Damage detection for a cable-stayed Bridge under the effect of moving loads using Transmissibility and Artificial Neural Network

Abstract

Artificial Neural Network (ANN) has been widely used for Structural Health Monitoring (SHM) in the last decades. To detect damage in the structure, ANN often uses input data consisting of natural frequencies or mode shapes. However, this data is not sensitive enough to accurately identify minor structural defects. Therefore, in this study, we propose to use transmissibility to generate input data for the input layer of ANN. Transmissibility uses output signals exclusively to preserve structural dynamic properties and is sensitive to damage characteristics. To evaluate the efficiency of the proposed approach, a cable-stayed bridge with a wide variety of damage scenarios is employed. The results show that the combination of transmissibility and ANN not only accurately detect damages but also outperforms natural frequencies-based ANN in terms of accuracy and computational cost.

Keywords

• artificial neural network structural health monitoring transmissibility natural frequencies