subhani-animproved-2021.pdf (11.14 MB)
An Improved Hilbert–Huang Transform for Vibration-Based Damage Detection of Utility Timber Poles
journal contribution
posted on 2021-01-01, 00:00 authored by Ipshita Das, Mohammad Taufiqul ArifMohammad Taufiqul Arif, Aman Maung Than OoAman Maung Than Oo, Mahbube SubhaniMahbube SubhaniIn this study, vibration based non-destructive testing (NDT) technique is adopted for assessing the condition of in-service timber pole. Timber is a natural material, and hence the captured broadband signal (induced from impact using modal hammer) is greatly affected by the uncertainty on wood properties, structure, and environment. Therefore, advanced signal processing technique is essential in order to extract features associated with the health condition of timber poles. In this study, Hilbert–Huang Transform (HHT) and Wavelet Packet Transform (WPT) are implemented to conduct time-frequency analysis on the acquired signal related to three in-service poles and three unserviceable poles. Firstly, mother wavelet is selected for WPT using maximum energy to Shannon entropy ratio. Then, the raw signal is divided into different frequency bands using WPT, followed by reconstructing the signal using wavelet coefficients in the dominant frequency bands. The reconstructed signal is then further decomposed into mono-component signals by Empirical Mode Decomposition (EMD), known as Intrinsic Mode Function (IMF). Dominant IMFs are selected using correlation coefficient method and instantaneous frequencies of those dominant IMFs are generated using HHT. Finally, the anomalies in the instantaneous frequency plots are efficiently utilised to determine vital features related to pole condition. The results of the study showed that HHT with WPT as pre-processor has a great potential for the condition assessment of utility timber poles.
History
Journal
Applied SciencesVolume
11Issue
7Pagination
1 - 25Publisher
MDPILocation
Basel, SwitzerlandPublisher DOI
Link to full text
eISSN
2076-3417Language
engPublication classification
C1 Refereed article in a scholarly journalUsage metrics
Categories
No categories selectedKeywords
non-destructive testingtimber utility polesvibration based damage detectionwavelet packet transformempirical mode decompositionHilbert–Huang transformScience & TechnologyPhysical SciencesTechnologyChemistry, MultidisciplinaryEngineering, MultidisciplinaryMaterials Science, MultidisciplinaryPhysics, AppliedChemistryEngineeringMaterials SciencePhysicsHilbert–Huang transform
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC