基于神经网络和遗传算法结合的桥梁结构损伤诊断

Identification of Damages in Bridge Structures Based on Combining Neural Networks and Genetic Algorithm

利用神经网络对有限元计算的样本数据建立起桥梁结构的损伤参数——损伤位置、损伤程度(输入)以及位移差——损伤结构与完好结构的位移差(输出)的全局性映射关系,以获得遗传算法求解桥梁结构损伤问题所需的目标函数值,从而识别桥梁的损伤参数.某连续梁桥算例表明,此方法可以在较少的有限元分析次数下获得较好的损伤诊断结果.

A non linear mapping function from multiple input data （damage parameters-damage locations and damage extents） to multiple output data （differences of displacements between damaged structure and intact structure calculated by FEM） is constructed within BP neural networks in order to obtain the obieetive function values that are necessary in identification of the damages in bridge structures using genetic algorithms and get the optimum damage parameters. An example of a pre-stressed continuous frame bridge is provided to indicate that the better optimum damage parameters can be obtained with less finite element analyses.