Resumen
This research integrates the Taguchi method, analysis of variables (ANOVA), back-propagation neural networks (BPNN), and hybrid PSO-GA to develop an intelligent optimization system of micro electroforming process for the mesh filter. From the outset of discussions with engineers in terms of past related literature survey of the micro electroforming process, the quality characteristics of product and control variables can be well ascertained, then transforming the problem of multiple quality characteristics into a single quality characteristic via the Taguchi method and ANOVA. However, the optimal parameter settings (solution) through the Taguchi experimental planning is still belong to a discrete optimal solution which is impossible to meet the process stability and quality goals. Therefore, this study first identifies the initial weight of BPNN,using hybrid PSO-GA with multilayer perceptron (MLP),in order to improve training efficiency and precision of BPNN. Furthermore, the study constructs the signal-to-noise (S/N) ratios (BPNNS/N) and quality predictors (BPNNQ) based on hybrid PSO-GA and BPNN with the experimental data. The optimal parameter settings are obtained through a combination of BPNNS/N and BPNNQ with modified PSO-GA. Finally, confirmation experiments are performed to assess the effectiveness of the proposed system. The results show that the proposed system can create the best performance, and the optimal parameters not only enhance the stability in the micro electro forming process but also effectively improve the product quality.