基于在线神经网络算法的混合试验方法

王 涛1,2,翟绪恒2,孟丽岩2,王 贞3

振动与冲击 ›› 2017, Vol. 36 ›› Issue (14) : 1-8.

PDF(3017 KB)
PDF(3017 KB)
振动与冲击 ›› 2017, Vol. 36 ›› Issue (14) : 1-8.
论文

基于在线神经网络算法的混合试验方法

  • 王 涛1,2,翟绪恒2,孟丽岩2,王 贞3
作者信息 +

Hybrid testing method based on online neural network algorithm

  • WANG Tao1,2,ZHAI Xuheng2,Meng Liyan2, WANG Zhen3
Author information +
文章历史 +

摘要

混合试验是一种将数值模拟与物理试验相结合的新兴结构抗震试验方法,得到了相关研究者们的广泛关注。如何模拟具有强非线性的数值子结构仍是混合试验亟待解决的问题。本文在传统的离线神经网络基础上提出一种在线学习的神经网络算法,并应用于混合试验中来在线预测数值子结构恢复力。在线学习算法仅利用当前步的系统输入和观测样本,采用递推形式更新每一步的权值和阈值。针对两个自由度非线性结构,分别进行了基于在线学习和离线学习神经网络的混合试验数值仿真。研究表明:与离线学习神经网络算法相比,在线学习神经网络算法具有更好的自适应性,能够有效提高恢复力预测精度和计算效率;基于在线学习神经网络算法的结构混合试验方法可以提高混合试验结果精度。

Abstract

Hybrid testing is an advanced structure seismic experimental method that combines numerical simulation and physical testing and is increasingly being recognized by researchers. One of the challenging issues is how to model the numerical substructure with strong nonlinearity. An online learning neural network algorithm (NN) is proposed based on conventional offline NN algorithm and applied in the hybrid testing to online predict the restoring force for the numerical structure. Weights and thresholds of the novel algorithm can be updated using recursive form only based on the current step systematic inputs and observations. Numerical simulations of hybrid testing based the online and the offline NN algorithms are conducted for a two freedom nonlinear system. The results show the new online NN hybrid testing method has better adaptation, computational efficiency and prediction accuracy for restoring force, compared with offline NN hybrid testing method. In general, the online NN hybrid testing method can effectively improve the accuracy of hybrid testing results.

 

关键词

混合试验 / 神经网络 / 在线预测 / 恢复力

Key words

 hybrid testing method / neural network / online prediction / restoring force

引用本文

导出引用
王 涛1,2,翟绪恒2,孟丽岩2,王 贞3. 基于在线神经网络算法的混合试验方法[J]. 振动与冲击, 2017, 36(14): 1-8
WANG Tao1,2,ZHAI Xuheng2,Meng Liyan2, WANG Zhen3. Hybrid testing method based on online neural network algorithm[J]. Journal of Vibration and Shock, 2017, 36(14): 1-8

