Multi Response Optimization of WEDM Process on OHNS Die Steel Using ANN, SA and GA

In the present research work Oil Hardened Naturally Shrinking (OHNS) work-material which is commonly used in plastic industries is considered for machining by WEDM process. Four different control parameters are deliberated to study the effect on the responses like material removal rate, overcut and surface roughness. To reduce the total number of experiment, L27 orthogonal array is used. Analysis of Variance is applied to attain the significant process parameters affecting the responses. The effect of the responses with the control parameters is plotted through S/N ratio graphs. To find the effect of the parameters on the responses and thereby developing a mathematical model regression analysis is done. The response data are examined using artificial neural network. Single objective parametric combination for each response is obtained using simulated annealing. A multi response optimization for the responses is done initially by using genetic algorithm and finally by applying Grey relational analysis.