Damage Detection of Jiangyin Bridge After Collision Accident Based on Auto-associative Neural Network
A multi-layered feed forward network-auto-associative neural network (AANN) was applied to detect damage of Jiangyin bridge after ship collision accident in this paper. Network structure was determined by taking the mean square deviation of simulating result as a standard. Then modal frequency vectors of bridge were input AANN to construct singular index. Whether the bridge had damaged could be distinguished by comparing singular index in unknown condition and normal state. Computing results indicated that there was no remarkable difference of singular index sequences around ship collision, which mean accident had not caused damage to Jiangyin Bridge.
Damage detection AANN Jiangyin Bridge Collision accident
Jun Wang Xiaoyong Ge Yu Cheng Yufeng Zhang
Jiangsu Transportation Research Institute, Nanjing 210017, China Nanjing Tap Water General Company, 210036, China
国际会议
南京
英文
808-813
2007-10-16（万方平台首次上网日期，不代表论文的发表时间）