Combined Neural Network and PCA for Complicated Damage Detection of Bridge

In this paper, an efficient bridge damage detection algorithm is reported. The measured frequency response functions (FRF) is used as the input to artificial neural networks (ANN). Since full size of FRF data is too much for the ANN, a data reduction technique based on principal component analysis (PCA) is applied to extract the features. The extracted features are used as the input data of ANN instead of the raw FRF data. The self-organizing map neural network is chosen because of its superiority in analyzing high-dimensional data without supervising. A steel box girder model with multi damage states is presented to demonstrate the effectiveness of the method. The results showed that it is possible to distinguish the states with good accuracy.