Prediction of single jersey plain cotton knitted fabric width using ANFIS.

Abstract

Knitted fabric structures have considerable advantages. over woven Fabric structure normally high elasticity, flexibility, easy-care property, better air permeability etc. There is however a number of problems with single jersey knitted structures encountered at the stages of knitting- make up and of end use. The major problem is high rate or fabric width shrinkage which causes dimensional. instability during the usage of the fabric. Achieving required fabric width with an acceptable shrinkage value is, always the ultimate. target of' a knitted fabric manufacture. This presented study was undertaken to develop an adaptive neuro-fizzy model for the prediction of width of single Jersey plain cotton knitted fabric. Creating this model helps knit. fabric manufacturers in optimizing manufacturing processes to control knitted fabric width thus improving dimensional stability. Multiple linear regression models as well as ANN have been, used in the past for the prediction of finished width of the single jersey cotton knitted fabric from the input machine. and knitting parameters. Prediction by ANN was found to be more accurate than those obtained from multiple linear regression models. The focus of this research was to develop a more reliable model to predict the fabric width of 100% single. jersey plain cotton knitted fabric in wet relaxed state. Adaptive neural fuzzy interference system was used develop efficient model to predict the fabric width. Yarn count and stitch length were considered for input parameters.38 fabric samples knitted with different stitch lengths and yarn counts were considered in developing the model. Model creation was done using fizzy logic tool box of the Matrix Laboratory (MatlabR2010a) software. Microsoft excel was used for drawing graphs (data analysis). The model was successfully created and validated. From the summary of goodness, it can be concluded that ANFIS performed better than its linear regression counterpart as it had better R2=0.95 and RMSE=I .912 compared to linear regression which gave R2=0.89 and RMSE=2.765. Thus this model can be comfortably used by knit Fabric manufacturers of 100% cotton knitted Fabrics.