基于深度强化学习的拉索智能减振算法

陈孝聪1,张恩启1,程斌1,王浩2

振动与冲击 ›› 2022, Vol. 41 ›› Issue (23) : 175-181.

PDF(2478 KB)
PDF(2478 KB)
振动与冲击 ›› 2022, Vol. 41 ›› Issue (23) : 175-181.
论文

基于深度强化学习的拉索智能减振算法

  • 陈孝聪1,张恩启1,程斌1,王浩2
作者信息 +

Intelligent vibration reduction algorithm of cable based on deep reinforcement learning

  • CHEN Xiaocong1,  ZHANG Enqi1, CHENG Bin1, WANG Hao2
Author information +
文章历史 +

摘要

为了降低振动控制算法对拉索-MR阻尼器系统动力学模型精度的依赖性,提出一种基于深度强化学习理论的无模型减振算法。该方法利用控制模块与环境之间的交互实现对拉索振动的自适应半主动控制,依据拉索特定点响应状态在线调节阻尼器施加电压,降低反馈控制要求。为验证智能控制算法的有效性,采用Galerkin法建立拉索-MR阻尼器环境模型,并以实桥拉索减振设计为例对比分析了粘滞阻尼器多模态控制、Bang-Bang控制、深度强化学习控制对拉索的风振控制效果。结果表明:在随机风荷载作用下,深度强化学习控制算法不仅能够实现对拉索的无模型振动控制,且控制效果优于粘滞阻尼器多模态控制和MR阻尼器的Bang-Bang控制。
关键词:拉索;振动控制;深度强化学习;MR阻尼器;风致振动

Abstract

In order to reduce the dependence on the accuracy of cable-MR damper system’s dynamic model, a model-free vibration control algorithm was proposed based on deep reinforcement learning theories. The algorithm utilized the interaction between the control module and environment model to control vibrations adaptively. The input voltage of MR damper was adjusted online according to the responses at specific location of the cable, which reduced feedback requirements. In order to verify the effectiveness of proposed intelligent control algorithm, Galerkin method was used to establish the environment simulation model, and comparative analysis of an actual stay cable vibration responses under optimal viscous damper multi-mode vibration control, Bang-Bang control and deep reinforcement learning control strategies was conducted. The results show that the deep reinforcement learning control algorithm can achieve model-free vibration control of the stay cable, and its control efficiency is better than the optimal passive control of viscous damper and Bang-Bang control of MR damper.
Key words: cable; vibration control; deep reinforcement learning; MR damper; wind-induced vibration

关键词

拉索 / 振动控制 / 深度强化学习 / MR阻尼器 / 风致振动

Key words

cable / vibration control / deep reinforcement learning / MR damper / wind-induced vibration

引用本文

导出引用
陈孝聪1,张恩启1,程斌1,王浩2. 基于深度强化学习的拉索智能减振算法[J]. 振动与冲击, 2022, 41(23): 175-181
CHEN Xiaocong1, ZHANG Enqi1, CHENG Bin1, WANG Hao2. Intelligent vibration reduction algorithm of cable based on deep reinforcement learning[J]. Journal of Vibration and Shock, 2022, 41(23): 175-181

