Fuzzy Logic Model to Predict the Color Properties of Pineapple Leaf Fiber Fabric

Abstract

This study aimed to fabricate a Fuzzy Logic (FL) model based on dye, salt, and alkali concentrations to predict the color properties of pineapple leaf fiber (PALF) fabric dyed with reactive dyes. The nonlinear and interactive nature of these factors makes it challenging to make an exact model using mathematical or statistical methods. Additionally, artificial neural networks and neuro-fuzzy models require extensive trial data for parameter optimization, which is often difficult to obtain from dyeing industries. In this context, FL is an effective modeling tool, capable of handling nonlinear complex domains with minimal trial data. The test data confirmed the model established in this study, showing a mean absolute error (MAE) for all parameters within the acceptable range (<5%) and coefficients of determination (R²) exceeding 0.96. It is concluded that the FL model is efficient in the prediction of color properties of PALF fabric in nonlinear complex dyeing with high accuracy.

Journal of Engineering Science 15(2), 2024, 13-27