Armor Damage Analysis Using Neural Networks

Autor: Joseph R Medina, M Esther Gomez-Martin, Joaquin Garrido, César Vidal
Rok vydání: 2004
Předmět:
Zdroj: Scopus-Elsevier
DOI: 10.1061/40733(147)20
Popis: This paper compares to methods to estimate the rubble-mound breakwater armor damage evolution in nonstationary wave conditions. The first method is based on the exponential model on individual waves proposed by Medina (1996). The second method is based on Neural Networks pruned using Evolutionary Strategies. Both methods give reasonably good agreement with laboratory observations.
Databáze: OpenAIRE