Applicability of the modified back-propagation algorithm in tool condition monitoring for faster convergence

The search for an adequate strategy for on-line Tool Condition Monitoring (TCM) in an automated manufacturing environment is a distant goal, yet to be achieved. The present impetus is towards the application of neural networks with different learning schemes through the use of computers for faster processing. In this paper, the performance of the back-propagation neural network has been studied for various parameters. Moreover, the efficacy of a modified back-propagation algorithm for faster convergence has been evaluated for its applicability with a set of data on TCM, where reduction in computation time is very important. The results of the modified algorithm are quite encouraging for future applications in the area of on-line TCM.