Modal Frequency Extraction from High-Noise Ambient Vibration Data Using Iterative Wavelet Denoising: A Case Study on Shiraz Petrochemical Tower

This study focuses on using the Discrete Wavelet Transform (DWT) as the main tool for denoising high-noise data from ambient vibration tests on a petrochemical tower in Shiraz, excited by natural wind. DWT enables multi-resolution analysis by decomposing signals into approximation and detail components, effectively capturing both global and local features. Two DWT-based approaches are evaluated: a conventional single-step method with fixed thresholds, and an iterative "peeling" method that refines the signal through repeated decomposition. While the iterative process improves adaptability, its role is to extend the denoising capabilities of DWT for highly contaminated signals. After denoising, Continuous Wavelet Transform (CWT) is applied for time-frequency analysis, and Cross Wavelet Transform (XWT) is used to identify modal frequencies through spectral power detection.

کلیدواژه ها:

Discrete wavelet transform ، Ambient Vibration Testing ، Iterative Denoising ، Modal Frequency Identification ، Cross Wavelet Transform (XWT)