基于HHT变换的非线性结构损伤识别研究

发布时间：2018-10-10 14:13

【摘要】：由于材料、连接和变形等因素的非线性导致实际结构都不同程度地存在非线性行为。特别是当结构遭受损伤时，损伤可能导致结构非线性加剧，并造成结构动态响应不满足平稳性假定。结构的损伤和非线性行为导致基于线性、平稳性假定的结构损伤识别方法不再适用。本文基于经验模态分解和Hilbert-Huang Transform（HHT），研究了基于HHT的非线性系统损伤识别方法。主要内容包括：对非线性系统识别方法进行了比较系统的归纳和综述，主要包括时域方法、频域方法、时频域方法和基于混沌的方法。本文研究采用时频域的HHT方法。采用集成经验模态分解方法，针对模态分解过程中筛选次数恒定，可能会导致分解结果不精确、信号分解过程缺乏自适应性的缺点，基于Huang提出的改进差值和函数方法，并结合筛选信号极值点与过零点数目差最大为1的要求，对集成经验模态分解方法的筛选停止条件进行了改进。通过算例分析，验证了改进方法增加了信号分解的精确性及自适应性。分析比较了端点效应影响范围和时间特征尺度大小之间的关系，，信号的高频分量时间特征尺度小，低频分量时间特征尺度大。验证了端点效应对于信号的高频分量相对于低频部分影响较小，因此提出在不需要整个信号特征时，可以选取高阶固有模态函数分量作为分析对象。采用改进的希尔伯特-黄变换分析方法，并以基于系统输出信号第一阶固有模态函数分量的瞬时频率作为损伤指标，对地震波作用下的二维双线性框架结构的损伤进行识别。通过瞬时频率的时域变化，能够有效识别结构非线性损伤及损伤时刻。通过在三维框架模型底层引入非线性元件来模拟非线性，杆件截面变化模拟损伤，将改进的HHT变换用于非线性框架结构的损伤识别。利用结构顶层输出信号来进行损伤识别，将系统输出信号第一阶固有模态函数分量的瞬时频率及瞬时能量作为损伤指标。结果表明：根据结构对称性选择合适的信号，利用响应第一阶固有模态函数分量的瞬时频率和瞬时能量可以有效检测出损伤的存在，并能进行损伤的程度识别。
[Abstract]:Due to the nonlinearity of material, connection and deformation, the nonlinear behavior of the actual structure exists in varying degrees. Especially when the structure is damaged, the damage may lead to the aggravation of the nonlinearity of the structure, and the dynamic response of the structure does not satisfy the assumption of stationarity. Due to the damage and nonlinear behavior of structures, the method of structural damage identification based on linear and stationary assumption is no longer applicable. Based on empirical mode decomposition (EMD) and Hilbert-Huang Transform (HHT), a method of nonlinear system damage identification based on HHT is studied in this paper. The main contents are as follows: the methods of nonlinear system identification are summarized and summarized systematically, including time domain method, frequency domain method, time and frequency domain method and chaos based method. The HHT method in time-frequency domain is studied in this paper. Based on the improved difference and function method proposed by Huang, the integrated empirical mode decomposition method is used to solve the problem that the filtering times are constant in the process of modal decomposition, which may lead to the imprecision of decomposition results and the lack of adaptability in the process of signal decomposition. According to the requirement that the maximum difference between the extreme point and the zero crossing point of the filter signal is 1, the screening stopping condition of the integrated empirical mode decomposition (EMD) method is improved. An example is given to verify that the improved method increases the accuracy and adaptability of signal decomposition. The relationship between the influence range of the end point effect and the size of the time characteristic scale is analyzed and compared. The time characteristic scale of the high frequency component is small and the time characteristic scale of the low frequency component is large. It is verified that the endpoint effect has little effect on the high frequency component of the signal compared with the low frequency part, so it is proposed that the higher order inherent mode function component can be selected as the object of analysis without the need of the whole signal feature. Using the improved Hilbert-Huang transform method and the instantaneous frequency of the first order natural mode function component of the system output signal as the damage index, the damage of the two-dimensional bilinear frame structure subjected to seismic waves is identified. The time domain variation of instantaneous frequency can effectively identify the nonlinear damage and damage time of the structure. The nonlinear element is introduced into the bottom layer of the three-dimensional frame model to simulate the nonlinearity and the change of the member section is used to simulate the damage. The improved HHT transform is applied to the damage identification of the nonlinear frame structure. The structural top output signal is used to identify the damage. The instantaneous frequency and instantaneous energy of the first order natural mode function component of the output signal of the system are taken as the damage index. The results show that the existence of damage can be effectively detected by using the instantaneous frequency and instantaneous energy of the component of the first order natural mode function according to the symmetry of the structure and the degree of damage can be identified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TU317

【参考文献】