参考文献

[1] Wu B, Bao H, Ou J, et al. Stability and accuracy analysis of the central difference method for real-time substructure testing[J]. Earthquake Engineering and Structural Dynamics, 2005, 34:705-718. 
[2] Jung R, Shing P, Stauffer E, et al. Performance of a real-time pseudo dynamic test system considering nonlinear structural response[J]. Earthquake Engineering and Structural Dynamics, 2007, 36:1785-1809.
[3] 王贞. 实时混合试验的控制和时间积分算法[D]. 哈尔滨: 哈尔滨工业大学, 2012:1-193.
   WANG Zhen. Control and time integration algorithms for real-time hybrid simulation[D]. Harbin: Harbin Institute of Technology. 2012:1-193.
[4] 曾 聪,许国山,张树伟,王 涛,耿 悦,吴 斌. 力-位移混合控制方法在大型多功能试验加载系统拟静力试验中的应用[J]. 振动与冲击,2016,35(7):161-166.
   Zeng Cong, Xu Guoshan, Zhang Shuwei, Wang Tao, Geng Yue, Wu Bin. Application of force-displacement hybrid control rnethod in quasi-static tests of a rnrtlti一funcaional testing ystcto
[5] Wu B, Wang Z, Bursi O. Actuator dynamics compensation based on upper bound delay for real-time hybrid simulation [J]. Earthquake Engineering and Structural Dynamics, 2013, 42(12):1749-1765.
[6] Chang S. Nonlinear error propagation analysis for explicit pseudo dynamic algorithm[J]. Journal of Engineering mechanics, 2003,129(8):841-850.
[7] 杨 格, 王 贞, 吴 斌, 杨 婧, 许国山, 陈永盛. 建筑结构混合试验平台HyTest开发研究[J]. 建筑结构, 2015,36(11):149-155.
   YANG Ge, WANG Zhen, WU Bin, YANG Jing, XUG Guo-Shan, CHEN Yong-sheng. Development of HyTest for structural hybrid simulation[J]. Building Structures. 2015,36(11):149-155.
[8] Kwon O, Kammula V. Model updating method for substructure pseudo-dynamic hybrid simulation[J]. Earthquake Engineering and Structural Dynamics, 2013, 42(13): 1971-1984.
[9] Hashemi M, Masroor A, Mosqueda G. Implementation of online model updating in hybrid simulation[J]. Earthquake Engineering and Structural Dynamics, 2014, 43(3): 395-412.
[10] Wu B, Wang T. Model updating with constrained unscented Kalman filter for hybrid testing[J]. Smart Structures and Systems, 2014.14(6):1105-1129.
[11] 王 涛. 基于模型更新的土木结构混合试验方法[D].哈尔滨:哈尔滨工业大学,2014.
    Wang Tao. Hybrid testing method for civil structures based on model updating[D]. Harbin: Harbin Institute of Technology. 2014.
[12] 王 涛, 吴 斌. 基于UKF模型更新的混合试验方法[J]. 振动与冲击, 2013,32(5):100-109.138-143.
    Wang Tao, Wu Bin. Model updating for hybrid testing with unscented Kalman filter[J]. Vibration and Shock. 2013,32(5):100-109.138-143.
[13] Wu B, Chen Yongsheng, Xu Guoshan, Zhu Mei, Pan Tianlin, Zeng Cong. Hybrid simulation of steel frame structures with sectional model updating[J]. Earthquake Engineering & structural Dynamiacs.2016, 45  (8):1251-1269.
[14] 庄镇泉. 神经网络与神经计算机[M]. 北京: 科学出版社. 1996.
    ZHUANG Qing-quan. Neural network and Neural computer[M]. Beijing: Science Press. 1996.
[15] 涂建维, 戴 葵, 瞿伟廉. 磁流变阻尼器的磁滞效应与神经网络预测调整[J].华中科技大学, 2007, 35 (3): 110-112.
    TU Jian-wei, DAI Kui, QU Wei-lian. Magnetic hysteresis of magnetorheological dampers and its compensation by neural network prediction[J]. J. Huazhong Uni. Of Sci.&Tech, 2007, 35 (3): 110-112.
[16] 周大兴, 闫维明,陈彦江,唐贞云,李振宝.神经网络在振动子结构试验中的应用[J].振动与冲击. 2011, 30 (12): 14-18.
     ZHOU Da-xing, YAN Wei-ming, CHEN Yan-jiang, TANG Zhen-yun, LI Zhen-bao. Application of neural network in real-time substructure testing with shaking table[J]. Vibration and Shock. 2011, 30 (12): 14-18.
[17] YANG W J,  NAKANO Y. Substructure online test by using real-time hysteresis modeling with a neural network[J]. Advances in Experimental Structural Engineering, 2004(38): 267-274.
[18] YUN G Y, GHABOUSSI J, ELNASHAI A S. A new network-based model for hysteretic behavior of materials[J]. International Journal for Numerical Methods in Engineering, 2008, 73(4): 447-469.
[19] 张 健. 自适应子结构拟动力试验方法[D]. 哈尔滨:哈尔滨工业大学,2010.
    ZHANG Jian. Adaptive substructure pseudo-dynamic testing method[D]. Harbin: Harbin Institute of Technology. 2010.
[20] Ma F, Zhang H, Bockstedte A, et al. Parameter analysis of  the differential model of hysteresis[J]. J. Appl. Mech. Trans. ASME, 2004, 71(3): 342-349.

PDF(3017 KB)

Accesses

Citation

Detail

段落导航
相关文章

/