EMD and VARMA model based structural damage detection
Dong Yinfeng 1; Li Yingmin 1; Lai Ming 2
1. College of Civil Engineering, Chongqing University, Chongqing 400045, China；2. Department of Science and Technology, Ministry of Construction, Beijing 100835, China
Abstract：Based on EMD and VARMA model, a method for structural damage detection is proposed. By using the time varying VARMA model to represent the intrinsic mode functions (IMFs) obtained from the EMD of structural response signal, the Kalman filter is used to estimate the VARMA coefficients. Then, a novel damage index derived from the VARMA coefficients is used to detect structural damages. To investigate the real performance of the damage index, the Imperial County Services Building and Van Nuys Hotel are used as the benchmark structures in the examples given. The analysis of the actual response records during earthquake shows that the anti-noise ability and sensitivity of damage index is satisfactory in real environments with the presence of actual noise and it can indicate the occurrence and relative severity of structural damages at multiple locations in an efficient manner. Since the damage index is defined based on the feature extraction of a signal with no additional information required, it can be used to detect both global and local structural damages. Also, the damage index can be potentially used in real time (online) damage detection or structural health monitoring, for it is directly derived from time varying VARMA coefficients. Finally, some recommendations for future research are provided.
2010, Vol. 29 
