基于LSTM的自复位中心支撑钢框架地震响应预测
Prediction of Seismic Response for Self-Centering Braced Steel Frames Based on LSTM
为预测自复位中心支撑钢框架结构在地震作用下的响应,本文通过深度学习算法搭建长短期记忆神经网络模型(LSTM),构建了输入地震加速度与输出结构地震位移响应之间的非线性映射关系,探究数据窗口尺寸和数据集划分比例因素对模型预测性能的影响。研究结果表明:所搭建LSTM神经网络模型具有稳健的预测性能,对结构屋顶位移的预测效果较好,峰值相对误差为2.36%,相关系数为0.94;适当增大数据窗口尺寸,仍能准确预测结构屋顶位移和层间位移,并提高预测效率,不会明显改变预测响应峰值的相位分布;对于样本量较小的情况,当已知数据集划分比例位于2∶1~3∶1之间时,模型整体预测效果较好,训练集和验证集充分覆盖各种时间步长类型时,能有效提高模型整体的泛化能力,但对小微数据敏感性略差。
To predict the response of self-centering braced steel frame structures under seismic loading, this paper employs a deep learning algorithm to construct a Long Short-Term Memory (LSTM) neural network model. This model establishes a nonlinear mapping relationship between input seismic acceleration and output structure seismic drift responses, exploring the effects of data window-size and dataset splitting ratios on the predictive performance of the model. The research findings indicate that the proposed LSTM neural network model demonstrates robust predictive capabilities, particularly in predicting the roof drift of the structure, achieving a peak relative error of 2.36% and a correlation coefficient of 0.94. Increasing the data window-size appropriately can still accurately predict both the roof drift and residual inter-story drift of the structure while enhancing prediction efficiency, without significantly altering the peak phase distribution of the predicted responses. For scenarios with limited sample sizes, when the known dataset splitting ratio falls within the range of 2∶1 to 3∶1, the model exhibits satisfactory overall predictive performance. When the training and validation sets adequately cover various time-step types, the model's overall generalization ability is effectively improved. However, it displays slightly less sensitivity to micro-data.
| [1] |
FANG C,WANG W,QIU C,et al.Seismic resilient steel structures:A review of research,practice,challenges and opportunities[J].Journal of Constructional Steel Research,2022,191:107172. |
| [2] |
HU S,WANG W,ALAM M S,et al.Life-cycle benefits estimation of self-centering building structures[J].Engineering Structures,2023,284:115982. |
| [3] |
ZHANG R,CHEN Z,CHEN S,et al.Deep long short-term memory networks for nonlinear structural seismic response prediction[J].Computers & Structures,2019,220:55-68. |
| [4] |
韩小雷,吴梓楠,杨明灿,基于深度学习的区域RC框架结构震损评估方法研究[J].建筑结构学报,2020,41(S2):27-35. |
| [5] |
HAN Xiaolei,WU Zinan,YANG Mingcan,et al.Research on seismic damage assessment method of regional RC frame structure based on deep learning[J].Journal of Building Structures,2020,41(S2):27-35.(in Chinese) |
| [6] |
蔡燕飞.基于深度学习的区域RC框架结构地震时程响应预测[D].广州:华南理工大学,2022. |
| [7] |
CAI Yanfei.Seismic time history response prediction of regional RC frame structure based on deep learning[D].Guangzhou:South China University of Technology,2022.(in Chinese) |
| [8] |
田壮,樊启武,王昌杰.深度学习在桥梁响应预测与健康监测中的应用[J].铁道工程学报,2021,38(6):47-52. |
| [9] |
TIAN Zhuang,FAN Qiwu,WANG Changjie.Application of deep learning in bridge response prediction and health monitoring[J].Journal of Railway Engineering Society,2021,38(6):47-52.(in Chinese) |
| [10] |
王琛,樊健生.具有历史依赖效应的材料及结构响应预测通用深度学习MechPerformer模型[J].建筑结构学报,2022,43(8):209-219. |
| [11] |
WANG Chen,FAN Jiansheng.MechPerformer model for material and structure response prediction with historical dependence[J].Journal of Building Structures,2022,43(8):209-219.(in Chinese) |
| [12] |
LIAO Y,LIN R,ZHANG R,et al.Attention-based LSTM (AttLSTM) neural network for seismic response modeling of bridges[J].Computers & Structures,2023,275:106915. |
| [13] |
FEMA P-751:2009 NEHRP recommended seismic provisions:Design examples[R].Washington,DC:Federal Emergency Management Agency, 2012. |
| [14] |
MCKENNA F,FENVES G L,SCOTT M H,et al.OpenSees:Open system for earthquake engineering simulation[CP].Berkeley,CA:Pacific Earthquake Engineering Research Center,University of California,2006. |
| [15] |
HU S,WANG W,QU B.Seismic economic losses in mid-rise steel buildings with conventional and emerging lateral force resisting systems[J].Engineering Structures,2020,204:110021. |
| [16] |
QIU C X,ZHU S.Performance-based seismic design of self-centering steel frames with SMA-based braces[J].Engineering Structures,2017,130:67-82. |
| [17] |
SOMERVILLE P G.Development of ground motion time histories for phase 2 of the FEMA/SAC steel project[M].Sacramento,CA:SAC Joint Venture,1997. |
| [18] |
ASCE,Minimum Design Loads for Buildings and Other Structures[S].ASCE 7-10,American Society of Civil Engineers,Reston,Virginia,2010. |
| [19] |
建筑抗震设计规范:GB 50011—2010[S].北京:中国建筑工业出版社,2010. |
| [20] |
Code for seismic design of buildings:GB 50011—2010[S].Beijing:China Architecture & Building Press,2010.(in Chinese) |
| [21] |
ZHANG R,WANG W,ALAM M S.Performance-based seismic design of full and partial self-centering steel braced frames using modified lateral force distribution[J].Journal of Constructional Steel Research,2022,196:107438. |
| [22] |
李文静,王潇潇.基于简化型LSTM神经网络的时间序列预测方法[J].北京工业大学学报,2021,47(5):480-488. |
| [23] |
LI Wenjing,WANG Xiaoxiao.Time series prediction method based on simplified LSTM neural network[J].Journal of Beijing University of Technology,2021,47(5):480-488.(in Chinese) |
| [24] |
GAL Y,GHAHRAMANI Z.A theoretically grounded application of dropout in recurrent neural networks[C]//NIPS'16:Proceedings of the 30th International Conference on Neural Information Processing Systems,2016:1027-1035. |
| [25] |
ABADI M,AGARWAL A,BARHAM P,et al.TensorFlow:Large-scale machine learning on heterogeneous systems[EB/OL].Mountain View,CA:Google,2015.https://www.tensorflow.org/. |
| [26] |
GLOROT X,BENGIO Y.Understanding the difficulty of training deep feedforward neural networks[C]//Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics.Sardinia,Italy:JMLR Workshop and Conference Proceedings,2010:249-256. |
| [27] |
KINGMA D P,BA J.Adam:A method for stochastic optimization[C]//Proceedings of the 3rd International Conference on Learning Representations.San Diego,CA,USA:ICLR,2015:1-15. |
国家自然科学基金项目(52378182)
国家自然科学基金项目(52308195)
/
| 〈 |
|
〉 |