Automatic quality detection system for structural objects using dynamic output method: Case study Vilnius bridges

Abstract

Paper provides an attempt to create a methodology for automated structure health monitoring procedures using vibration spectrum analysis. There is an option to use autoregressive (AR) spectral analysis to extract information from frequency spectra when conventional Fast Fourier transformation (FFT) analysis cannot give relevant information. An autoregressive spectrum analysis is widely used in optics and medicine; however, it can be applied for different purposes, such as spectra analysis in electronics or mechanical vibration. This paper presents an automated structural health monitoring approach based on the algorithm-driven definition of the first resonant frequency value from a noisy signal, acquired from traffic-created bridge vibrations. We implemented the AR procedure and developed a peak detection algorithm for experimental data processing. The functionality of the proposed methodology was evaluated by performing research on six bridges in Vilnius (Lithuania). We compared three methods of data processing: FFT, filtered FFT and AR. Bridges vibrations under different excitation conditions (wind, impulse and traffic) in normal direction were measured using accelerometers. AR provided one peak representing the lowest resonant frequency in all cases, while FFT and filtered FFT provided up to 12 peaks with similar frequency values. Such results allow implementing our method for remote automated structures health monitoring and ensure structures safety using a convenient and straightforward diagnostic method.

A bearing fault diagnosis approach based on an improved neural network combined with transfer learning

“…This method achieves fault diagnosis by collecting bearing vibration signals and performing their feature extraction, analysis and identification. The classical methods include root mean square [5], crest factor [6], fast Fourier transform [7], wavelet transform and wavelet packet transform [8], etc. The main advantages of the methods are that they can be used for noise reduction and feature extraction with no need of actual mathematical modeling [9].…”