Prediction of single yarn tenacity of ring-and rotor-spun yarns from HVI results using artificial neural networks

Artificial neural network (ANN) models for predicting the single yarn tenacity of ring- and rotor- spun yarns form the cotton fibre properties, measured by high volume instrument, have been presented. Seven cotton fibre properties and yarn fineness have been used as the inputs to the neural network. Different network structures have been used to optimize the prediction performance. The relative importance of all the cotton fibre properties has also been quantified. The ANN models could predict the single yarn tenacity with less than 5% and 2% mean error in case of ring- and rotor- spun yarns respectively. Yarn fineness, fibre bundle tenacity, elongation and length uniformity are the dominant input parameters which influence the single yam tenacity of spun yarns.