Parameter estimation in equivalent circuit analysis of dielectric cure monitoring signals using genetic algorithms.

This communication concerns the treatment of dielectric data obtained from experiments following the chemical hardening process (cure) in thermosetting resins. The aim is to follow, in real time, the evolution of the individual parameters of an equivalent electrical circuit that expresses the electrical behavior of a curing thermoset. The article presents a methodology for the sequential inversion of impedance spectra obtained in cure monitoring experiments. A new parameter estimation technique based on genetic algorithms is developed and tested using different objective functions. The influence of the objective functions on the modelling performance is investigated. The new technique models successfully spectra contaminated with high noise levels. The introduction of regularization in the optimization function rationalizes the effects of outliers usually detected in cure monitoring dielectric spectra. The technique was successfully applied to the analysis of a series of spectra obtained during the cure of an epoxy thermosetting resin.