基于模型智能修正的桥梁健康状态评估

Evaluation of Bridge Health Condition Through Model Updating by Neural Network Approach

利用神经网络方法来修正结构模型,然后再进行桥梁状态评估。在模型修正中引入模态频率和模态柔度指标作为神经网络系统的输入指标,修正后的结构数值模型能够准确反映桥梁的实际健康状态。通过巴东长江大桥实例进行了可行性验证,然后在此基础上进行各种计算工况的加载,可以评估当前桥梁的实际状态。

The intelligent updated model of neural network structure was set up to assess the bridge health condition in this paper.The mode flexibility and the mode frequency were introduced to the input parameter of neural network system.Through the network training and amending,the updated FEM model was carried out.By an example,it proved that the updated model could reflect the bridge actual state accurately.Then,with the various loads combinations on the updated model,it could evaluate the real state condition of...