Analytical Study of Factors Affecting Yarn Coefficient of Mass Variation Estimated by Artificial Neural Networks

Authors

  • Manal R. Abdel-Hamied Textile Department-Faculty of Engineering, Alexandria University, Lotfy El-Sayed St., Alexandria, 21544, Egypt
  • Sherien ElKateb Textile Department-Faculty of Engineering, Alexandria University, Lotfy El-Sayed St., Alexandria, 21544, Egypt
  • Adel El-Geiheini Textile Department-Faculty of Engineering, Alexandria University, Lotfy El-Sayed St., Alexandria, 21544, Egypt

DOI:

https://doi.org/10.3993/jfbim00345

Keywords:

Yarn coefficient of mass variation;Image Processing;Artificial Neural Networks;ANOVA;Correlation

Abstract

Manufacturers aim to achieve the optimal quality, therefore, the evaluation of yarn parameters and the determination of factors that influence yarn quality is of great importance. The yarn coefficient of mass variation (CVm%) reflects the irregularity of the yarn which reflects the yarns' quality. This study investigates the parameters affecting the CVm% that was previously estimated using image processing and artificial neural networks. Yarn images and data were used as inputs into neural networks and CVm% was evaluated. In addition, two statistical methods were used which were: correlation and ANOVA to research the effect of yarn count, twist factor, blend ratio, and cotton type on CVm%. It was found that the yarn count and twist factor were the highest correlated parameters regarding CVm%.

Published

2023-06-16

Issue

Section

Articles