Engineering design of polyester-viscose blended suiting fabrics using radial basis function network: Part II—Prediction of fabric constructional parameters from its properties

Abstract

The application of artificial neural network approach to re-engineer the design of woven polyester-viscose blended suiting fabric to be used by the weavers has been described. The fabric constructional parameter have been predicted for specific fabric property requirements using the same network with an approach called reverse engineering. It is observed that the radial basis function neural network could successfully predict the trends in variation of fabric constructional parameters. Evaluation of the model for each fabric property specification shows good agreement between predicted and generally accepted fabric and yarn structure-property relationships.