L’application des réseaux neuronaux artificiels pour l’identification et la détection de l’endommagement dans les ponts

Abstract.This paper presents the development of an automatic monitoring method for the detection of structural damage. In this feasibility study, we have explored the use of self-organisation of artificial neural networks in structural damage assessment combined with the finite element method. The basic strategy is to train the network to recognise the behaviour of the structure with various possible damage states. When the trained network is subjected to the measurements of the structural response, it should be able to detect any existing damage. Three examples are examined and the results are promising.