Application of ANFIS for Prediction Micro Holes in EDM

Abstract

The present work demonstrates the optimization process of Micro- hole of Electrical Discharge Machining (EDM) by Adaptive Neuro Fuzzy Inference System (ANFIS). The workpiece material was copper alloy. The current, gap distance and pulse on time were the control parameters of EDM. The process has been successfully modeled using ANFIS model constructs a fuzzy inference system in MATLAB 7.2 Software Gaussian type for optimization of micro diameter, were adopted during the training and testing process of ANFIS model in order to compare the prediction accuracy of micro diameter by one membership function. Finally, the comparison of ANFIS results with experimental data indicates that adoption of Gaussian membership function in proposed system achieved satisfactory accuracy. Prediction using ANFIS model compared with experimental values of micro holes at correspond ratio 98.37%.