地震研究

为研究RC框架结构在主余震序列地震动作用下的易损性,设计了一个6层RC框架结构,挑选不同余震持时的主余震地震动合成得到主震-长持时余震和主震-短持时余震序列地震动各20组,选用地震动强度指标作为预测变量,将全部地震序列输入到结构中进行IDA分析,以IDA分析结果作为神经网络数据库,训练出最佳BP神经网络模型,得到BP神经网络的结构易损性曲面,从而综合考虑余震持时和主震强度的向量型IM与结构最大层间位移角之间的关系。结果表明:在主震-长持时余震和主震-短持时余震两种地震作用下,前者对结构的影响较大,造成结构的失效概率更大; 基于BP神经网络的易损性函数能更好地反映结构损伤。

In order to study the vulnerability of the RC frame structure under the action of the mainshock-aftershock earthquake sequence,we design a 6-layer RC frame structure.By selecting recordings of the ground motion of the mainshock-aftershock sequence containing aftershocks with different durations,we respectively synthesize 20 sets of mainshock-aftershock sequence containing long-duration aftershocks and 20 sets of mainshock-aftershock sequence containing short-duration aftershocks.We select the intensity index of the ground motion IM as the predictive variable,and input all the ground motions of the 40 sets of mainshock-aftershock sequence into the structure for IDA analysis.Then we take the IDA analysis results as the neural network database to train the best BP neural network model.Thus,we get the vulnerable surface of the BP neural network.On this basis we obtain the relationship between the IM and the maximum interlayer displacement angle of the structure.We find that the mainshock-aftershock sequences containing long-duration aftershocks impacts the structure greater than the ones containing short-duration aftershocks.The vulnerability function based on BP neural network can better reveal the damage to the structure.

引言
1 BP神经网络理论
2 RC框架结构模型建立
3 地震动选取
4 基于BP神经网络的结构主余震易损性分析
5 结论

图1 3层神经网络模型示意图<br/>Fig.1 A three-layer neural network model

图1 3层神经网络模型示意图
Fig.1 A three-layer neural network model

图2 结构一层平面图(a)及结构剖面图(b)
Fig.2 The plane graph of the building's 1st floor(a)and the building's profile(b)

图3 构建的有限元模型<br/>Fig.3 A constructed finite element model

图3 构建的有限元模型
Fig.3 A constructed finite element model

表1 结构配筋信息<br/>Tab.1 The building's reinforcement information

表1 结构配筋信息
Tab.1 The building's reinforcement information

表2 前3阶周期对比<br/>Tab.2 Comparison of the first three orders' periods

表2 前3阶周期对比
Tab.2 Comparison of the first three orders' periods

图4 主余震序列地震动构成<br/>Fig.4 Construction of the mainshock-aftershock sequence ground motion

图4 主余震序列地震动构成
Fig.4 Construction of the mainshock-aftershock sequence ground motion

图5 长(a)、短(b)持时地震动谱匹配结果<br/>Fig.5 The matching of the long-duration(a)and the short-duration(b)ground motion spectra

图5 长(a)、短(b)持时地震动谱匹配结果
Fig.5 The matching of the long-duration(a)and the short-duration(b)ground motion spectra

表3 主余震地震动记录<br/>Tab.3 Ground motion records of the mainshocks and aftershocks

表3 主余震地震动记录
Tab.3 Ground motion records of the mainshocks and aftershocks

图6 BP 神经网络模型示意图<br/>Fig.6 A BP neural network model diagram

图6 BP 神经网络模型示意图
Fig.6 A BP neural network model diagram

图7 主震-长持时(a)和主震-短持时(b)余震序列地震动作用下结构的BP神经网络模型训练结果<br/>Fig.7 Results from the training of the BP neural network model for the structures subjected to the ground motion of the mainshocks and their long-duration aftershocks(a)and short-duration(b)aftershocks

图7 主震-长持时(a)和主震-短持时(b)余震序列地震动作用下结构的BP神经网络模型训练结果
Fig.7 Results from the training of the BP neural network model for the structures subjected to the ground motion of the mainshocks and their long-duration aftershocks(a)and short-duration(b)aftershocks

图8 4种破坏状态时主震-长持时(a)及主震-短持时(b)余震序列地震动作用下结构的易损性曲面<br/>Fig.8 Susceptibility surfaces of structure subjected to the mainshocks and their long-duration(a)and short-duration(b)aftershocks under four damage conditions

