Modelling Air Permeability of Woven Fabrics by Artificial Neural Networks

Abstract

The objective of this paper is to investigate the predictability of air permeability of cotton woven fabrics from their construction variables by using a feed-forward back-propagation network in an artificial neural network (ANN) system. In order to achieve this objective, a number of grey cotton fabrics meant for shirting end use are desized, scoured, and relaxed. The fabrics are then conditioned and tested for constructional particulars and air permeability. A multiple linear regression based approach for modelling is attempted and the prediction issues are briefly summarized. For neural network modelling, a three layer feed-forward network is formed and trained by using the Broyden– Fletcher–Goldfarb–Shanno (BFGS) quasi-Newton algorithm. The predictive ability of the neural model is examined by comparing their results with experimental data. From the results, it is seen that high correlation exists between the actual and predicted values of the neural network model. The overall predictability of the model is good.

Citation

Behera, B.K. and Guruprasad, R. (2010), "Modelling Air Permeability of Woven Fabrics by Artificial Neural Networks", Research Journal of Textile and Apparel, Vol. 14 No. 3, pp. 77-84. https://doi.org/10.1108/RJTA-14-03-2010-B008