Bridge damage detection via improved completed ensemble empirical mode decomposition with adaptive noise and machine learning algorithms

Structural health monitoring field is growing in the use of more modern techniques and tools order to identify damages civil structures. The improvements signal processing data mining have, recently, been employed due their powerful computational ability detect damage bridges. Despite majority researchers have studying laboratory-scale implementations theoretical developments, limited structural faults real bridges are still a problem. current study presents novel approach for identification by using two improved methods such as decomposition machine learning algorithms. Since obtained from traffic vibration non-linear time varying, Hilbert–Huang transform used process data. Additionally, phenomenon mode mixing presented methods, empirical (EMD). Therefore, completed ensemble EMD with adaptive noise (ICEEMDAN) was adopted. After parameter, symbolic analysis clustering-based were developed. an unsupervised algorithm group substructures similar behavior then damages. This method automatically classifying moving windows sequentially applied response bridge. validity demonstrated collected truss results show that proposed mixed effective can endow better bridge monitoring.