图8 4种破坏状态时主震-长持时(a)及主震-短持时(b)余震序列地震动作用下结构的易损性曲面
Fig.8 Susceptibility surfaces of structure subjected to the mainshocks and their long-duration(a)and short-duration(b)aftershocks under four damage conditions

表4 损伤状态的划分<br/>Tab.4 Classification of the structure's damage conditions

表4 损伤状态的划分
Tab.4 Classification of the structure's damage conditions

图9 4种破坏状态时S/sub>a(T1,0.05)罕遇下主震-长持时(a)和主震-短持时(b)余震序列地震动作用下结构超越概率<br/>Fig.9 The exceeding probability of the structure when the structure is subjected to the rare mainshocks and their long-duration(a)and short-duration(b)aftershocks under four damage conditions

图9 4种破坏状态时S/sub>a(T1,0.05)罕遇下主震-长持时(a)和主震-短持时(b)余震序列地震动作用下结构超越概率
Fig.9 The exceeding probability of the structure when the structure is subjected to the rare mainshocks and their long-duration(a)and short-duration(b)aftershocks under four damage conditions

蔡斌,许龙飞,郝丽妍.2019.基于BP神经网络的受火后RC梁钢筋温度预测[J].北方建筑,4(4):23-27.
Cai B,Xu L F,Hao L Y.2019.Prediction of reinforcement temperature of post-fire reinforced concrete beams based on back-propagation neural networks[J].Northern Architecture,4(4):23-27.(in Chinese)
程诗焱,韩建平,于晓辉,等.2021.基于BP神经网络的RC框架结构地震易损性曲面分析:考虑地震动强度和持时的影响[J].工程力学,38(12):107-117.
Cheng S Y,Han J P,Yu X H,et al.2021.Seismic fragility surface analysis of RC frame structures based on BP neural networks:Accounting for the effects of ground motion intensity and duration[J].Engineering Mechanics,38(12):107-117.(in Chinese)
程诗焱.2020.考虑地震动持时影响的RC框架结构易损性研究[D].兰州:兰州理工大学.
Cheng S Y.2020.Investigation on seismic fragility analysis of RC frame structures accounting for ground motion duration[D].Lanzhou:Lanzhou University of Technology.(in Chinese)
范宣梅,王欣,戴岚欣,等.2022.2022年MS6.8级泸定地震诱发地质灾害特征与空间分布规律研究[J].工程地质学报,30(5):1504-1516.
Fan X M,Wang X,Dai L X,et al.Characteristics and spatial distribution pattern of MS6.8 Luding earthquake occurred on September 5,2022[J].Journal of Engineering Geology,30(5):1504-1516.(in Chinese)
韩建平,李军.2020.考虑主余震序列影响的低延性钢筋混凝土框架易损性分析[J].工程力学,37(2):124-133.
Han J P,Li J.2020.Seismic fragility analysis of low-ductile RC frame accounting for the influence of mainshock-aftershock sequences[J].Engineering Mechanics,37(2):124-133.(in Chinese)
何正风.2016.MATLAB R2015b神经网络技术[M].北京:清华大学出版社.
He Z F.2016.MATLAB R2015b neural network technology[M].Beijing:Tsinghua University Press.(in Chinese)
孔令峰.2019.地震动持时对结构地震反应影响[D].哈尔滨:中国地震局工程力学研究所.
Kong L F.2019.Effect of earthquake duration on seismic response of structures[D].Harbin:Institute of Engineering Mechanics,China Earthquake Administration.(in Chinese)
李喜梅,李明睿,母渤海.2021.人工主余震作用下桥墩结构损伤预测评估分析[J].地震工程学报,43(6):1402-1408.
Li X M,Li M R,Mu B H.2021.Damage prediction and evaluation of bridge pier structures under artificial mainshock-aftershock sequences[J].China Earthquake Engineering Journal,43(6):1402-1408.(in Chinese)
刘全明.2022.基于粗粒度并行遗传算法和机器学习的隔震结构优化设计研究[D].兰州:兰州理工大学.
Liu Q M.2022.Research on optimal design of isolated structure based on parallel genetic algorithm and machine learning[D].Lanzhou:Lanzhou University of Technology.(in Chinese)
孙小云,韩建平,党育,等.2018.地震动持时对考虑梁柱节点区不同破坏模式RC框架的地震易损性影响[J].工程力学,35(5):193-203.
