Merging fuzzy logic, neural networks, and genetic computation in the design of a decision-support system

The main goal of evolutionary computation is to provide a near optimal tech nique between exploration and exploitation of a search space. This approach is based on a genetic "engine" that operates the search of the optimal sol ution via biological-based assumptions. Selection of the optimal maintenanc e interventions activity, that can be tackled with success thanks To an evo lutionary approach able to correct the distresses on the road pavement, is a very complex task. This paper presents an experimental architecture that improves the evolutionary aspect with additional benefits deriving from a s ynergistic combination of other powerful techniques, in particular neural n etworks and fuzzy logic. The best rules for managing pavement maintenance a ctivities, developed through a genetic selection, are judged by a neural ne twork. By an appropriate introduction of simple and efficient fuzzy identif iers, the features of the distress to treat can be described in an efficien t and natural way. We describe the main advantages arising from this hybrid approach discussing the applicability of the method with experimental resu lts. (C) 2000 John Wiley & Sons, Inc.