Comparison Research Of Bridge Structural Operational Modal Parameter Identification Methods

One of the most important componenet in bridge health monitoring system is operational modal parameter identification. In order to meet the real-world engineering requirement of a certain long span cable-stayed bridge health monitoring system, a whole year measured data acquired by the monitoring system are analyzed and treated, and common operational modal parameter identification methods in frequency domain, time domain and time-frequency domain are adopted to identify the operational modal parameters of main grider, pylon and stay-cable. After the comparison analysis of identified results derieved from different identification methods, the specific operational parameter identification method which is more feasible in the real-world engineering is found out. The main researching works are listed as following:1) After the analysis of the real-world requirement for bridge operational modal parameter identification, the summarization of structural operational modal parameter identification methods in frequency domain, time domain and time-frequency domain is figured out.2) The power spectrum-based frequency domain methods, the state equation-based time domain methods and data-driven based time-frequency domain method for operational modal parameter identification are briefly introduced. The identification procedures for different methods are carried out clearly, and the corresponding software are compiled based on MATLAB platform.3) The exploratory data analysis is introduced to analyze the one year measured data obtained from a certain long span cable-stayed bridge under normal operation, and the favorable measurrands are selected.4) The operational modal parameters of the main girder, pylon and stay-cable in 12 months are identified by the typical methods in frequency domain, time domain and time-frequency domain, and the comparison analysis of identified results derieved from different identification methods are carried out.5) It is shown that the more effective and applicable operational modal parameter identification method for the main grider is the combination algorithm with improved ensemble empirical mode decomposition and data-driven stochastic subspace method, and the fast and more accurate operational modal parameter identification method for stay-cable is Peak Picking method.