Predicting air permeability of pile loop knit fabrics using fuzzy logic with type-2 fuzzy inference system

Abstract

Recent studies have shown that pile loop knit fabrics hold promise for being used in tissue engineering in addition to their usual application areas (e.g., sportswear). Understanding the air permeability property of fabrics would be of importance in terms of time and cost when designing new fabrics. This study develops a Fuzzy Logic (FL) model with type-2 fuzzy inference system for predicting the air permeability of pile loop knit fabrics. For this purpose, pile loop knit structures with different areal densities were produced by using textured polyethylene terephthalate (PET) yarns from four different filament fineness. FL model with type-2 sytem analysis was performed. The root mean square error (RMSE) of the developed model was compared with those of the Multiple linear regression (MLR), FL model with type-1 system, and Artificial Neural Network (ANN) model from the previous study. The RMSE of the MLR, FL model with type-1, and type-2 sytems, and ANN were found to be 14.93, 12.41, 11.58, and 2.42 respectively. Thus, the FL model with type-2 system gave less RMSE in comparison to the MLR, and FL model with type-1 system. However, the ANN model provided superior performance over the MLR and FL models in predicting air permeability.