Inexpensive Health Monitoring of Cable-Stayed Brides Using Synchrosqueezed Wavelet Transform and Nonlinear Main Components Analysis
1گروه سازه و زلزله، دانشکده عمران، دانشگاه علم و صنعت ایران، تهران، ایران
2گروه عمران، دانشکده فنی و مهندسی، دانشگاه آزاد اسلامی، واحد رودهن، رودهن، ایران
3گروه سازه و زلزله، دانشکده مهندسی عمران، دانشگاه علم و صنعت ایران، تهران ، ایران، ریاست دانشکده عمران
Nowadays, the structural health monitoring has standardized in many countries. The health monitoring systems for large and important structures are equipped with a number of sen-sors. Therefore, their usage due to high prices is not yet prevalent in our country. The main purpose of this research is to providing an inexpensive method of structures health monitor-ing based on signal processing. Accordingly, only three sensors have been used to identify-ing the structure damage. In this article, a new damage detection method has been proposed. This method has driven from a combination of the signal processing techniques using the synchrosqueezed wavelet transform method, the data clustering, and the regression analysis using autoassociative dependent neural networks method. For this purpose, the Yonghe bridge data references have been used, whose data are recorded based on the actual vibration of the bridge. The results indicate that the proposed signal processing method has sufficient accuracy level to identifying the damage.