Quantitative Identification Of The Cmse And The Structure Of The Niche Ga-based Damage Localization Method Of Research

The large and complex structures always damage because of the loads and erosion from environment when it is serving. The dynamic properties of the damaged structures will change. So the damage detection methods using the change of the dynamic properties have been studied in recent years. In this paper, the vibration-based damage detection methods are studied and the effectiveness of the detection is also checked.In the first part, the finite element dynamics models of the intact and damaged airfoil case are established using ANSYS software, and the modes and frequencies of the case are gotten. Then the concept of the cross modal strain energy is given and the damage index is designed. The damage detection of the cantilever board is emulated. The efficiency and calculation of the CMSE method will be influenced by the complexity and the size of the structure. In the paper, the niche GA is imported to improve the efficiency of the detection. The emulation of the cantilever airfoil case indicates that the detection method based on CMSE and Niche GA can detect the single and complex damage of the structure, which the damage degree is small. And the error of the detection damage degree is very small. Compared with the ordinary model strain energy methods, the CMSE method needs a little mode information; even only one mode of the damaged structure is needed. The analytical and the measured modes are also not required to be consistent in scale, or to be normalized.In the second part, how to extract the damage information in the online damage detection is studied. If the structural damage occurs when it is serving, the collected signal will have impulses or breaks. So the study purpose is to extract the transient part of the collected signal. Because of the uncertainty of the signal, the blind signal processing is imported which is the pop study in recent years. In the paper, the part contains the damage information is extracted using filter, predictive theory, short time Fourier transform and spectral kurtosis. The signal processing is emulated using the MATLAB. The results indicate that the method can extract the abrupt damage information of the serving structure and confirm the damage time. The work can help detect the damage location and degree.