Study of melting temperature behavior of polymer nanocomposites using fuzzy logic-based approach of artificial intelligence

Abstract

This work presents a fuzzy logic-based artificial intelligence approach for predicting the melting temperature of polymer nanocomposites based on polylactic acid and carbon nanotubes (CNTs). A Mamdani-type fuzzy inference model was developed using the degree of crystallinity, carbon nanotube concentration, and nanotube diameter as input parameters. The constructed model reproduced nonlinear relationships between the structural characteristics and the thermal behavior of the nanocomposites and demonstrated good agreement with experimental calorimetric data. The resulting response surfaces revealed the existence of an optimal CNT concentration range associated with the maximum nucleating effect of the nanotubes. The predictive capability of the model was confirmed by an adjusted coefficient of determination R2 = 0.86, indicating the applicability of fuzzy logic methods for intelligent modeling of polymer nanocomposite systems.