【作者】 徐士代；

【导师】 汪凤泉；

【作者基本信息】 东南大学 ， 测试计量技术及仪器， 2006， 博士

【摘要】 模态参数识别对于大型工程结构的健康检测、损伤识别、主动(模糊)控制等工程应用具有重要意义。大型工程结构由于体积庞大、材料芸杂、约束条件复杂等原因,难以对其进行人为激励和难以对激励信号进行有效测量,因此,传统的基于输入输出信号识别结构模态参数的理论和技术在大型工程结构中难以适用。环境激励(风荷载、大地脉动、车辆激励等)下的大型工程结构模态参数识别方法,具有无需施加人为激励、仅对输出信号进行力学和数学分析、在时域内或者频域内进行模态参数识别,具有理论上的可信度和工程应用上的可行性,因此该方法在工程界得到了广泛的使用。本文以大型工程结构为研究对象,以近代实验模态分析理论及环境振动识别技术为基础,着重研究了两类基于环境振动的模态参数识别方法。对随机子空间法进行改进,提高了其模态参数识别精度和计算稳定性,并对非平稳激励信号下的随机子空间法模态参数识别进行了研究,拓展了该方法的适用领域。对跨点功率谱法识别结构模态参数进行了深入研究和分析,全面论证了该方法的理论依据,填补了该方法在特征值识别以及其它重要模态参数识别等方面的薄弱环节,使之成为完整有效的模态识别方法,该方法不仅应用方便、适用性广,而且还能有效地应用目前普遍拥有的多通道模态分析系统以及成熟的FFT分析技术。综合这两类时域内和频域内的模态参数识别方法的特点,本文建立起一套基于环境激励下,大型工程结构模态参数识别的理论和应用模式。通过对一些结构的试验测试和仿真计算,验证了本文中给出的模态参数识别方法在理论上具有可信度,在工程应用上具有可行性。全面地研究了环境激励下工程结构基于频域内或者时域内的模态参数识别方法,并对这些方法的理论适用范围和工程应用可行性进行了系统分析。研究了大型工程结构常见激励形式,给出了激励信号平稳性检测的流程。非平稳激励下大型工程结构的模态参数识别方法研究对激励信号的假定更加贴近现实,因此该类别的方法研究是发展方向,本文对此进行了分析和讨论。本文的研究结果显示,本文给出的基于环境激励下的模态参数识别方法能够快速、经济、有效地识别大型工程结构模态参数。更多还原

【Abstract】 The modal parameter identification of large engineering structure has a important sense to applied civil engineering such as health monitoring, structure damage estimate, active control (fuzzy control), etc. It is difficult to exert artificial excitation and difficult to measure excitation signal on large engineering structures for them having huge volume, mixed material, complex bind condition, etc. So it is not easy to use traditional modal parameter identification technology with testing input and output signal on large engineering structures. Having theory reliability and engineering applied feasibility, the ambient excitation method to identify large structure modal parameter is widely used in engineering for it can identify structure modal parameter in time domain or frequency domain only requiring analysis output data with mechanics and mathematics theory.Two modal identification methods undergoing ambient excitation are investigated for the object of large structures based on modern experimental modal analysis theory and ambient vibration technology. Stochastic subspace identification is improved in modal accuracies and in calculating stability. The new stochastic subspace identification method can be used in non-stationary signal domain. Special research and analysis is focused on cross-point spectrum modal parameter identification method in the field of presenting a rounded parameter identification method in which can be used with existent testing instruments and with the FFT technology. After integrating characteristic of the two methods, the dissertation presents a theory and engineering applied mode aiming at modal parameter identification of large engineering structure. The theory reliability and engineering applied feasibility of the modal parameter identification methods presented by the paper are validated through some experiments and simulation computes of engineering structures. The basic theories of the ambient excitation modal parameter identification in time domain or frequency domain are roundly investigated, and their advantages and disadvantages are discussed. Some familiar ambient excitations of large engineering structures are investigated and a flow is presented to examine the input signal’s stationary characteristics. Modal parameter identification of large structure undergoing non-stationary excitation is the next important research domain for the presupposition of excitation signal being close to reality, so it is discussed in this paper.The research results of this paper show that the modal parameter identification undergoing ambient excitation presented by the paper can identify large engineering structure modal parameters fleetly,economy and efficiently.更多还原