Comparing the efficeincy of several time-frequency signal processing methods for damage detection

Detection of damage in structures has been a major concern for structural engineers. Among the various damage detection methods introduced, signal based methods are attracting more attention today. However, these methods show weaknesses in confronting with civil structures with numerous members. For this reason, a reliable signal processing tool is of major importance since error in extraction of signal may result in faulty damage detection. In this study, the efficiency of several modern signal processing methods is investigated to extract signal features. Four methods are studied including Hilbert-Huang Transform (HHT), Direct Quadrature (DQ), Teager Energy Operator (TEO) and Higher-order Energy Operator (HEO). Among these methods HHT and TEO are well-known and are widely used in engineering applications. DQ and HEO are recently proposed and not very well-known for engineers. Results show that although HHT is the oldest procedure, it still has the highest accuracy to extract frequencies and amplitudes of structural signals.

Keywords:

signal processing , Hilbert-Huang transform (HHT) , direct quadrature (DQ) , Teager energy operator (TEO) , higher-order energy operator (HEO)