algorithm for optimization of the wire EDM cutting of mild steel

In the current study, the impact of mild steel kerf width, material removal rate (MRR), and surface roughness of wire Electrical Discharge Machining (EDM) cutting process parameters are examined. This process is modelled using an ANN, and correlations between the process variables, as well as responses are created. On kerf width and MRR, a little impact of pulse-on time and gap voltage is discovered. However, the kerf width and MRR are getting smaller as the pulse-off time increases. With a rise in wire feed rate, it is discovered that kerf width and MRR increase. By the rise in wire feed rate, GV, and TON duration, the surface roughness also increases. An inverse relation is found between pulse-off time and surface roughness. Quantum-inspired Evolutionary Algorithm (QEA) is applied to optimize the process in MATLAB software. The optimum range of TON, TOFF, WF, and GVwere found 25-27 µs, 25-26 µs, 5-6 mm/min and 63-65 Volt respectively considering the machine constraints.