| Popis: |
Global Navigation Satellite Systems (GNSS) Realtime Kinematic (RTK) positioning technique has been widely used for the structural health monitoring (SHM) of different structures in the past two decades. Through post processing and analysis, it has been demonstrated that the displacements and natural frequencies identified with GNSS data are highly consistent with those obtained by using a finite element (FE) model. However, structural health monitoring needs to measure all spectrum of the dynamic responses of bridges such as deformations, natural frequencies, damping, etc. in real-time in order to support the timely decision making for the bridge operation and maintenance, particularly under extreme loading conditions caused by busy traffic, severe wind or even earthquake etc. This paper proposes a new quasi real time time-frequency analysis strategy based on the Fast Fourier Transform (FFT). With the support of the European Space Agency (ESA) and the University of Nottingham in the UK, one week of real-life GNSS data gathered from the Forth Road Bridge in Scotland has been used since the traffic loading has an approximate repetition period of one week from the weekdays to weekend. Firstly, the approximate frequency distribution is achieved by using the whole date set. Then a sliding window method is proposed to simulate a quasi real time mode for the time-frequency analysis, and a set of experiments are carried out to decide the optimal window length and the overlapped sliding step, through which the natural frequencies and relevant deformation amplitudes can be calculated at the same time. Finally, the results show that the natural frequencies calculated by FFT are quite stable which indicates the frequency responses are not sensitive enough to the changing loadings. However, the relevant amplitude time series of each frequency can clearly display the influence caused by different kinds of loading respectively, such as vehicles and wind etc., which would be a reliable indicator of bridge dynamic responses to assess the structural health conditions in the future. |
