A Neurogenetic Method for System-Identification

Models of real systems are of fundamental importance in virtually all disciplines. Models make possible to predict or simulate a system’s behavior. Since the quality of the model typically determines an upper bound on the quality of the final problem solution, modeling is often the bottleneck in the development of the whole system. As a consequence, a strong demand for advanced modeling and identification schemes arises. In this paper it is shown, via the development of a system for the analysis of the seismic response of rockfill dams, that neurogenetic techniques are a better option than analytically-based procedures in this kind of studies. To buttress this assertion, El Infiernillo dam, which has an ample history of being shaken by a great variety of seismic events, is used to this end. Key-Words: system-identification, modeling, neurogenetic networks, seismic response, soil-dynamic properties, finite element method, rockfill dams