Prediction of clothing comfort sensation with different activities based on fuzzy comprehensive evaluation of variable weight

Abstract

There is no universal model for evaluating the clothing comfort sensation under various conditions due to the different effects of the sensory factors for different activities or environments. The current study aimed to develop a prediction method of the clothing comfort sensation for different activities. Two models were built using fuzzy comprehensive evaluation based on the constant weight (FCE-CW) and variable weight (FCE-VW). Firstly, four sensory factors (wetness, stickiness, hotness, roughness) were selected as input data. Subsequently, grey relation analysis was introduced to obtain the constant weight during the complete experimental process and variable weight for different activities. Finally, the FCE-CW and FCE-VW models were built. A psychological wearing evaluation experiment with different activities was conducted to examine the validity of the model. The results showed that the determination coefficient ( R2) of the FCE-VW model was above 85%, which was superior to that of the FCE-CW model, especially in the running and recovery phases. The high accuracy of the FCE-VW model indicated that grey relation analysis was a practical algorithm for determining the weight of sensory factors. The roughness and hotness sensations demonstrated higher weights during the resting and walking phases, respectively, whereas the wetness and stickiness sensations had a high weight during the running and recovery phases. Moreover, the associations between the physiological parameters and weights of the sensory factors were also explored. In the future, a more comprehensive model based on physiological and psychological parameters can be developed using this method.

Research on Clothing Color Classification Method based on Improved FCM Clustering Algorithm

“…Fuzzy cluster analysis is a clustering analysis method constructed on the basis of fuzzy theory. Its correlation analysis of data is performed on the basis of considering regional characteristics and classifying them [18]. In pattern recognition and image segmentation, this method can be used to describe fuzzy objects more objectively.…”