Damage Alarming Method Research For Bridges Based On Long-term Strain Monitoring Data

Bridge occupies an important position in the national economy, assuming the responsibilities of resource transportation and commodity circulation. It will cause heavy economic losses and bad social influence once an accident occurs. The bridge will be under constant effects of normal load, environmental load, corrosion as well as reduction, causing catastrophic accident in extreme cases. Therefore damage can identify early structural damage and avoid the occurrence of major accidents. Damage alarming and safety assessment of the long-span bridges based on the structural health monitoring system(SHMS)has become one of present research hotspots in the field of civil engineering.How to effectively process, analyze and interpret the long-term data collected by the SHMS installed on the bridge is the most essential problem in the field of bridge health monitoring. To solve the above problems, the main work in this thesis are as follows:1. Through summarizing domestic and grandson damage alarming methods of bridge,discovery that existing methods has a good detection efficiency for simple bridge,but for large–scale complicated bridge still having a lot of space for further development.2. In order to realize the damage alarming of highway bridge based on the quasi-static signal from bridge health monitoring system, a moving principal component correlation analysis(MPCCA) method is proposed and is programmed through the MATLAB software. The method does not rely on structural model and has a good adaptability.3. Considering the coupled effect of the vehicle load and the ambient temperature, a three-span prestressed reinforced concrete box girder bridge is simulated as an illustrative example. The efficiency and accuracy of the proposed method for alarming the damage in different scenarios is verified. The results indicate that the present hybrid approach-the MPCCA method is better than the(moving principal component analysis) MPCA and the(moving correlation coefficient analysis) MCA methods in precision and efficiency. Especially, the MPCCA method has an excellent ability to locate the damage.4.Based on continuous signal from bridge health monitoring system, put forward method on how to pre-process the data. The pre-process of the data method and damage alarming methods mentioned above are used on measured data from monitoring system of a real bridge. The results show the practicability and reliability of the technique.