The main purpose of this paper is to develop a system to predict the amount of cloudiness and non-homogeneous appearance of the circular knitted fabric, by utilizing the various cotton yarn quality characteristics which produced that fabric. A novel approach was developed using Artificial Neural Network (ANN) and satisfied results were obtained. A learning phase for the system was initially conducted on circular knitting machine fitted with an image processing system to capture an on-line image of the weft knitted fabric which was produced using one yarn cone (the machine was modified and adapted to use one cone instead of 24 yarn cones) in order to eliminate yarn cone to cone variation and to differentiate between 14 different typesof yarn from 14 different producers, All kinds of physical tests on yarn quality characteristics were conducted on the yarns in order to be used as a comprehensive input set of data for the prediction of the appearance and cloudiness of knitted fabric. The developed (ANN) System was successfully capable of predicting the appearance and cloudiness with satisfied agreement with actual results.
)
View on SCiNiTO