两层三跨地铁车站结构抗震性能等级量化指标

Quantitative index of seismic performance levels for subway station structure with two-story and three-span

针对目前浅埋两层三跨地铁车站结构抗震性能指标在限值标定方面存在的不足,以典型两层三跨地铁车站结构为研究对象,建立土-地下结构非线性动力相互作用有限元模型,考虑工程场地条件及地震动的多样性,基于增量动力分析(IDA)方法,开展了两层三跨地铁车站结构地震响应的900个工况计算。利用Python软件实现了大量计算数据的高效处理,并进一步利用提出的地铁车站结构易损截面的平均地震损伤值(DCPs)概念,建立了预测地铁车站结构抗震性能等级的随机森林预测模型(RFPM),该预测模型精度为95%。从地铁车站结构地震损伤的直观分析及层间位移角(IDR)数据的参数检验分析两个角度,验证了该地铁车站结构地震损伤与IDR之间的强相关性,论证了IDR用作该地铁车站结构抗震性能量化指标的可靠性,最终提出了基于概率统计的该类地铁车站结构抗震性能等级量化指标,对应的4个IDR限值分别为0.126%、0.266%、0.450%及0.813%。

In response to the deficiencies of quantitative index of seismic performance levels(SPLs)for shallow-buried subway station structure with two-story and three-span,especially considering the limit value calibration at present given the variation of sites and the diversity of input ground motions,a finite element model of nonlinear dynamic soil-underground structure interaction was established.Based on the incremental dynamic analysis(IDA)method,the seismic response of the subway station structure was calculated considering 900 cases.Python was used to realize the efficient processing of massive data.According to the seismic damage characteristics of the subway station structure,this paper put forward a concept of average seismic damage value of vulnerable sections of the subway station structure,and established a random forest prediction model(RFPM,and its accuracy is 95%)to predict the SPL.The seismic damage of the subway station structure was analyzed intuitively and a parametric test of the IDRs was carried out to verify the strong correlation between the seismic damage and the IDR.Thus,the reliability of IDR used as a quantitative index of seismic performance was verified.The quantitative index for SPLs based on probability statistics is further improved,and the corresponding four limit values are 0.126%,0.266%,0.450%and 0.813%.