Experimental Studis On Damage Detection Of Cable Stay And Suspention Combination Sysytem Bridge

In recent years, because of the outstanding classification performance of support vector machine (SVM), many studies have been engaging on damage identification based on SVM. A lot of research accumulation on this subject, including mumerical simulations and experiment studies, carried out by the Smart Bridge Study Group (SBSG) of Southwest Jiaotong University as well. Some identification methods figured out by SBSG have achieved good results in numerical simulations, they are asked to verify by model test, and some other studies are already verified by experiments, but still required to be verified for future study.The damage identification experiment of one certain cable stay and suspension combination system model bridge is taken as the studied object; the experiment is presented form the viewpoints of sturctural system, excitation, damage modeling and data acquisition etc; then static and dynamic identification methods based on SVM are validated by using experimental measurement. The main work is listed as following:1. The state-of-the-art of damage identification experiment is reviewed from the viewpoint of Structural system, excitation, damage modeling, data-acquisition and verification methods.2. The process of damage identification experiment is introduced from Structural system, excitation, damage modeling and data acquisition as well.3. The numerical simulation model and numerical error analysis are discussed from Structural system simulation, exicitation and damage modeling.4. The static and dynamic identification methods carried out by SBSG based on SVM are verified by using the experiment measurement.