Indian Journal of Fibre &
Vol. 31, September 2006, pp. 401-408
B K Behera & S B Muttagi
Received 15 June 2005; revised received 9 September 2005; accepted 17 October 2005
A complete engineering design of polyester-viscose blended suiting fabrics has been presented using radial basis function neural network algorithm. Fabric low-stress mechanical properties, such as extension, bending rigidity, shear rigidity, breaking strength have been predicted from the structural parameters of the fabric such as weave, yarn tex, thread density, crimp, fabric mass and fabric cover. It is observed that the radial basis function neural network could successfully predict the trends in variation of fabric property with corresponding change in structural parameters.
Keywords: Engineering design, Neural network, Polyester-viscose blend, Prediction error, Radial basis function
IPC Code: Int. Cl.8 G06N3/02