Predicting Circular Knitting Machine Parameters for Cotton Plain Fabrics Using Conventional and Neuro-Fuzzy Methods

Abstract

Although there are many studies related to predicting the dimensions of knitted fabrics, there are no studies available about predicting the knitting machine used to produce the knit. Until now, the decision to select a knitting machine was made through experience, trial and error, or rough calculations. This study predicts the machine gauge and diameter more accurately for cotton plain knitted fabrics. Both conventional (multiple linear regression) and neuro-fuzzy methods are used for the predictions. High correlation coefficients are obtained with multiple linear regression analysis. A neuro-fuzzy method using an adaptive network based fuzzy inference system (ANFIS) is also used for predic tions, and the results of both systems are compared.