【作者】 陈建群；

【导师】 段晨东；

【作者基本信息】 长安大学 ， 环境工程， 2009， 硕士

【摘要】 土木结构服役期间由于荷载和其他有害效应的作用可能损害建筑结构,造成重大的损失和安全事故。对已建成的结构和设施采取有效的手段进行长期的结构健康监测和损伤诊断具有重要的科学理论意义和工程应用价值。结构损伤发生后将导致结构固有频率、刚度和阻尼等模态参数的改变。因此,模态参数的改变可视为结构损伤发生的标志,可以利用损伤前后结构模态参数的变化来检测结构损伤。本文研究了一种基于信号处理的结构模态参数提取方法,为结构健康状态检测提供依据。从模态参数提取角度出发,研究了基于Laplace小波的相关滤波方法的特性。通过仿真实验研究相关滤波各种参数对滤波过程的影响,并将此方法应用于工程模型,研究证明小波分析是一种相关分析过程,只有小波函数与待提取的特征匹配时,才能取得较大的小波系数,凸现信号中的隐含的特征信息。因此,选择小波函数是特征提取的关键。具有单边衰减特性的Laplace小波与结构冲击信号的响应波形相匹配。可识别出结构的模态参数。当结构的各阶响应交叠在一起时对模态参数的提取会产生不利的影响,提取出的模态参数,与信号中各阶模态分量的幅值大小有关。由于实际结构的振动响应信号是各阶模态响应的综合,直接对实际结构冲击响应信号进行Laplace小波相关滤波难以正确的提取结构的模态参数。研究EMD方法分解原理和算法,并对EMD分解方法存在的问题进行了探讨,研究了EEMD方法及其特点;通过实验验证了EMD和EEMD方法自适应分解信号的一些特点。并根据实际结构振动信号的特点,提出了基于EEMD的相关滤波的参数提取方法。首先采用EEMD方法从结构的冲击响应信号中提取出结构的某阶响应信号,避免各阶响应信号交叠在一起对相关滤波造成的不利影响。再对其分解出的固有模式分量进行相关滤波以提取结构的某阶模态参数。将此方法应用于工程模型的各种损伤工况,准确的提取了结构的三阶模态频率。实验证明:当结构出现损伤时其固有频率将会改变,对于不同类型的损伤,结构的固有频率有显著的差别。因此,结构模态参数的变化可以反映结构的健康状态,采用本方法在线提取结构的模态参数,可以为结构的健康状态的在线监测和诊断提供依据。更多还原

【Abstract】 Damage will appear in the civil engineering structures during their lifetime with the effect of loads and other harmful factors. As a result, it sometimes will bring about significant economic losses and safety incidents. So it is necessary to make efficient diagnosis, evaluation and prognosis for the health condition of serving structures.The structure modal parameters, such as nature frequency, stiffness, damping parameter can be changed when damage occurs. So the modal parameter variety may be regarded as the index that the structure damage occurs, we may use the structure modal parameter variety to detect the different structure damage. In this paper the damage feature extraction techniques for engineering structure based on signal processing method are studied to provide a basis for structural health monitoring.In order to extract modal parameters, a structure damage detection method based on the Laplace wavelet correlation filtering is studied. The effect of the various parameters on the correlation filtering process is also studied through the simulation. By making correlation filtering to the response signal of the ASE structure’s impulse vibration, conclusions were as follows: Wavelet analysis is a process of correlation analysis, larger wavelet coefficients can be obtained only when the Wavelet function matches with characteristics of the feature to be extracted. Therefore, the choice of wavelet function is crucial to the feature extraction. The Laplace wavelet which has a unilateral attenuation characteristic resembles the impulse response signal of the structure. When using the Laplace wavelet correlation filtering, it treats the target as a single degree of freedom system, so the result will be one modal parameter, not all modals. It will have bad effect on modal parameter extraction when all modals responses mixed together. In this case, the extracted modal parameter is obtained depends on their amplitude. Because of the vibration response signal of the actual structure is as the situation mentioned above, it is difficult to use the Laplace wavelet correlation filtering to extract modal parameter correctly.The principle, algorithm and problems of the Empirical Mode Decomposition （EMD） method are studied. In order to fix the Mode Mixing Problem of the EMD, the Ensemble Empirical Mode Decomposition method （EEMD） and its characteristics are introduced. Some issues in using the EMD and EEMD method to decompose signals by simulation are investigated. Due to all modals responds mixed together in vibration response signal of the actual structure, a modal parameter extraction method based on EEMD and the Laplace wavelet correlation filtering are developed. firstly, use the EEMD method to decompose the vibration response signal to get a modal response signal from the mixed responds of all modals, which can eliminate the bad effect of the mixed modals responds. Then the Laplace wavelet correlation filtering is adapted to the IMF which related to the modal response signal to get the right modal parameter. By using this method to different cases of the ASCE structure, the 3^rd order nature frequency is extracted accurately. It’s shown by experiments that: The structure modal parameters will change when damage occurs, nature frequency of the structure have significant difference between different damage situations. So the changes of the structure modal parameters can reflect the situation of structure health condition. Through using this method to extract modal parameters of structure online, the result can give the evidence for the assessment of structure healthy condition.更多还原