参考文献

[1] Pacheco B M, Fujino Y, Sulekh A. Estimation curve for modal damping in stay cables with viscous damper[J]. Journal of Structural Engineering, 1993, 119(6): 1961-1979.
[2] Wang Z H, Gao H, Xu Y W, et al. Vibration control of a stay cable with a rotary electromagnetic inertial mass damper[J]. Smart Structures and Systems, 2019, 23(6): 627-639.
[3] Li Y M, Shen W A, Zhu H P. Vibration mitigation of stay cables using electromagnetic inertial mass dampers: Full-scale experiment and analysis [J], Engineering Structures, 2019, 200: 109693.
[4] Chen L, Nagarajaiah S, Sun L M. A unified analysis of negative stiffness dampers and inerter-based absorber for multimode cable vibration control[J]. Journal of Sound and Vibration, 2021,494:115814.
[5] 汪志昊,寇琛,许艳伟,等. 自供电MR阻尼器复合减振系统对斜拉索振动控制试验研究[J]. 振动与冲击,2019, 38(10): 1-5.
WANG Zhihao, KOU Chen, XU Yan wei, et al. Experimental study on the cable vibration control using a self-powered MR damper hybrid system[J]. Journal of Vibration and Shock, 2019, 38(10):1-5.
[6] Xu Y W, Xu Z D, GUO Y Q, et al. Single input magnetorheological pseudo negative stiffness control for bridge stay cables[J]. Smart Materials and Structures, 2021, 30(1):1-14.
[7] Johnson E A, Christenson R E, Spencer B F. Semiactive damper of cables with sag[J]. Computer-Aided Civil and Infrastructure Engineering, 2003, 18(2):132-146.
[8] Zhou H J and Sun L M. Damping of stay cable with passive-on magnetorheological dampers: a full-scale test[J]. International Journal of Civil Engineering, 2013, 11(3): 154-159.
[9] Duan Y F, Ni Y Q, Zhang H M,et al. Design formulas for vibration control of taut cables using passive MR dampers[J], 2019, 23(6): 521-536.
[10] 李惠,刘敏,欧进萍,等. 斜拉索磁流变智能阻尼控制系统分析与设计[J]. 中国公路学报,2005, 18(4): 37-41.
LI Hui, Liu Min, Ou Jingping, et al. Design and analysis of magnetorheological dampers with intelligent control systems for stay cables[J]. China Journal of Highway and Transport, 2005, 18(4): 37-41.
[11] Weber F, Distl H. Semi-active damping with negative stiffness for multi-mode cable vibration mitigation: approximate collocated control solution[J]. Smart Materials and Structures, 2015, 24(11), 115015.
[12] Chen Z H, Lam K H, Ni Y Q. Enhanced damping for bridge cables using a self-sensing MR damper[J]. Smart Materials  and Structures, 2016, 25(8), 085019.
[13] Zhao Y L, Xu Z D, Wang C. Wind vibration control of stay cables using magnetorheological dampers under optimal equivalent control algorithm[J]. Journal of Sound and Vibration, 2019, 443(4): 732-747.
[14] Javanbakht M, Cheng S H, Ghrib F. Control-oriented model for the dynamic response of a damped cable[J]. Journal of Sound and Vibration. 2019, 442:249-267.
[15] 邬喆华,楼文娟,陈勇,等. MR阻尼器半主动控制对拉索减振效果的仿真分析[J]. 工程力学,2007, 24(10):11-18.
WU Zhehua, LOU Wenjuan, CHEN Yong, et al. Simulation analysis for vibration mitigation of stay-cable by MR dampers under semi-active control strategy[J]. Engineering Mechanics, 2007, 24(10):11-18.
[16] Zhou Q, Nielsen S R K, Qu W L. Semi-active control of three-dimensional vibrations of an inclined sag cable with magnetorheological dampers[J]. Journal of Sound and Vibration, 2006, 296(12): 1-22.
[17] Liu M, Song G, Li H. Non-model-based semi-active vibration suppression of stay cables using magneto-rheological fluid dampers[J], Smart Materials and Structures, 2007, 16(4): 1447-1452.
[18] Sutton R S, Barto A G, Reinforcement learning: An Introduction[M]. Cambridge, USA: MIT Press, 2018.
[19] Minh V, KavuKcuoglu K, Silver D,et al. Human-level control through deep reinforcement learning[J]. Nature, 2015, 518(7540): 529-533.
[20] Hasselt H V, Guez A, Silver D. Deep reinforcement learning with double Q-learning[C]//Proceedings of the 30th AAAI Conference on Artificial Intelligence. Phoenix: AAAI, 2016: 2094-2100.
[21] Wang Z Y, Schaul T, Hessel M,et al. Dueling network architectures for deep reinforcement learning[C]// Proceeding of the 33rd International Conference on Machine Learning. New York: ACM Press, 2016, 1995-2003.
[22] 王修勇,陈政清,倪一清. 斜拉桥拉索风雨振观测及其控制[J]. 土木工程学报,2003, 36(6): 53-59.
WANG Xiuyong, CHEN Zhengqing, Ni Yiqing. Wind-rain induced vibration and its control on stay cables of cable-stayed bridges[J]. China civil engineering journal, 2003, 36(3): 53-59.
[23] 中交公路规划设计院. 公路桥梁抗风设计规范[M]. 北京:人民交通出版社, 2018.
[24] Spencer B F, Dyke S J, Sain M K, et al. Phenomenological model for magnetorheological dampers[J]. Journal of Engineering Mechanics, 1997, 123(3): 230-238.
[25] Liao W H, Lai C Y. Harmonic analysis of a magnetorheological damper for vibration control[J]. Smart Materials And Structure, 2002, 11(2), 288–296.

PDF(2478 KB)

395

Accesses

0

Citation

Detail

段落导航
相关文章

/