ARTIFICIAL NEURAL NETWORKS AS A TOOL IN URBAN STORM DRAINAGE

The introduction of Artificial Neural Networks (ANNs) as a tool in the field of urban storm drainage is discussed. Besides some basic theory on the mechanics of ANNs and a general classification of the differen t types of ANNs, two ANN application examples are presented: The predi ction of runoff coefficients and the restoration of rainfall data. Fro m the results, it can be concluded that ANNs can deal with problems th at are traditionally difficult for conventional modelling techniques t o solve. Their advantages include good generalisation abilities, high fault tolerance, high execution speed, and the ability to adapt and le arn. However, ANNs rely strongly on the quantity of data examples, the ir training is occasionally slow, and they are not transparent and obs truct any closer analysis and interpretation of their performance. Fin ally, it is expected that the future of ANNs will lie in its integrati on with other conventional and more advanced modelling techniques, cre ating so-called hybrid models. (C) 1997 IAWQ. Published by Elsevier Sc ience Ltd.