Jacobian algorithm on Features Selection and Evaluation of CHES-FY System

摘要

the main content dealt with in the paper was to adopt fuzzy algorithm to select features for characterizing multiple physical properties, which were acquired from the pulling-out force and distance curve by a new apparatus. It was a comprehensive handle evaluation system for fabrics and yarns (CHES-FY), which can reflect weight, bending, friction and tension properties as well as stiffness tactile. The paper presented the structure and principle of the CHES-FY system, and the Jacobian algorithm and Karhunen-Loeve (K-L) transformation were used to obtain several features selected from the pulling-out force and distance curve for the experimental fabrics. These features were proved to be effective in characterizing weight, bending, friction and tensile properties of fabric by comparisons with measured results by standard testers. There existed good correlations in utilizing these features selected by fuzzy algorithm to evaluate physical behaviour of fabrics. It indicated that it was greatly helpful for fuzzy method in the application of textile fabrics fields.