Comparison between Levenberg Marquart, Bayesian and Scaled Conjugate Algorithm for Prediction of Cutting Forces in Face Milling Operation

Abstract

In machining operation, there is frequent tool wearing and breakage. Many a times, the root cause for early failure of tool is inappropriate cutting forces. In milling operation, cutting force is an intrinsic phenomenon which plays important role in machine tool conditioning. Prediction of the cutting force by traditional modelling methods becomes time consuming and costly. Artificial Neural Network (ANN) gives an easier way of predicting force model in case of machining operations and it is best technique to be used when the process is non-linear and as the milling operation is non-linear and time dependent. This technique gives good result in lesser time. In this paper three feed forward backpropagation algorithms are compared on the basis of six different criterion to decide the best suited algorithm for predicting the cutting force in face milling.