Using the FAST system to establish translation equations for the drape coefficient

Abstract

The regression and neural network methods have been used to develop new translation equations for calculating the drape coefficient of fabric using three fabric mechanical properties, namely bending rigidity, shear rigidity and formability, by the FAST system. Fabric mechanical properties show significant correlations with drape coefficient. Both regression models and neural network models have high correlation coefficients and low mean absolute values. It is observed that the neural network method has better ability of fitting than regression analysis.