The Study On Application Of FBG Sensor To Bridge Damage Detection And Evaluating System

Along with the development of the national economy, more and more large-span bridges are constructed, and the existing bridges suffer from the weather, environment factors and the action of the static and active loads imposed on them. The strength and stiffness will degrade with the time running, and damages will occur as the result, which endanger the safety of the bridges. Accordingly, the bridge safety monitoring has been received more and more concern from structure experts.A truly bridge safety monitoring system is not just the simple improve of the monitoring methods. It should be involved the functions as following: Monitoring the environment load and the entire and local state variable with various sensors; Acquire the response data of the bridge synchronously and manage the data effectively; Analyze the acquired data, process the system identification, structural infinite model modification and damage detection; the last but not least, is the after-damage management, quantificationally evaluate the damage state of the bridge according to the correct determinant of damage detection. The scientific repairment and maintenance methods are presented consequently. The last two functions are the core and the final goal of safety monitoring system.Based on the fiber grating vibration sensor and vibration response acquisition system, the core technology of bridge safety monitoring system is discussed in this paper. Neural network and rough set algorithm are suggested to be used in bridge damage detection and damage state evaluation respectively. Therefore, the following research work has been carried out in this doctoral dissertation.(1) As vibration testing has become the most effective method for the damage detection of bridge, an effective and dependable vibration testing system is developed. Fiber grating vibration sensor based on matched filtering demodulation is adopted to monitoring the vibration response data of the bridge. The problems of wave length demodulation speed and temperature error are solved effectively. The data acquisition system is developed based on USB. The software is carried out in LabVIEW by calling dynamic link library functions.(2) The measured frequency response functions (FRF) is used as the input to artificial neural networks (ANN) because it can provide more structural information in frequency domain. Since full size of FRF data is too much for the ANN, a data reduction technique based on principal component analysis (PCA) is applied to extract the features. The extracted features are used as the input data of ANN instead of the raw FRF data.(3) The self-organizing map neural network is chosen because of its superiority in analyzing high-dimensional data without supervising. A steel box girder model with multi damage states is presented to demonstrate the effectiveness of the method.(4) Damage state evaluation of the damaged bridges can provide scientific reference for the bridge management and maintance. Take the advantage in processing incomplete information without any preliminary or additional information about data, rough set theory is used to mine rules between the discretized testing data of each branch construction and the state of the bridge. The evaluating results can offer valuable perspectives of structural design and analysis as well.