Predicting the Performance of Air-Jet Weaving Machine in Relation to Filling Yarns’ Twist Loss Using (ANN) and Linear Regression Models

Document Type : Original research articles

Authors

1 Textile Research and Technology Institute, National Research Center, Dokki, Cairo, Egypt.

2 Assistant Professor, and Head of Garment Department-Sammanoud

Abstract

During their insertion on Air-Jet weaving machines, the leading end of the filling yarns lose some of their twist values. The loss of twist amount is around 10-15% of the nominal twist. This amount of lost twist will adversely affect the performance of this type of weaving machine and also deteriorates the physical properties of the produced fabrics. In this study, the amount of lost twists in the weaving yarns on Air-Jet weaving machines was predicted using two different methodologies, namely regression and Artificial Neural Networks (ANN). The predicted models were derived and compared for both methodologies. The efficiency of predicting power was also evaluated using the coefficient of determination (R2 value), mean bias error (MBE), and Root Mean Square Error (RMSE) for both models. It is evident that R2 values have high values and the remaining parameters have low values for the ANN technique in comparison with regression one. This means that the ANN technique is more efficient in predicting than regression.

Keywords

Main Subjects

Files

Share

How to cite

Statistics