Jorge M. Rosa and Ana M. F. Fileti
Artificial neural network, Engineering, Dyeing of cotton, Reactive dyestuff
Considerable attention has been given to process of cotton dyeing in order to minimize the presence of colored compounds in effluents generated from textile industries. This work proposes the development of an empirical model of dyeing of cotton made with Reactive Black 5 dyestuff (RB5), by using a perceptron multilayer neural network. Aiming at better colorist intensity and minimizing the salts in the effluents, the following values of best condition of dyeing were found by the neural model developed: 60°C; 97,0 g L-1 of NaCl; 14,5 g L-1 of Na2CO3; 2,5 mL L-1 of NaOH; 90 min of process and 4,8% of RB5.