Sun X Y,Han J P,Dang Y,et al.2018.Effect of ground motion duration on seismic fragility of RC frames with different beam-column joint failure modes[J].Engineering Mechanics,35(5):193-203.(in Chinese)
王新悦.2018.框架—核心筒混合结构在主余震序列作用下的易损性分析[D].哈尔滨:哈尔滨工业大学.
Wang X Y.2018.Fragility analysis of a frame-core tube hybrid structure under mainshock-aftershock sequences [D].Harbin:Harbin Institute of Technology.(in Chinese)
温卫平,籍多发,虞亦琦,等.2021.考虑余震影响的RC框架结构抗震设计方法[J].土木工程学报,54(2):43-49.
Wen W P,Ji D F,Yu Y Q,et al.2021.Seismic design method of RC frame structures considering effects of aftershocks[J].China Civil Engineering Journal,54(2):43-49.(in Chinese)
徐峥匀,钱松荣.2022.基于BP神经网络的桥梁损伤识别研究[J].软件导刊,21(12):53-57.
Xu Z Y,Qian S R.2022.Research on bridge damage identification based on BP neural network[J].Software Guide,21(12):53-57.(in Chinese)
于晓辉,马富梓,吕大刚.2020.考虑多次余震的钢筋混凝土框架结构地震损伤分析[J].建筑结构学报,41(S2):19-26.
Yu X H,Ma F Z,Lyu D G.2020.Seismic damage analyses of reinforced concrete frame structures considering multiple aftershocks[J].Journal of Building Structures,41(S2):19-26.(in Chinese)
张挺.2018.主余震序列作用下SRC框架-核心筒结构易损性分析[D].西安:西安建筑科技大学.
Zhang T.2018.Seismic fragility analysis of SRC frame-core wall structure under mainshock-aftershock sequences[D].Xi'an:Xi'an University of Architecture and Technology.(in Chinese)
周敉,张玥,彭挺.2004.RC梁桥承载力BP神经网络预测模型[J].公路与汽运,(3):68-70.
Zhou M,Zhang Y,Peng T.2004.BP Neural network prediction model for RC girder bridge load carrying capacity[J].Highways & Automotive Applications,(3):68-70.(in Chinese)
GB 50011—2010,建筑抗震设计规范(2016年版)[S].
GB 50011—2010,Code for seismic design of buildings(2016 edition)[S].(in Chinese)
Chandramohan R,Baker J W,Deierlein G G.2016.Quantifying the influence of ground motion duration on structural collapse capacity using spectrally equivalent records[J].32(2):927-950.
FEMA 356.2000.Prestandard and commentary for the seismic rehabilitation of builings[S].Washington D C.
Hammer B.2001.Neural Smithing-supervised learning in feedforward artificial neural networks[J].Pattern Analysis & Applications,(1):73-74.
Qiao Y,Lu D,Yu X.2022.Shaking table tests of a reinforced concrete frame subjected to mainshock-aftershock sequences[J].Journal of Earthquake Engineering,26(4):1693-1722.
Ruiz-Garcia J.2012.Mainshock-aftershock ground motion features and their influence in building's seismic response[J].Journal of Earthquake Engineering,16(5):719-737.
Simon H.2008.Neural networks and learning machines(3rd Edition)[M].Prentice Hall Hamilton,Ontario,Canada.
Wen W,Ji D,Zhai C.2022.Cumulative damage of structures under the mainshock-aftershock sequences in the near-fault region[J].Journal of Earthquake Engineering,26(4):2088-2102.
Yin Y J,Li Y.2011.Loss estimation of light-frame wood construction subjected to mainshock-aftershock sequences[J].ASCE Journal of Performance of Constructed Facilities,25(6):504-513.
Zhang Y,Wang Z,Jiang L,et al.2023.Seismic fragility analysis of masonry structures considering the effect of mainshock-aftershock sequences[J].Engineering Structures,275(Part B):115287.

备注

引言

1 BP神经网络理论

2 RC框架结构模型建立

3 地震动选取

4 基于BP神经网络的结构主余震易损性分析

5 结论

期刊信息