基于传递率的工作模态参数识别方法的研究

发布时间：2018-05-01 20:34

本文选题：模态参数识别 + 工作模态　；参考：《太原理工大学》2012年硕士论文

【摘要】：大型工程设备、土木结构、机车及航天器等在设计时需要关注其动态特性,在正常工作时会受到意外的超大载荷、风载荷、大地脉动、地震和气流等的影响。结构不合理的动态特性和这些随机的自然激励会引起结构的振动、疲劳以及引发的噪声和破坏。结构优化设计、故障诊断、损伤识别、健康检测等具有重大的工程意义,而模态分析以及模态参数识别是解决这些问题的重要手段。实验模态分析理论成熟,但使用范围相对较小；工作模态分析使用相对广泛,但仍然需要假设激励为白噪声。为了能从非白噪声激励环境中使用工作模态分析方法精确地提取模态参数,本文在基于功率谱的工作模态参数识别方法的基础上,研究了基于传递率的工作模态参数识别方法。在纯白噪声激励时,功率谱方法能够获得清晰的稳态图；而传递率方法的稳态图不够清晰,其极点的数目较多,稳定性较差。在白噪声和谐波共同激励时,功率谱方法会有伪极点出现,并且该伪极点比系统的物理极点更加稳定,这使得系统的物理极点不够明显,甚至会被伪极点淹没；传递率方法没有因为谐波的引入而出现伪极点,相比而言其稳态图更加清晰,保留了纯白噪声激励时的“单峰值”特性。功率谱方法和传递率方法在白噪声与谐波共同激励时都有伪极点：功率谱方法因为噪声问题带来伪极点,且伪极点在该工况下稳定；传递率方法因为传递率函数的构成特点带来伪极点,但伪极点在两种不同工况下不稳定。系统的物理极点与激励无关,并不随着拟合阶次的改变而改变,而伪极点却是与激励有关,会随着拟合阶次的变化而变化。模态稳态图不断改变拟合多项式的阶次,能反映出稳定的极点,达到了识别系统物理极点的目的。研究表明,传递率工作模态参数识别方法和模态稳态图的配合使用,能在谐波干扰下准确地提取系统的模态参数,克服了功率谱工作模态参数识别方法在这方面的不足,达到了研究的目的。
[Abstract]:Large engineering equipment, civil structures, locomotives and spacecraft should pay attention to their dynamic characteristics when they are designed, and will be affected by unexpected super large load, wind load, earth pulsation, earthquake and airflow during normal operation. The unreasonable dynamic characteristics of the structure and these random natural excitations will cause the vibration, fatigue, noise and damage of the structure. Structural optimization, fault diagnosis, damage identification and health detection are of great engineering significance. Modal analysis and modal parameter identification are important means to solve these problems. The experimental modal analysis theory is mature, but the range of application is relatively small, and the working modal analysis is widely used, but it is still necessary to assume that the excitation is white noise. In order to extract modal parameters accurately from non-white noise excitation environment, this paper studies the method of modal parameter identification based on transfer rate on the basis of power spectrum based working modal parameter identification method. Under pure white noise excitation, the power spectrum method can obtain a clear steady-state diagram, while the transfer rate method is not clear enough, the number of poles is more, and the stability is poor. When the white noise and harmonics are excited together, the pseudo poles appear in the power spectrum method, and the pseudo poles are more stable than the physical poles of the system, which makes the physical poles of the system not obvious enough, or even submerged by the pseudo poles. The transfer rate method has no pseudo-pole due to the introduction of harmonics, and its steady-state diagram is more clear, and it retains the "single peak" characteristic of pure white noise excitation. Both the power spectrum method and the transfer rate method have pseudo poles when the white noise and harmonic are excited together. The power spectrum method brings pseudo poles because of the noise problem, and the pseudo poles are stable under this condition. The transfer rate method leads to pseudo poles because of the constitution of the transfer rate function, but the pseudo poles are unstable under two different working conditions. The physical poles of the system are independent of the excitation and do not change with the change of the fitting order, but the pseudo poles are related to the excitation and change with the change of the fitting order. The order of the fitting polynomial is changed continuously by the mode steady state map, which can reflect the stable poles and achieve the purpose of identifying the physical poles of the system. The results show that the combined use of transfer rate working modal parameter identification method and modal steady-state diagram can accurately extract the modal parameters of the system under harmonic interference, which overcomes the shortcomings of the power spectrum working modal parameter identification method in this respect. The purpose of the research has been achieved.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【相似文献】