深度學習於提升地震模擬振動臺加速度性能之研究

Improved Acceleration Tracking Performance of Seismic Simulators using Supervised Deep Learning

賴魁星(Kui-Xing Lai) ； 陳沛清(Pei-Ching Chen)

《結構工程》 37卷4期 (2022/12) Pp. 86-107

https://doi.org/10.6849/SE.202212_37(4).0005

摘要

地震模擬振動臺是地震工程研究重要的實驗設備，已廣泛應用在各種結構系統的耐震性能試驗，諸如鋼筋混凝土結構、鋼結構、隔減震結構以及精密設備等，因此振動臺加速度控制之精確與否在振動臺實驗中特別重要。本研究提出使用深度學習的方式，使用了長短期記憶（Long Short-Term Memory, LSTM）神經網路，透過大量實驗資料訓練出振動臺的外迴路控制器。本研究使用油壓驅動之大型單軸向振動臺，設計一組鋼構造試體安裝其上並進行大量的振動臺實驗，以得到LSTM之訓練資料，再將訓練所得到的LSTM作為前饋控制器，加裝在振動臺既有的三參數控制架構中，藉此補償振動臺加速度控制的動態響應。實驗結果證明LSTM前饋控制器可有效地降低振動臺的加速度誤差，並減少試體與振動臺的互制現象，使振動臺之加速度性得到顯著的提升。本研究成果顯示應用深度學習於提升振動臺加速度表現的發展潛能，未來可進行更多的相關研究，以改善振動臺實驗之測試品質。

關鍵字

地震模擬振動臺 ； 加速度控制 ； 三參數控制器 ； 深度學習 ； 長短期記憶神經網路

並列摘要

Seismic shake table testing has been widely used for various structural systems such as steel structures, reinforced-concrete structures, energy-dissipated and base-isolated buildings, and nonstructural components etc. Therefore, accurate replication of shake table acceleration is particularly important to these tests. In this study, supervised deep learning approach is applied as an alternative for seismic shake table control. The Long Short-Term Memory (LSTM) neural network is built for training the controller to improve acceleration performance of the shake table. A large-scale servo-hydraulic uniaxial shake table is adopted. A steel specimen is designed and fabricated for performing a large number of shake table tests. Then, the shake table testing data are used to train a feedforward controller using LSTM which is implemented close to an existing Three-Variable Control (TVC) loop. The validating experimental results prove that the acceleration tracking performance is improved compared with conventional TVC. The control-structure interaction is also suppressed. The experimental results demonstrate the proposed control scheme reduces the acceleration tracking error effectively compared with conventional TVC control. The research results also show great potential for deep learning application to seismic shake table control in the future. Keywords: Constitutive model, anisotropy, shear-slip and re-contact, mesh-sensitivity, non-proportional loading, concrete, finite element.

並列關鍵字

Seismic shake table ； acceleration control ； three-variable control ； deep learning ； long short-term memory neural network

參考文獻

Chen, P. C., Kek, M. K., Hu, Y. W., and Lai, C. T. (2018). Statistical reference values for control performance assessment of seismic shake table testing. Earthquakes and Structures, 15(6), 595-603.

Yao, J., Dietz, M., Xiao, R., Yu, H., Wang, T., and Yue, D. (2016). An overview of control schemes for hydraulic shaking tables. Journal of Vibration and Control, 22(12), 2807–2823.

Dyke, S. J., Spencer, Jr. B. F., Quast, P., and Sain, M. K. (1995). Role of control-structure interaction in protective system design. Journal of Engineering Mechanics, ASCE,121(2): 322-338.

Kiam Heong Ang., Chong, G., and Yun Li. (2005). PID control system analysis, design, and technology. IEEE Transactions on Control Systems Technology, 13(4), 559-576.

Tagawa, Y., and Kajiwara K. (2007). Controller development for the E-Defense shaking table. Proceedings of the Institution of Mechanical Engineers. Journal of Systems and Control Engineering, 221(2), 171-181.