PREDICTION AND INFLUENCE OF PROCESS PARAMETERS IN WIRE CUT ELECTRIC DISCHARGE MACHINING FOR HSS BY USING ARTIFICIAL NEURAL NETWORKS

M. JAYASREE
(M.TECH Student, Mechanical Engineering, JNTUACEA, ANANTHAPUR, A.P,) , IN

Dr.K.HEMACHANDRA REDDY
(M.TECH Student, Mechanical Engineering, JNTUACEA, ANANTHAPUR, A.P,) , IN

Dr.S.SIVANAGA MALLESWARA RAO
(M.TECH Student, Mechanical Engineering, JNTUACEA, ANANTHAPUR, A.P,) , IN

Keywords: WEDM, High Speed Steel, Artificial Neural Networks, MAPE and R

Abstract

performance characteristics like cutting speed, surface roughness, wire vibration and spark gap for different thickness ranging from 5mm to 80mm for machining 18-4-1 grade high speed steel (HSS) in wire electric discharge machining (WEDM). Experiments were performed at different levels of discharge current on different levels of plate’s thickness and experimental results of spark gap, cutting speed, surface roughness, and wire vibration were taken. In WEDM, there is a risk of breakage of wire that affects the overall efficiency of the process. To decrease the breakage of wire and to save the time with the lengthy calculation , Artificial neural network (ANN) are very useful for predicting the required targets. The selected inputs are thickness, current and the selected outputs are spark gap, cutting speed, wire vibration and surface roughness. By comparing the three algorithms Levenberg-Marquardt, Bayesian Regularization and Scaled Conjugated Gradient, the results obtained from Levenberg-Marquardt are more accurate than other two
algorithms when correlated with experimental results from the base paper using statistical tool such as MAPE and R2 . By
using L-M Algorithm the MSE value of best validation performance is 0.013364 at epoch 10 which is in acceptable limit

How to Cite

References

NA