Introduction to the theory of neural computation
Introduction to the theory of neural computation
Hi-index | 0.01 |
In this study, the approach of using neural networks is implemented for demonstrating its effectiveness in the real-time application of microelectroform on the different geometry forms. Three back-propagation neural networks are established via the training process with the numerical database to predict the distributions of Sh/Shmax, Cf/Cfmax and I/Imax. Comparisons of the predictions with the test target vectors indicate that the averaged root-meansquared errors from three back-propagation neural networks are well within 4.15%. The trained neural networks can verify the prediction capability for agent technology. Then, to fabricate the microstructure of higher surface accurate, higher hardness, lower residual stress and can be duplicated perfectly. Nevertheless, the instant knowledge of micro-electrforming characteristics is practically needed for many industrial agents technology applications.