The prediction of fiber diameter of melt blowing by means of intelligent techniques and mathematical-physical methods

This work gives a comparative study of two modeling methods to predict the fiber diameter of melt blowing nonwovens from the processing parameters (polymer flow rate, polymer melt temperature, initial air velocity and die-to-collector distance). After measuring fiber diameter, evaluations of data were performed by using ANN model and mathematical-physical model. On the basis of the results obtained, with the aid of both ANN and linear regression analysis. The results demonstrated that artificial neural network is more accurate and reliable than the mathematical-physical model, which is a useful alternative approach for the development of melt blowing processing prediction models.