Frequency Identification of Practical Bridges through Higher-Order Spectrum

Abstract

Identifying the frequencies of practical bridges can help in understanding the bridge dynamic property. However, the vibration amplitude excited by wind load and traffic load is too small, resulting in a small signal noise ratio, which affects the performance of frequency identification. This paper proposes frequency identification procedures for practical bridges to reduce the influence of noise through higher-order spectrum. First, a higher-order spectrum is introduced. Then, the frequency identification procedures are presented for practical bridges. Finally, the proposed procedures are applied to a practical bridge. A higher-order spectrum can eliminate the influence of Gaussian white noise (GWN) or reduce nonstationary random noise during frequency identification. The advantage of using the higher-order spectrum to identify frequencies is verified by simple artificial signals combined with sinusoidal signals with different frequencies and GWN. The results show that the second-order spectrum obtained by the higher-order spectrum has better frequency identification performance than the tradition spectrum and that the frequencies of the practical bridge can be identified successfully.