Research on Damage Identification Method for Continuous Beam Bridges Based on SDAE
The identification of bridge damage based on dynamic detection often uses methods such as BP neural network and SVM.These methods have a long learning time,complex calculation process,and are closely related to the selection of damage indicators.The difficulty and sensitivity of obtaining damage indicators play a crucial role in the recognition effect.To determine the status of bridge structures in real-time,a continuous beam bridge structural damage identification and localization method based on SDAE(stacked denoising autoencoder)network is proposed,utilizing the advantages of cloud computing and deep learning computing capabilities of big data.Two span concrete continuous beam bridges are used as objects,and the acceleration response values of the structure are directly collected.The stacked denoising autoencoder is used to identify single and multiple damage locations in different noise environments,it is compared with existing machine learning methods such as BP neural network and SVM support vector machine in terms of recognition accuracy and noise resistance to verify its reliability and applicability.
NETL
NSTL
万方数据
维普
