材料的密度、杨氏弹性模量和泊松比分别为2700 kg/m3,7.0×1010 Pa和0.3。结构被离散为50个节点,由于结构受到气动弹性力作用,这里设置对结构的刚度矩阵做了修改。结构的示意图和有限元离散图如图4所示。
结构的固有频率和阻尼比如表2所示,相对应的有限元模态振型如图5所示。
记录数据的采样频率fs=600 Hz,两个激励服从正态分布N(0,106)施加在节点6和50。仿真时间ts和相关函数信号长度tcor分别为7200 s和4 s。不失一般性,设置节点1的位移响应信号为参考信号,计算位移相关函数R1(T)。
首先验证了相关函数与自由衰减响应的等效性,不失一般性,图6展示第10点与第20点对应的相关函数与自由衰减响应。
图6中相关函数和自由衰减响应完全吻合,进一步证明了相关函数与自由衰减响应的等效性。
下面根据节点的随机响应进行辨识,首先使用所有节点响应对节点1做相关函数,得到位移相关函数向量,接着使用ERA辨识算法对采集的位移响应进行初步辨识,最后使用本文提出的IULS算法对响应信号进行辨识。ERA稳态图如图7所示。
由图7可见,对于较为复杂的结构稳定极点的选取仍然依靠人工判断。通过图7的稳态图选取迭代的初值,利用IULS算法估计模态参数。含有能量贡献的结构模态振型可以根据式(37)得到。估计的频率和阻尼比如表3所示。
由表3可见IULS对初值的精度有一定的提高,可以实现更为精准的模态参数辨识。重要的是,该方法在稳定极点选择初值不够理想的前提下仍然可以收敛到满意的结果。
4 结 论
本文首先基于复模态理论推导了非自伴随系统在宽频随机和自激反馈力作用下的响应,并且通过响应相关函数的理论推导证明了相关函数与自由衰减响应的等价性。在相关函数与自由衰减响应等价前提下,推导了针对自由衰减响应的IULS辨识算法。最后通过了桥梁节段模型和机翼模型的仿真验证了理论的可靠性。本文的结论如下:
1.从理论上证明了非自伴随系统响应的相关函数与给定初始条件下的自由衰減响应等价。
2.针对随机响应的相关函数引入了IULS算法,实现了更为精确的模态参数辨识。
3.通过两自由度桥梁节段模型和机翼型板的有限元模型验证了响应的相关函数与自由衰减响应的等价性以及辨识算法的有效性。
参考文献:
[1] 刘海标, 宋汉文. 主动结构动力学特征与频响特性研究[J]. 振动与冲击, 2014, 33(22):121—126+145.
Liu Haibiao, Song Hanwen. Dynamic characteristics and frequency response features of active structures[J]. Journal of Vibration and Shock, 2014, 33(22):121—126+145.
[2] 刘志文, 吕建国, 刘小兵,等. 串列双幅断面颤振稳定性气动干扰试验研究[J]. 振动工程学报, 2016, 29(3):403—409.
Liu Zhiwen, Lü Jianguo, Liu Xiaobing, et al. Experimental investigations of aerodynamic interference effects on flutter stability of cylinders in tandem arrangement[J]. Journal of Vibration Engineering, 2016, 29(3):403—409.
[3] Garrick I E, Reed W H. Historical Development of Aircraft Flutter[J]. Journal of Aircraft, 1981, 18(11):897—912.
[4] 张景绘, 龚 靖, 王永刚. 线性主动结构及模态(Ⅰ)——基本概念及属性[J]. 应用数学和力学, 2004, 25(8):771—778.
Zhang Jinghui, Gong Jing, Wang Yonggang. Linear active structures and modes(Ⅰ) : Basic concepts and properties[J]. Applied Mathematics and Mechanics, 2004, 25(8):771—778.
[5] 郭其威, 吴 松, 刘 芳,等. 航天器模态分析—试验体系工程实践研究[J]. 动力学与控制学报, 2014, (3):274—278.
Guo Qiwei, Wu Song, Liu Fang, et al. Research on Engineering Practice of Modal Analysis-Test of Spacecraft[J]. Journal of Dynamics and Control, 2014, (3):274—278.
[6] 谭万军, 杨 亮, 吴行让,等. 基于ODS与试验模态分析的方向盘摆振优化[J]. 振动工程学报, 2011, 24(5):498—504.
Tan Wanjun, Yang Liang, Wu Xingrang, et al. Steering wheel shimmy optimization based on ODS analysis and experimental modal analysis[J]. Journal of Vibration Engineering, 2011, 24(5):498—504.
[7] Brincker R. On the application of correlation function matrices in OMA[J]. Mechanical Systems and Signal Processing, 2017, 87:17—22.
[8] Reynders E, Maes K, Lombaert G, et al. Uncertainty quantification in operational modal analysis with stochastic subspace identification: validation and applications[J]. Mechanical Systems and Signal Processing, 2016, 66—67:13—30.
[9] 谭德先, 周 云, 米斯特,等. 环境激励下高层建筑结构模态测试与有限元建模分析[J]. 土木工程学报, 2015, 48(9):41—50.
Tan Dexian, Zhou Yun, Mi Site, et al. Ambient vibration dynamic test and finite element analysis for high-rise buildings[J]. China Civil Engineering Journal, 2015, 48(9):41—50.
[10] Gu M, Zhang R X, Xiang H F. Identification of flutter derivatives of bridge decks[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2000, 84(2):151—162.
[11] Bartoli G, Contri S, Mannini C, et al. Toward an improvement in the identification of bridge deck flutter derivatives[J]. Journal of Engineering Mechanics-ASCE, 2009, 135(8):771—785.
[12] Boonyapinyo V, Janesupasaeree T. Data-driven stochastic subspace identification of flutter derivatives of bridge decks[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2010, 98(12):784—799.
Abstract: Non-self-adjoint dynamical system commonly appears in rotor dynamics, flutter analysis and control synthesis, where the symmetry of the system matrices are destroyed. The asymmetry of the system matrices leads to challenges to system identification when the difference arises between the right and left eigenvectors corresponding to the same eigenvalue. The identification of non-self-adjoint system is of great importance for the prediction of flutter boundary, the identification of control law, the optimal design of structures etc. However, for the non-self-adjoint system in engineering (e.g. bridge flutter, the aerodynamic drag forces acting on airplane wings and fuselages, the forces acting on the rotor in turbines, brake system of a vehicle), the identification is based on the output data of the system because of the unknown input data. This research concerns the operational modal analysis (OMA) of a typical non-self-adjoint system. Specifically, the equivalence between the correlation functions of random responses and the free decay responses of the original structure is proved for the non-self-adjoint system. The ERA method is applied to reconstruct the non-self-adjoint system. Case examples on the identification of a six-degree-of-freedom system and the flutter derivatives of bridge sections are performed to validate the method.
Key words: non-self-adjoint dynamic system; system identification; operational modal analysis; asymmetry
作者簡介: 陈 伟(1994—),男,博士研究生。电话: 15317058025; E-mail: meshiawei@tongji.edu.cn
通讯作者: 宋汉文(1962—),男,教授,博士生导师。电话: 18019787293; E-mail: hwsong@tongji.edu.cn