Prediction of pilling grade of alkali-treated regenerated cellulosic fabric using fuzzy inference system

Fabric pilling is formed due to the entanglement of fibers during mechanical actions including abrasion, domestic washing process, and wearing cycles. Pilling resistance is one of the important aesthetic property of fabrics controlling the appearance of fabric quality. Pilling can be controlled by different methods; alkali treatment is one of the methods for the lyocell fabric. In this study, the prediction of pilling resistance on the alkali treated lyocell fabric has been done through a fuzzy inference system (FIS). In general, conventional pilling measurements has some hindrance due to human error and time consuming. This work utilizes the algorithm to provide accurate quantitative values of pilling grades. The correlation coefficient 'r' between predicted and experimental values shows 0.93 for TmAH and 0.81 for NaOH treatments. The forecast between experimental and predicted data sets of pilling grades were evaluated by mean absolute percentage error (MAPE) method and observed the values for NaOH, KOH, LiOH, and TmAH treatments as 16.6%, 18.87%, 21.09%, and 8.72% respectively.

Automated summary

This study predicted the pilling resistance of alkali treated lyocell fabric using a fuzzy inference system, with a high correlation coefficient between predicted and experimental values. The comparison of predicted and experimental data sets using the mean absolute percentage error method showed accurate predictions for various treatments.