中文摘要：

利用时程分析计算结构地震反应时，不同的地震动记录或不同的地震动记录调整方法得到的结果差别很大，因此如何选择和调整地震动记录，减小结构地震反应分析结果的离散性，是提高结构抗震分析结果可靠性的必然要求。针对基于模态推覆的地震动记录调整方法的不足，提出引入以振型参与重量系数为加权系数、可考虑前若干阶振型贡献的综合调整系数调整初选的地震动记录，其中每一阶振型对应的调整系数利用该阶振型相应等效单自由度体系非线性时程分析所得最大位移匹配其目标位移确定。进一步，以一6层钢筋混凝土框架结构为例，分别采用改进的基于模态推覆、基于模态推覆和基于ASCE7的调整方法调整地震动记录进行时程分析，并对不同调整方法下结构最大水平位移、最大层间位移角等结构响应计算结果的离散性进行了对比。对比结果表明，与基于模态推覆和基于ASCE7的地震动记录调整方法相比，基于改进模态推覆的地震动记录调整方法所得结构地震响应分析结果的离散性明显下降。

英文摘要：

For time history analysis of seismic response of structures, different ground motion records or different scaling procedure to modify the ground motion records will lead to obvious record-to-record variation. Thus how to select and modify ground motion records to reduce the variation is necessary for improving the reliability of seismic analysis results. Aiming at overcoming disadvantages of the modal-pushover-based ground motion scaling procedure, a new integrated scaling factor is proposed to scale ground motion records, which uses effective modal gravity coefficients as weight coefficients to consider the contribution of the first several modes. For each mode, an individual scaling factor is determined by minimize the discrepancy between the maximum displacement of the corresponding equivalent SDOF system by nonlinear time history analysis and target displacement. Furthermore, a 6-storey reinforced concrete frame was taken as an example structure. The modified modal-pushover-based, the modal-pushover-based and the ASCE7-based ground motion scaling procedures were used to scale the preliminary ground motion records for time history analysis of the example structure. Comparison of the maximum lateral displacement and the maximum interstory drift ratio indicates that the modified modal-pushover-based scaling procedure is better than the modal-pushover-based and the ASCE7-based ground motion scaling procedures in the sense of reducing the variation of analytical results for seismic responses of structures.