Selection of Manufacturing Method Using Artificial Neural Network

Abstract

The selection of a manufacturing method for developing new products with optimal quality, minimal cost in the shortest time possible is a important phase of the industry. This paper uses artificial neural network to facilitate for product manufacturing method selection. Initially, general sorting is employed to select an initial product platform. Then using repertory grids method, designers contribute importance ratings to the design options. These ratings are employed to reduce the number of the derived design options, and thereby used as input data to a neural network. The neural network is then trained by using Levenberg-Marquart Algorithm in Mat lab software. The trained neural network is applied to classify the set of options into different patterns. The classification results can subsequently serve as base for the screening of preferred manufacturing options.