DOI QR코드

DOI QR Code

Research on Hyperparameter of RNN for Seismic Response Prediction of a Structure With Vibration Control System

진동 제어 장치를 포함한 구조물의 지진 응답 예측을 위한 순환신경망의 하이퍼파라미터 연구

  • Kim, Hyun-Su (Division of Architecture, Sunmoon University) ;
  • Park, Kwang-Seob (Boru E&C)
  • 김현수 (선문대학교 건축학부) ;
  • 박광섭 ((주)보루)
  • Received : 2020.03.05
  • Accepted : 2020.03.20
  • Published : 2020.06.15

Abstract

Recently, deep learning that is the most popular and effective class of machine learning algorithms is widely applied to various industrial areas. A number of research on various topics about structural engineering was performed by using artificial neural networks, such as structural design optimization, vibration control and system identification etc. When nonlinear semi-active structural control devices are applied to building structure, a lot of computational effort is required to predict dynamic structural responses of finite element method (FEM) model for development of control algorithm. To solve this problem, an artificial neural network model was developed in this study. Among various deep learning algorithms, a recurrent neural network (RNN) was used to make the time history response prediction model. An RNN can retain state from one iteration to the next by using its own output as input for the next step. An eleven-story building structure with semi-active tuned mass damper (TMD) was used as an example structure. The semi-active TMD was composed of magnetorheological damper. Five historical earthquakes and five artificial ground motions were used as ground excitations for training of an RNN model. Another artificial ground motion that was not used for training was used for verification of the developed RNN model. Parametric studies on various hyper-parameters including number of hidden layers, sequence length, number of LSTM cells, etc. After appropriate training iteration of the RNN model with proper hyper-parameters, the RNN model for prediction of seismic responses of the building structure with semi-active TMD was developed. The developed RNN model can effectively provide very accurate seismic responses compared to the FEM model.

References

  1. Kim, H. S., & Kang, J. W., "Seismic Response Control of Retractable-roof Spatial Structure using Smart TMD", Journal of Korean Association for Spacial Structures, Vol.16, No.4, pp.91-100, 2016, doi: 10.9712/KASS.2016.16.4.091 https://doi.org/10.9712/KASS.2016.16.4.091
  2. Bathaei, A., Zahrai, S. M., & Ramezani, M., "Semi-active seismic control of an 11-DOF building model with TMD+MR damper using type-1 and -2 fuzzy algorithms", Journal of Vibration and Control, Vol.24, No.13, pp.2938-2953, 2018 https://doi.org/10.1177/1077546317696369
  3. Nagarajaiah, S., & Narasimhan, S., "Smart base-isolated benchmark building. Part II: phase I sample controllers for linear isolation systems", Structural Control and Health Monitoring, Vol.13, No.2-3 pp.589-604, 2006, doi: 10.1002/stc.100 https://doi.org/10.1002/stc.100
  4. Kim, H. S., & Kang, J. W., "Vibration Control Performance Evaluation of Hybrid Mid-Story Isolation System for a Tall Building", Journal of Korean Association for Spatial Structures, Vol.18, No.3 pp.37-44, 2018 doi: 10.9712/KASS.2018.18.3.37 https://doi.org/10.9712/KASS.2018.18.3.37
  5. dProgrammer. Recurrent Neural Network (RNN), Long-Short Term Memory (LSTM) & Gated Recurrent Unit (GRU) (April 2019 ed.). Retrieved from http://dprogrammer.org/rnnlstm-gru
  6. Sues, R. H., Mau, S. T., & Wen, Y. K., "System Identification of Degrading Hysteretic Restoring Forces", Journal of Engineering Mechanics, Vol.114, No.5, pp.833-846, 1988, doi: 10.1061/(ASCE)0733-9399(1988)114:5(833) https://doi.org/10.1061/(ASCE)0733-9399(1988)114:5(833)