Prediction the Effect of Tensile and Elongation Properties of Dry and Wet Cotton Fibre on the Fabric Wrinkle Grade Based on Polynomial Type2 Fuzzy Neural Network Modelling

Abstract

In this article, we intend to use a novel tool in computational intelligence to model the effect of the cotton fibre tensile properties on the fabric wrinkle appearance. For this purpose, two groups of cotton fibres were provided in dry and wet conditions. In both groups, the linear density, tensile strength and breaking elongation of the fibres were considered as important tensile parameters. These parameters were measured in the laboratory for various cotton fibres. The wrinkle degree of the fabric samples manufactured by these fibres was ratted objectively, by the method based on the narrow optical technique. Then, were developed three fuzzy neural network modelling as type1, type2 and polynomial type2. In mentioned models, the cotton fibre linear density, tensile strength and breaking elongation parameters of the fibres were the input parameters, and the wrinkle degree of cotton fabrics was considered as output. The experimental results show the tensile properties of the cotton fibres are increased with the wetting process. Based on the modelling, there is a nonlinear relation between output and input parameters. Second, the wrinkle grade parameter was affected due to the wetting of cotton fibres. According to this results, the wrinkle grade of the fabrics is related to the fibre linear density and tensile parameters. In modeling, three models based on the type 1, type 2 and type 2-nomial fuzzy neural network were developed. Comparisons of the modelling and experimental results show the type2-nominal fuzzy model with accuracy greater than 98.8% and error value smaller than of 0.08% can predict more accuracy amongst three developed models.