A Theoretical Approach for Prediction of Yarn Strength in Textile Industry

Abstract This paper describes advanced techniques that can be used to predict ring spun yarn strength from HVI/FMT measured properties of cotton fibers such as span length, bundle strength, fineness, uniformity ratio and maturity of cotton fibers. Neural networks, neuro-fuzzy, multiple-linear regression techniques and Sitra’s expressions were used to predict yarn strength from cotton fibers properties. The results of intelligence system were better than multiple-linear regression techniques and Sitra’s expressions. The best results were obtained by neuro-fuzzy method.