Tactile Comfort Prediction of Functional Fabrics from Instrumental Data Using Intelligence Systems

Subjective and objective evaluations of the handle of textile materials are very important to describe its tactile comfort for next-to-skin goods. In this paper, the applicability of artificial neural-network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) modeling approaches for the prediction of the psychological perceptions of functional fabrics from mechanical properties were investigated. Six distinct functional fabrics were evaluated using human subjects for their tactile score and total hand values (THV) using tactile and comfort-based fabric touch attributes. Then, the measurement of mechanical properties of the same set of samples using KES-FB was performed. The RMSE values for ANN and ANFIS predictions were 0.014 and 0.0122 and are extremely lower than the variations of the perception scores of 0.644 and 0.85 for ANN and ANFIS, respectively with fewer prediction errors. The observed results indicated that the predicted tactile score and THV are almost very close to the actual output obtained using the human judgment. Fabric objective measurement technology, therefore, provides reliable measurement approaches for functional fabric quality inspection, control, and design specification.