Using adaptive neuro-fuzzy and genetic algorithm for simultaneously estimating the dye and AgNP concentrations of treated silk fabrics with nanosilver

Abstract

Purpose

Artificial intelligence (AI) methods, such as genetic algorithm (GA) and adaptive neuro-fuzzy inference system (ANFIS), are capable of providing superior solutions for the simulation and the modeling of complex problems. The purpose of this study is to estimate the dye and the silver nanoparticle (AgNP) concentrations of silver nanoparticle-treated silk fabrics by the aforementioned methods.

Design/methodology/approach

In this study, the color and the antimicrobial properties of silver nanoparticle-treated silk fabrics were matched by using the GA technique based on spectrophotometric color matching. The ANFIS method was also used; this method is based on the grid partitioning algorithm across four different methods. The first and second methods are provided for dye concentration prediction, and the third and the fourth methods are given for AgNP concentration prediction.

Findings

The mean of absolute error and root mean square (RMS) of the best dye concentration prediction by the ANFIS method based on the second method are 0.087 and 0.103, respectively. In addition, the mean of the absolute error and the RMS of the best results for AgNP concentration prediction by the ANFIS method by using the third method is 0.002 and 0.003, respectively. The obtained results indicate that the performance of the ANFIS method is better than the GA method.

Originality value

The simultaneous prediction of the color and the antimicrobial properties of silver nanoparticle-treated silk fabrics was performed by using the GA and the ANFIS. The suggested method led to acceptable accuracy for color and antibacterial matching.

Citation

Shams Nateri, A., Hasanlou, E. and Hajipour, A. (2019), "Using adaptive neuro-fuzzy and genetic algorithm for simultaneously estimating the dye and AgNP concentrations of treated silk fabrics with nanosilver", Pigment & Resin Technology, Vol. 48 No. 1, pp. 20-28. https://doi.org/10.1108/PRT-11-2017-0096