Signal Quality Evaluation, Based On A Smooth And Snr Bridge Ambient Vibration Response

Dynamic testing of civil engineering structures under ambient vibration excitation has many advantages, such as no valuable excitation equipment needed, no interruption of structural operational conditions and less test time, which is now widely used system identification technical of civil engineering structures. As only output is measured in the ambient vibration testing, most of the modal parameter identification require the input data for mathematical processing, such as assuming excitation is white noise, which does not meet the white noise assumed actually, and the output signal will inevitably affected by noise interference, resulting in the uncertainty of the results and seriously affected the recognition effect. Therefore, how to evaluate testing vibration signal quality, which is a active research topic of system identification. The state-of-the-art studies are carried out in this thesis on the vibration data stationary test method, the theory, algorithms of the output signal SNR estimation, real application of bridge testing. The main work and conclusions include:1. The theoretical and hypothesis background of random data stationary test is discussed. According to the data non-stationary can be characterized by time trend of mean square values assumptions, testing random data stationary verse the run test method. The simulation proves that the run test method is well to test the stationary properties of representative stationary signal (white noise) and non-stationary signals (seismic).2. Singular spectrum analysis, the singularity entropy increment and its derivative are proposed to determine the system order incorporated for SNR estimation; Discussing self-regression model based on time series analysis to estimate SNR; Using the self-regression model and singular value decomposition for calculations respectively. The simulation show that two methods can be for SNR estimation calculations, by comparison with the self-regression model of SNR estimation algorithm can be more accurately.3. Finish the field ambient vibration tests on Zhanjiang Bay Bridge under operational conditions. With the run test to test stationary on the 12 hours record data; Based on time series analysis of self-regression model to evaluate the measured sections of response records of the Zhanjiang Bay Bridge. The experimental results show that from stationary and SNR of two aspects on random data to evaluate the measured signal quality is feasible, can evaluate the real bridge signal quality preferably.