加速遗传算法在预应力连续箱梁桥有限元模型修正中的应用

Application of accelerating genetic algorithm to finite element model updating of prestressed concrete continuous box-girders

针对预应力混凝土连续箱梁桥的通扬运河大桥,建立其有限元模型。基于评价模型修正效果的频率指标与振型相关系数指标、柔度矩阵指标和挠度指标,定义三个目标函数,利用环境激励的试验模态参数,通过实数编码加速遗传算法对有限元模型进行了修正。相对于基于频率指标与振型相关系数指标定义的目标函数,基于柔度矩阵指标或挠度指标定义的目标函数修正的有限元模型,能同时较准确地反映其动态和静态特性,也能实现其试验模态振型较精确的质量归一化。

nal bridge, The finite element model（FEM） of a prestressed concrete continuous box-girders, named Tongyang ca- is built and updated based on the results of ambient vibration testing（AVT） using real-coded accelera- ting genetic algorithm, whose objective functions are defined based on frequency index and modal assurance criteri- on（MAC） index, flexibility matrix index, deflection index for evaluating the updated FEM. The updated FEM of the bridge, based on flexibility matrix index and deflection index, can not only better represent its dynamic and static characteristics, but also achieve the precise mass normalization of its experimental modal shapes