Optimal selection of ANN training and architectural parameters using Taguchi method: A case study

Abstract

In recent years, there is increasing interest in using artificial neural networks (ANNs) for modeling and optimization of machining processes. The advantages that ANNs offer are numerous and are achievable only by developing an ANN model of high performance. However, determining suitable training and architectural parameters of an ANN still remains a difficult task. These parameters are typically determined in trial and error procedure, where a large number of ANN models are developed and compared to one another. This paper presents the application of Taguchi method for the optimization of ANN model trained by Levenberg-Marquardt algorithm. A case study of a modeling resultant cutting force in turning process is used to demonstrate implementation of the approach. The ANN training and architectural parameters were arranged in L18 orthogonal array and the predictive performance of the ANN model is evaluated using the proposed equation. Using the analysis of variance (ANOVA) and analysis of means (ANOM) optimal ANN parameter levels are identified. Taguchi optimized ANN model has been developed and has shown high prediction accuracy. Analyses and experiments have shown that the optimal ANN training and architectural parameters can be determined in a systematic way, thereby avoiding the lengthy trial and error procedure.

Keywords

• ann
• optimization
• taguchi method