航空学报 > 2017, Vol. 38 Issue (2): 220458-220465   doi: 10.7527/S1000-6893.2016.0253

多自由度非高斯随机振动控制

孟韩1, 黄海1, 黄舟2   

  1. 1. 北京航空航天大学 宇航学院, 北京 100083;
    2. 中国工程物理研究院 总体工程研究所, 绵阳 621900
  • 收稿日期:2016-05-20 修回日期:2016-09-07 出版日期:2017-02-15 发布日期:2016-10-09
  • 通讯作者: 黄海,E-mail:hhuang@buaa.edu.cn E-mail:hhuang@buaa.edu.cn

Multi-degree-of-freedom non-Gaussian random vibration control

MENG Han1, HUANG Hai1, HUANG Zhou2   

  1. 1. School of Astronautics, Beihang University, Beijing 100083, China;
    2. Institute of Systems Engineering, China Academy of Engineering Physics, Mianyang 621900, China
  • Received:2016-05-20 Revised:2016-09-07 Online:2017-02-15 Published:2016-10-09

摘要:

在振动试验台上进行多自由度(MDOF)随机振动激励时,传统的控制方法生成的驱动信号及试验台的响应信号都是高斯信号。但真实的振动干扰信号多是超高斯的;而相比于高斯激励,亚高斯激励可降低驱动信号的最大幅值。为实现多自由度亚高斯和超高斯振动控制,提出一种多自由度非高斯随机振动控制方法,该方法采用系统辨识解决系统耦合问题,而后通过选择特殊的相位生成非高斯伪随机驱动信号,再经过时域随机化得到真随机非高斯驱动信号。基于Hexapod平台的多自由度微振动试验台的亚高斯和超高斯实验表明,在试验台的响应功率谱(PSD)满足工程中常用的±3 dB精度的同时,亚高斯驱动信号的最大幅值相比于高斯驱动信号的最大幅值降低了20%以上;超高斯响应信号的峭度与参考峭度的误差在0.2之内。实验结果验证了所提方法的有效性。

关键词: 非高斯随机振动控制, 多自由度, 功率谱密度, 峭度, 相位选择

Abstract:

The drive signal and the response signal generated by traditional multi-degree-of-freedom (MDOF) random vibration control method are both Gaussian signal. However, the real vibration interference signal is always super-Gaussian, while sub-Gaussian random excitation is mainly used to reduce the maximum amplitude of the drive signal. To achieve MDOF sub-Gaussian and super-Gaussian vibration control, an MDOF non-Gaussian random vibration control method is proposed, which solve the coupling problem through system identification, and select special phase to generate non-Gaussian pseudo-random drive signal, and then the pseudo-random drive signal is transformed to real random non-Gaussian drive signal through time domain randomization. The sub-Gaussian and super-Gaussian experiments based on a Hexapod-based MDOF micro vibration test bed show that the response power spectral density (PSD) of response signals obtained by the proposed method are limited to ±3 dB error band of reference PSD. Compared to that in the Gaussian experiment, the drive signal in the sub-Gaussian experiment decreases by more than 20%. In the super-Gaussian experiment, the error between the kurtosis of response signal and the reference value is within 0.2. Effectiveness of the proposed method can be validated by the experiment results.

Key words: non-Gaussian random vibration control, multi-degree-of-freedom, power spectrum density, kurtosis, phase selection

中图分类号: