Author : Aliwi Hussein Mohsin and Sadiq Muhsin Ihmood

In this research, a neural network was used to develop a model of EDM process and analysis of variance (ANOVA)
method was used to analyze the results. The pulse current(Ip), pulse time(Ton) and pulse of time (Toff) were selected as input to the neural network while material removed rate(MRR), tool wear rate(TWR) and surface roughness(Ra) represent the output of this neural network. Data obtained from practical experiments are used for training and testing the neural network. The effect of change in machining conditions on process performance can be tested and analyzed through the neural network model. Three models of neural networks were obtained. The first model predicts the rate of the removed metal, the second model predicts the wear rate of the tool and the third model predicts the roughness of the surface. The results indicate that the neural network model can predict the operation performance with reasonable accuracy for different machining conditions.

Keywords: Electro-discharge (EDM), Artificial neural network(ANN), Levenberg-maquardt (trainlm), ANOVA

Article published in International Journal of Current Engineering and Technology, Vol.7, No.4 (Aug-2017)