A neutral-network approach for the on-line monitoring of the electrical discharge machining process

In this paper. a feed-forward neural network is used to on-line monitor electrical discharge machining (EDM) processes. The relationships between tool-workpiece gap signals and various pulse types are established, based on the neural network through the back-propagation learning algorithm. As a result, the developed neural network can be used to monitor varying machining conditions in EDM. Experimental results have shown the effectiveness of this approach. (C) 1997 Published by Elsevier Science S.A.