Bridge Operational Modal Analysis

Bridge health monitoring system is one of the vital measurements to ensure the operational safety of existing bridge, and the accurate identification of bridge operational modal parameters can give a real-time presentation of natural dynamic properties for bridge structure and is significant for long-term bridge health monitoring.The dynamic response signals of bridge structure under operational conditions are characterized by nonstationarity, but the stationary signals are required by all of contemporary modal parameter identification methods; therefore, the mearsured dynamic signals should be treated for obtaining stationary signals with high singal noise ration and rich structural information, and the requirement of existing opearational modal parameters identification methods is met. The main works are listed as following.1. The fundmental principles of empirical mode decomposition (EMD) and ensemble empirical mode decomposition (EEMD) are summarized. As for the end effects and mode mixing problem raised by EMD and EEMD, one improved treatment method combination moving closed mirror extention with advanced EEMD is proposed, and the corresponding software is compiled based on MATLAB platform;2. The basic principle of correlation detection is presented from the viewpoint of signal processing, and improved coorelation detection method with phase search is figured out to depress the noise in mearsured response signals after the selection of intrisinc mode functions.3. Band-stop filter, selection of intrinsic mode functions and correlation detection are adopted to extract the free vibration responses from the bridge operational dynamic measurements for the purpose of structural dynamic properties identification.4. Four signal processing schemes are proposed in light of different orders of the extraction of free vibration and denoise, etc. Modal parameters of one SDOF system, one3DOFs system and one ceration simple supported bridge are identified aided by the data-driven stochastic subspace identification algorithm, and then the more suitable and feasible signal processing scheme is determined.5. Testing signals from one certain real long-span cable-stayed bridge are conditioned, and its operational modal parameters are identified by the previously determined signal processing scheme.