MODELING AND OPTIMIZATION OF PROCESS PARAMETERS USING NEURAL NETWORKS AND SIMULATED ANNEALING ALGORITHM FOR ELECTRICAL DISCHARGE MACHINING OF AISI2312 HOT WORKED STEEL

Abstract

The present study addresses the multi-criteria modeling and optimization of Electrical Discharge Machining (EDM) for AISI2312 hot worked steel parts via optimized back propagation neural networks (OBPNN) and Simulated Annealing (SA) algorithm. The process response characteristics considered are surface roughness, tool wear rate and material removal rate.The process input parameters include voltage, peak current, pulse off time, pulse on time and duty factor. The weighted normalized grades, obtained from Taguchi design of experiments, are used to develop the arteficial neural network (ANN) model. In order to enhance the prediction capability of the proosed model, its architecture is tuned by simulated annealing algorithm. Next, the developed model is embaded into the SA algorithm to determine the best set of process parameters values for a desired set of outputs. Validation of the results has been carried out through a test run under the optimal machining conditions. Experimental results indicate that the proposed modeling and optimization procedures are quite efficient in modeling and optimization of EDM process parameters.