Application of BP Neural Network for Predicting Anode Accuracy in ECM

It is difficult for numerical method to predict the anode accuracy in electrochemical machining (ECM) with an uneven interelectrode gap, so this paper introduces forward feed forward back propagation (BP) neural network to solve this problem. Based on analyzing effect of parameters including workpiece, electrolyte and cathode on machined accuracy, meanwhile considering the practical machining condition, the neurons of BP neural network in the input layer are confirmed. The trial and error procedure was employed to optimize the number of neurons in the hidden layer. The architecture of BP neural network is constructed to ensure the minimum total prediction error. Levenberg Marquadt (LM) algorithm is used to train this network. To verify the validity of the trained network, results obtained by BP neural network are compared with that obtained by the experiments. It shows that the former is close to the later, the maximum prediction error is lower than 10%, which indicates that it is feasible to apply BP neural network to predict anode accuracy.