Monitoring of the Infante D. Henrique Bridge with self organizing maps

Abstract

The early detection of some possible structural damage, requires the implementation of appropriate structural health monitoring systems, including efficient computational tools to extract the most relevant structural features characterizing the structural performance. In this context, this paper describes the application of Self Organizing Maps SOM, a type of neural network, to the processing of monitoring data of Infante D. Henrique Bridge, over Douro River in Porto, Portugal. This bridge has been monitored since 2007 (Magalhães, 2010) and the resulting database has been used as a source of information for training and testing the neural networks. In order to test the performance of SOM, four damaged models were simulated with the reduction of inertia in some elements. The results showed a good performance of the SOM, as the implemented method was able to detect relatively small structural changes. (c) 2016 Taylor & Francis Group, London.