中文摘要：

提出了基于改进遗传算法的公路桥梁损伤程度标定的两阶段法。第一阶段：应用静应变残差进行损伤定位；第二阶段：基于已经识别出的损伤位置，利用改进的遗传算法进行损伤程度的标定。两阶段方法有效地克服了同时进行常规的损伤位置识别和损伤程度的标定的收敛速度慢、存储空间大及可能误标定等问题。某三跨连续桥梁应用分析发现，在已知很少实测数据的情况下，对损伤程度的识别取得较理想的效果，证实了基于改进遗传算法的两阶段法用于损伤程度的识别具有更高的效率，更好的灵敏度、稳定性和可靠性。

英文摘要：

A new approach based on the detected position and the improved genetic algorithms is developed, namely the two stage method. It overcomes the shortcomings of the normal method, for instance, the slow convergence, requiting large memory space and mistake identification, and so on, in which the structural damage localization and identification are carried into execution at the same time. The numerical example for 3-span-girder reinforced con- crete bridge showc that the damage magnitude identification of the bridge is accurate even if only the dead displacement data are measured, and that the method has high efficiency, sensitivity, stability and reliability of the damage extent identification.