Author(s)

O. Giustolisi, M. Mastrorilli & F. Porcaro

Abstract

A connectionist model (using a Neural Network) applied to the transformation inflow-outflow phenomena of an experimental urban basin has been developed. It has been made a software, using a Matlab Toolbox, simulating a Standard Back Propagation Neural Network, which has been applied to the experimental data of Luzzi's basin. In a previous paper, this application gave encouraging results even if the authors had worked only with twenty five experimentally measured events of the inflow-outflow phenomena. In the present paper the authors present the results of the application of the Neural Network to the same inflow-outflow phenomena of Luzzi's basin, obtained working with fifty nine experimental inflow-outflow events. The authors improved the Neural Network efficiency using a var

Keywords